Talk:Retrospect (software)

Wrapping this up

The discussion was endless and not productive. I decided to take the bull by the horns and reduce the article to a manageable and appropriate size. I might've deleted one or two (one or two) important "selected" features - if so, let's discuss. I acknowledge that I broke a bunch of refs - I'll get to fixing those, but of course anyone who wants to get started on it sooner is welcome to. Thanks. JohnInDC (talk) 01:34, 27 September 2017 (UTC)

I've fixed the refs. I'll also note that I didn't do anything yet to the "Editions and Add-ons" section, which is probably too much as well. JohnInDC (talk) 01:40, 27 September 2017 (UTC)

I think JohnInDC took the bull by the horns about an hour before I would have posted a proposal to guide the bull through a chute to a mutually-acceptable pasture. What I realized earlier this evening is that we have all been suffering from a conceptual problem. I agree that the "Main features" section was too big for an article on a particular backup app. Instead most of the items describe features that distinguish an enterprise backup app from a personal backup app. Look—one-by-one—at the "Main features" described in this permalink of the article before JohnInDC's edits tonight, and see if you can find them in the Backup article. In most cases you won't be able to, because they are not features required in even a full-featured personal backup app. They are instead features required by an administrator responsible for keeping an entire multi-computer installation centrally and thoroughly backed up (including rapid-recovery off-site storage—which IMHO doesn't include cloud unless you are budgeted for an AWS Snowball—see 10/2015 and 11/2016).

What should have occurred to me at least a week ago is that there needs to be a separate article named "Enterprise Backup" or something similar. It would contain those "Main features" that aren't in the Backup article. I would initially write the new article using Retrospect terminology—because I don't currently know any alternative, but would otherwise mention Retrospect as little as possible. Because I did some research on Tolis BRU a few months ago, I could write an article on that backup app—which greatly resembles a Mac/Linux Retrospect clone feature-frozen around 2006—with links to the "Enterprise Backup" article. I would then write an email to Tolis Group telling them I'd started a WP article (which they don't already have) about their app, and inviting them to improve it. This would start a terminology and completeness battle about the "Enterprise Backup" article, which IMHO is what we need. At some point or other, subject to your OK, I would enhance the Retrospect article so that its "Main features" section would also refer to the "Enterprise Backup" article. Hopefully we'd get knowledgeable editors to write about Windows and Linux enterprise backup apps in the same vein, which would undoubtedly unleash further helpful terminology and completeness battles over the "Enterprise Backup" article.

All in all, I'm not too unhappy about what JohnInDC did tonight. The only thing I'm rather annoyed about is his deletion of the last paragraph and the translation table in the "History" section. That's fundamentally unfair, because it conceals the fact that modern Retrospect Windows and Retrospect Mac are two different apps—with different user interfaces and different modes of operation that happen to share 99% of their underlying code (in answer to scope_creep's challenge, Retrospect Windows can write to "superfloppies" and has an Immediate mode—facilities that Retrospect Mac has dropped). However, at the time he did the deletion, JohnInDC had not seen my reply to his claim that the last "History" prgf. contained "unsourced claims, OR, synthesis or opinion"—it doesn't. DovidBenAvraham (talk) 04:36, 27 September 2017 (UTC)

Hi JohnInDC, DovidBenAvraham, I see you have done it. Although I think it would have been worth keeping the history section as it provides additional context for the articles, and names the manufacturer, which had some historical and could have done with an article. Scope creep (talk) 12:08, 27 September 2017 (UTC)

I've now implemented Phase 1 of what I proposed in the first two paragraphs of my 04:36, 27 September 2017 post. The "Selected features" section has been renamed the "Small-group features" section, and I've combined a couple of items in it to make room for two new items. Thus the number of items and the number of lines in that section has not been increased, although the byte-count has gone up slightly because I used each line more fully and added 4 more refs to provide complete referencing. Thank you, scope_creep, for insisting that I increase the number of links; using links to various sections of the Backup article has enabled me to use WP-standard backup terminology in the items and not have to provide explanations. Later substituted Proactive script feature for "Script Hooks" feature; Proactive scripts are widely-used and have been in Retrospect so long that Dantz's U.S. patent expired in 2016, whereas "Script Hooks" were only officially announced in March 2017—although they may have quietly been added to Retrospect Windows in 2015. DovidBenAvraham (talk) 11:22, 4 October 2017 (UTC)

Phase 2 will be, as I said above, to create the separate "Enterprise Backup" article so the the "Retrospect (software)" article can link to its items. The creation of the basic contents of the "Enterprise Backup" article is not so daunting, because it will be built from those items from the "Main features" of the old "Retrospect (software)" article that were left out of the new "Small-group features" section. The daunting part will be my searching of Web documentation for other enterprise backup apps, in order to provide the refs and alternate terminology that will turn the "Enterprise Backup" article into something that is not Retrospect-specific.

I too am unhappy, scope_creep, that there isn't an article on Dantz Development Corp.. But AFAIK the material to write such an article simply isn't available. However it occurred to me the other day that the "Retrospect (software)" article could be significantly improved by rewriting the "History" section with an extra paragraph. I realized that what evidently happened after EMC bought Dantz is that EMC management (EMC being an enterprise-oriented corporation) put pressure on the Retrospect developers to upgrade it from a small-group backup app to a true enterprise backup app. That's why new enterprise-oriented features appeared in Retrospect Windows 7.0, and why more such features appeared in Retrospect Windows 7.5 and 7.7. And then Microsoft put a severe kink in the effort, by adding security features to Windows Vista that made it impossible for the Retrospect developers to add the free-standing GUI Console that the developers felt was necessary. So, after layoffs and re-hirings, the Retrospect developers focused on Retrospect Mac 8 to show what they could really accomplish. Because of lack of time, since they were bucking EMC management, the developers released a version that had lots of bugs—especially in the Console—and was incapable of running on the older PowerPC hardware that a lot of Retrospect Mac administrators used for running "backup servers". This is not speculation; the third-party articles I used as refs for the old version of the "History" section support this interpretation. If it's not too scary for you folks, I'd like to do that "History" rewrite. Please comment. DovidBenAvraham (talk) 04:08, 4 October 2017 (UTC)

If the articles directly state what you've summarized about the history of these companies then a couple of new sentences wouldn't be bad. But if the sources merely support, or are consistent with, this interpretation as you say, then that's SYNTHESIS and not proper. You can recast & rephrase a reliable source's research and conclusions on a matter but you can't collect information from a variety of sources, construct a story that appears to knit it all together (even if the pieces seem to fit together in only that one way) and add it. Editors' own understanding, their own interpretation of facts, aren't permissible. I hope that's helpful. JohnInDC (talk) 11:43, 4 October 2017 (UTC)

I reviewed today's addition and little by little realized that nearly every assertion in it was an inference drawn from the source material, or an interpolation of facts. This is not a narrative that appears in any of the sources and is inappropriate SYNTHESIS, and so after a couple of cleanup efforts, I restored the prior version. JohnInDC (talk) 00:24, 5 October 2017 (UTC)

I may have picked the wrong sources for the section, but the assertions themselves are not inferences or interpolations. For instance, I have belatedly discovered a 2014 Retrospect Inc. Knowledge Base article that directly confirms the main assertion of my second paragraph—that starting with Windows Vista "when Retrospect auto launches [via the Retrospect Launcher service] Windows puts Retrospect into a protective desktop environment where the user will not be able view and control Retrospect". However, if you won't accept that as a reference, I can simply use JG Heithcock's December 2009 Retrospect Developers' blog post about the just-released Retrospect Windows 7.7; he mentions "others unhappy it wasn't going to be the full-blown new User Interface as the Mac product got" and then states that "Given the resources we have today, not putting a brand-new UI saved us a lot of time, both in development and testing." So in that case I can simply make a well-supported assertion that EMC developers chose to delay implementing the same GUI in Retrospect Windows as in Retrospect Mac; Eric Ullman's September 2009 blog post says the same thing. If you don't like my Release Notes reference showing that Retrospect Inc. was still trying in September 2017 to provide a workaround for the Windows Vista "protective desktop environment" problem, I'll leave that assertion out.

As for my third-paragraph assertion about the EMC developers' success in splitting the Retrospect Macintosh backup server into separate separate communicating GUI Console user-space and server root processes, I have several third-party references I can include for that. As for the terminology change in Retrospect Mac 8, the latest edition of the Joe Kissell book (the one I paid $15 plus tax to download so I could view it freely) directly says the terminology was changed in Retrospect Mac 8. And as for the assertions about EMC layoffs/rehirings and customer dissatisfaction with Retrospect Mac 8, I used to have those in the old version of the article—and can easily put them back in.

I don't want to get into a revert war with you, JohnInDC. Therefore please reply as to whether you would accept the "History" section additions as I have re-proposed them in the two prgfs. directly above. DovidBenAvraham (talk) 03:14, 5 October 2017 (UTC)

Retrospect developer blog posts are not reliable third party sources. They are first-party sources, the opinions of one or two individuals with (apparent) knowledge of particular circumstances at a particular point in time - they may be right, or wrong about why the company was doing what it did; others within the same company with equal knowledge (but no blog) might've disagreed. They might've had extrinsic reasons for posting the things that they did, or for the spin they put on them. We don't know. I don't think that, as a general matter, those posts can be used to construct an otherwise unpublished, insider's view of the history of the software and the firms that owned & developed it. Find a third party reliable source that tells this story from start to finish and summarize that. Otherwise - the lack of any such sources are a strong indication that this is not sufficiently well documented to include at all. JohnInDC (talk) 03:32, 5 October 2017 (UTC)

In this case the developers who blogged were Eric Ullman, Retrospect product manager at the time who previously worked for Dantz from 1992-2004, and JG Heithcock, director of software engineering for Retrospect at the time who had been working for Dantz/EMC since 1998. Heithcock has now been CEO of Retrospect Inc. for several years, so we know he must be a master of spin. And in fact I don't necessarily believe Heithcock's December 2009 explanation, but the "blog" was evidently EMC's semi-official way of communicating with Retrospect customers less formally than with press releases.

But I don't have to use the "blog" as a ref. at all. I just mentioned the "blog" above in case—for some reason—you have a problem with my proposed official first-party refs. from Retrospect Inc.'s website: a 2014 Knowledge Base article, and Release Notes accompanying the September 2017 version of Retrospect Windows. All I'm trying to newly establish in the second paragraph of the "History" section is that EMC/Roxio/Retrospect Inc. has still not implemented a split between the Retrospect Windows GUI and "backup server engine" processes, and as a result Retrospect Windows requires special administrator operating procedures as of 2017. All I'm trying to establish in the third paragraph of the "History" section is that EMC did in 2009 implement for Retrospect Mac a split between the GUI and "backup server engine" processes, a change to a more Mac-like GUI, and a change in terminology. That justifies the statement I made in the last sentence of the third paragraph: "Retrospect Inc. has continued to sell two flavors of backup server software that, while having nearly identical non-GUI code, are operated differently by the administrator and have different terminology." I can justify that with official first-party and third party refs. without using any "blog" posts, and I propose to do so unless I hear from you to the contrary. DovidBenAvraham (talk) 11:53, 5 October 2017 (UTC)

First, this level of detail is excruciating and I object on that basis alone. Second, you need to find a third party source that pulls all this together for us. Not your observations tying together different statements or observations by internal sources at Dantz or EMC or wherever. Finally it seems to me that if you want to mark a difference in functionality or code between the Macintosh and Windows versions, the place to do it is not in the History section but at one of the locations where you describe the different flavors of software: "Retrospect for Windows has nearly identical non-GUI code as Retrospect for Macintosh but they are operated differently by the administrator and have different terminology." (If that captures the distinction properly.) The fact of the difference seems reasonably well sourced. How and why it came to be, and why it persists, is not. JohnInDC (talk) 13:13, 5 October 2017 (UTC)

Almost certainly no such third-party source exists; I've searched. Since reviewing publications are usually targeted to particular OS platforms, why would any reviewer bother writing a review that covers what—on the surface—appear to be two differently-functioning "backup server" apps that run on two different OS platforms and simply share the same name? As to how and why the difference came to be, and why it persists, the Knowledge Base article will make that crystal-clear for the technically inclined: it's because of security features added to Windows Vista that aren't in macOS. And as for the "excruciating" level of detail, I think I can cut my additional 11 lines in the "History" section down to 6 lines. The result will be very terse, and will be festooned with refs like a Christmas tree, but that shouldn't bother you zealous Wikipedia editors. After I've written the 6 lines, you can decide where they should be moved to in the article. However I don't think they belong in the "Editions and Add-Ons" section; that distinguishes between various pricing levels of the software, not the macOS-vs.-Windows "flavors" (as I have termed them). DovidBenAvraham (talk) 21:30, 5 October 2017 (UTC)

I don't know how many times I need to say it, but if you are not repeating or summarizing what a third party source has said, then you are contributing original research or synthesis, which are not how the encyclopedia is written. Editors here edit other independent, reliable material down into Wikipedia articles; they don't connect the dots among disparate sources to devise interpretations of their own. I'm sorry you think my concerns are extreme, but this discussion has dragged on for weeks and I'm tired of making the same points over and over. Again: If a third party has written up a history of Retrospect, and that history examines the different architecture of the program on the two platforms and how they came to be, then include it; but otherwise, please don't. JohnInDC (talk) 21:59, 5 October 2017 (UTC)

Here's the Retrospect Inc. Knowledge Base article "Auto Launching Guide for Retrospect for Windows". If I simply reference that single source in a brief quote, where's the inference and connecting the dots that I'll be supposedly making? I guess I must be really stupid not to understand your concern in this case. DovidBenAvraham (talk) 01:17, 6 October 2017 (UTC)

Since I must be really stupid, here's a Really Stupid Procedure for Resolving This: (1) I clone the existing article into a "Retrospect (software) for Windows" article; there'll be less than 2 dozen words difference between the two articles. (2) A shill (possibly I myself) suggests that the two articles be merged. (3) Some kind of online committee forms to consider this question; I testify that I'd be happy to merge the two articles, but JohnInDC won't let me because of a crucial 6-line difference between the articles. (4) I may come out with egg on my face, but somebody else may come out with egg on their face instead. How about it, JohnInDC, are you ready to put your WP reputation where your mouth is?

"Due to mandatory Windows security settings starting with Windows Vista/Server 2008, Retrospect when auto-launched does not interact properly with the user. The program must instead be launched manually and be minimized, or another workaround employed." The cited source goes into more detail, but this is not a user manual, and the foregoing captures the gist of it. It still doesn't read like "history" to me, but it's only a line and a half and fairly straightforward and if you want to include that, or something like that, I won't object. JohnInDC (talk) 02:10, 6 October 2017 (UTC)

I did not intend for you to add back in all of your synthesis along with that sentence. The sentence was it. I'm going through the material that you've restored and checking it against the sources. If the sources don't say what you cite them for then I will remove it. JohnInDC (talk) 23:38, 6 October 2017 (UTC)

I pared the material down a bit, leaving in the main points (sold to EMC, Windows 7.5 good, EMC hoped to make more improvements to Windows but were frustrated by Vista, didn't focus on Macintosh, v.8 sucked) but eliminating obscure / proprietary technical terms as well as recommended "fixes", which are, in the end, entirely beside the point. Now it's reads like a story - supported by the sources - and much less a how-to. The only not-purely-sourced inference now is to say that the hopes for Windows were to make it more like the Mac version (no one really says that) but it's a small point and helps the narrative flow. JohnInDC (talk) 00:19, 7 October 2017 (UTC)

No, the hopes for Retrospect Windows were—and still are—to give it a capability that EMC management had already determined is needed for any client-server backup system designed for use in enterprises larger than 20 people; that capability is complete two-way data interchange between the backup server and another computer running a Console app. For you to understand that, I must get you to look at the way modern larger offices are administratively structured. The last office I worked in before I retired had 60 people; of these, only 3 IT people were allowed access to the locked (and fire-hardened) server room. Once multi-client backups started being done to large-capacity hard disk drives instead of to tape drives, bosses everywhere realized that they could delegate administration of daily backups to an administrative assistant (who used to be called a "senior secretary")—not a scarce and highly-skilled IT person. The actual backup server, with its attached HDDs, can be kept in the locked server room, but the administrative assistant (referred to as the "administrator" by Retrospect Inc.) needs to be able to change schedules for backup scripts and monitor scripts for correct operation. To do that conveniently, the backup administrator really needs to be able to do two-way interaction with the backup server app from another computer that is not in the inaccessible server room. Under macOS that turned out not to be a problem; the Console-Server split was successfully (although buggily) implemented in Retrospect Mac 8 at the beginning of 2009. However the security enhancements in Windows Vista basically made it impossible to implement the Console-Server split in the planned Retrospect Windows 8 at the end of 2009, because complete two-way communication between two user-space processes is no longer allowed. So, as the Heithcock December 2009 blog post said (without giving the real reason why this was done), EMC instead decided to bring out a Retrospect Windows 7.7 version (for which they charged extra despite its being nominally a "point release") without a Console-Server split. Not having a separate Console makes it impossible for an administrator to do full two-way interaction with the backup server, unless the administrator is allowed physical access and account access to the backup server machine—which in most offices the administrator is not allowed. The two sentences JohnInDC deleted in my second "History" paragraph were an attempt to mention the only two practical operating alternatives for Retrospect Windows, and to say—briefly and relatively-non-technically referenced to the Knowledge Base article—that both these alternatives are unsatisfactory. If JohnInDC lets me put that statement back in, I can give a ref to the Preferences section of the Retrospect Windows User's Guide that demonstrates that "launch[ing] the program manually, leav[ing] it open, and minimiz[ing] it ... [also] causes administrator problems". As I've said above, Retrospect Inc. tried to implement one-way communication between the backup server and a stand-alone view-only Retrospect Dashboard app in 2014, botched the implementation, and refused to officially document the stand-alone Retrospect Dashboard app through 2017—although they claim to have finally fixed the implementation in September 2017 (as shown in the Release Notes ref that JohnInDC left in). BTW, note that—in order to keep as close as possible to my promised 6 additional screen lines—I've left out any mention of the differences in GUI and terminology between current Retrospect Windows and current Retrospect Mac. DovidBenAvraham (talk) 03:08, 7 October 2017 (UTC)

I'm sorry, but all of that is far beyond the scope of this article, which is about the existence, basic function, and general history of a piece of retail software; on top of which, none of what you've just laid out, really and truly none, is in any of the sources at hand. What EMC or Dantz was trying to do, the needs they were seeking to meet, and your highly technical descriptions of appropriate backup architecture as manifested in Retrospect (or not) - that's all your own inference, based on your own understanding of the industry, and how mid-sized enterprises work and what their needs are; and the material about how users should work around these limitations are precisely the kind of how-to guide that Wikipedia is not. Please stop now. The article, and its history section, are now comprehensible, they line up with the sources, and they're appropriate to the subject at hand. JohnInDC (talk) 03:21, 7 October 2017 (UTC)

OK, but you should be aware that all the client-server backup apps I've looked at so far have Consoles. NetBackup has one. Tolis BRU looks like it has more than one type of Console. Just sayin', but you'll be able to read more about this cross-app feature when I write the "Enterprise Backup" article. DovidBenAvraham (talk) 03:52, 7 October 2017 (UTC)

"Favorite Folder"

This is an OS X-specific word for the more general and commonly understood term, "subvolume". Retrospect (for both Mac & Windows) backs up subvolumes. On OS X these subvolumes are called "Favorite Folders". The function is the same, just the word is different. You don't need to introduce the Mac-specific terminology, define it in terms of the common word, so that you can use it a single time later in the article. It adds nothing to the understanding of the functions of the software. Again - again, again - this is not a user manual, it's not a glossary or lookup table. "Retrospect can be configured to back up subvolumes." That's all that needs to be said. Please stop making this change. JohnInDC (talk) 15:09, 15 October 2017 (UTC)

You're wrong in this specific case, JohnInDC; Favorite Folder is not an OS X term—I've Googled it. Instead it is an EMC-coined term for a subvolume that is defined and distinguished only within Retrospect. It is therefore distinct from a Btrfs subvolume, which AFAICT is—once it has been defined—more generally visible within the filesystem. I'm pretty sure that's why EMC changed the terminology for Retrospect Mac 8; subvolume is the older Retrospect term that is still used within Retrospect Windows. That's why I used the term Favorite Folder, and took pains to put in a ref to the Joe Kissell book—and improved that ref to show the page number corresponding to the definition for Favorite Folder (the actual ref is for "subvolume", because the only copy of the Kissell book that is freely accessible on the Web predates Retrospect Mac 8). DovidBenAvraham (talk) 16:13, 15 October 2017 (UTC)

I didn't mention that I'd also searched for "Favorite Folder" in posts over many years to the Ars Technica Macintoshian Achaia forum. I found none. I did the same search in all Ars' Operating System and Software forums; it found 5 posts with the phrase, all in the Microsoft OS & Software Colloquium. DovidBenAvraham (talk) 21:25, 15 October 2017 (UTC)

If it's a Retrospect term then it's only worse. It's simple: Restrospect can back up a volume, or a subset of that volume (aka a subvolume). We don't need the special Retrospect name to convey that concept. It's not necessary and it clutters the article with proprietary terminology that adds nothing. Better without it. JohnInDC (talk) 22:14, 15 October 2017 (UTC)

OK, but if scope_creep mindlessly tries to link "subvolume" to Btrfs subvolume I'm going to edit the link out as soon as I see it—because it's misleading in the article. DovidBenAvraham (talk) 23:50, 15 October 2017 (UTC)

Right after the word "subvolume", I've added a link to the latest version of the Retrospect Mac User's Guide. If anyone tries to find "subvolume" in that, they'll be taken straight to a Glossary of Terms entry that says "In previous versions of Retrospect, a folder you designate as an independent volume for use within Retrospect. Retrospect uses the term Favorite Folder." DovidBenAvraham (talk) 09:48, 16 October 2017 (UTC)

Consensus

DovidBenAvraham, we have consensus, which came about from the Rfc, to simplify the article, to remove what was essentially a manual, and create a new article, which JohnInDC did, to keep it simplified, salient and linked to WP. We updated the article to put in salient sections in history, which provided additionall context, linked it per WP:MOS for those terms that needed it, and removed all the advertising per WP Terms of Use. You now seem to adding more more stuff, which is outside the consensus, which nobody wants, and is breaking the article again. It is almost 6 weeks now since the Rfc, and you are still not sticking with it. If you continue to work on the article, against concensus, and adding more more detail which is well beyond what is required, its going to an administrator to sort, likely to WP:ANI. Also please stop reordering these comments, it is bad form and i'm sure it is illegal on WP. scope_creep (talk) 08:23, 18 October 2017 (UTC)

First, I have never been shown any consensus (I, at least, know how to spell "consensus" consistently) resulting from the RfC, and I prefer to see it as pointed to by someone other than scope_creep. Second, my article edits from 00:25 through 01:17 on 18 on 18 October added no detail except the two words "overly complicated" (with ref for the quote); everything else was actually a clarification of what was already in the "History" section—namely attempts to clarify (without adding a single screen line) that Windows Vista's changes prevented EMC from creating a fully-interactive console for Retrospect Windows. Third—as I have already stated above, there never was any "advertising" in the article except by scope_creep's definition—which as I have noted above is not Wikipedia's definition. Fourth, I have never rearranged any comments on this page; I just put in a new section header between existing comments. Finally, I posted an Active Disagreements entry at 04:28 on 18 October 2017, and then clarified its last sentence (while changing the date) at 05:18. Let's wait until that concludes before proceeding to any WP:ANI. DovidBenAvraham (talk) 10:24, 18 October 2017 (UTC)

Cleaning up this article has been long and laborious and exhausting. DBA, you've been communicative all along and I appreciate that but progress throughout has been so slow, and so incremental, with so much back-and-forth and with oh-so-many-words that when you come back and make just one small change back in the direction we're trying to move the article away from - well, I can appreciate scope's losing patience. We'll see what the Third Opinion request produces as regards this latest episode but really the best thing would be for you to turn away from this article and see what you can contribute to one of the other 5.5 million articles that are here on English Wikipedia. JohnInDC (talk) 10:45, 18 October 2017 (UTC)

That good. Some more progress. Let's get the Third opinion sorted out. I not sure what it is, but it has got to be better than this. I would like to draw your attention to JohnInDC suggestion about working on other articles. There is over 200k articles on WP, which don't have sources. They desperately need some help. scope_creep (talk) 11:24, 18 October 2017 (UTC)

First of all, JohnInDC, I'd appreciate it if you would show me the written consensus reached by the RfC. It's customary in the English-speaking world to give the defendant a copy of the judgement against him/her. The WP:ANI surely can't blame me for violating a judgement I never saw.

Second, as I said four paragraphs above this, my only "one small change back in the direction we're trying to move the article away from" was adding the two words "overly complicated" in my attempted first edit. My attempted second edit was entirely to clarify what was already in the article; in fact it was enough of a clarification that IMHO I now think the link to application isolation would enable me to eliminate the ref to the Retrospect Knowledge Base article—which is mostly an overly-technical description of how to get around the lack of a fully-interactive console in Retrospect Windows. I believe scope_creep wasn't thinking properly when he/she reverted that second edit.

Third, the only other Wikipedia article I want to contribute to at the moment is the "Enterprise Backup features" article I proposed above. But there's no point in my starting to draft it if scope_creep later decides that it is a pure extension of the "Retrospect" article, and that therefore he has the right to apply his non-WP standards on "advertising" and "marketing" to it. That's why I requested the Third Opinion, not that I enjoy the fight.

Fourth, I have a suggestion for something scope_creep can do in the meantime. By my count there are 31 occurrences of the phrase "Time Machine" in the article Time_Machine_(macOS). That is surely a "marketing" term ("Time Machine" certainly has a marketing ring to it, and everyone is a marketer over at Apple) in a backup app article by scope_creep's standards just as much as "Retrospect" is. Why doesn't he/she edit out most of those phrases, changing them to "the program" or something similar? I think the resultant reaction among Apple fans on Wikipedia would justify the phrase "came down on him like a ton of bricks".

Finally, I owe a slight apology to scope_creep. I forgot that I did move his/her 15:10 16 October 2017 post to the preceding section. But that's where he/she should have put in the first place; it has nothing to do with "Favorite Folder". We are all of us capable of using View History to see what changes have been made to a page, even if a new post was added to a section other than the last one. DovidBenAvraham (talk) 12:17, 18 October 2017 (UTC)

I realized the other day that the Comments in the Survey sub-section of this Talk page were supposed to be taken together as the "written result" of the RfC. However my Internet was down for 3 days as a result of Verizon stupidity, so my apology had to wait until it was back up. I'm sorry, there was no other "written result" that JohnInDC should have shown me. However, in those Comments nobody said I should not be allowed to edit the article without prior consensus, which is what scope_creep has been recently claiming in reverting my edits. DovidBenAvraham (talk) 14:43, 26 October 2017 (UTC)

Please, may I have consensus to add the words "GUI-scripted" (with a link—which I think was previously deleted from the article—to the GUI article) in front of "backup" in the last single-sentence paragraph of the lead? The article as it is has no hint that operations must be scripted (Retrospect Windows kept Immediate operations, but Retrospect Mac 8 replaced those with equivalent GUI buttons that create scripts and immediately run them) until the last three items in the "Small-group features". The word "GUI-" in front of "scripted" is my sneaky single-word way of inserting the fact that Retrospect (both variants) has a GUI for creating scripts. DovidBenAvraham (talk) 20:35, 19 October 2017 (UTC)

Since I hadn't heard any approval or objection in two days, I have made this edit—which included a link for GUI. I then did two more edits re-arranging the single sentence, for clarity. I've also made "Consensus" a section on the same level as "Favorite Folder" in this Talk page, since the discussion in this section has nothing to do with Favorite Folder. DovidBenAvraham (talk) 19:06, 21 October 2017 (UTC)

I assume that the consensus allows me to clarify or add mentions of features, so long as this doesn't add to the number of screen lines in the article and doesn't go into any more detail about a feature than what's already there. In conformity with that assumption, I have today squeezed in brief mentions of three additional "Small-group features": non-Desktop Editions can simultaneously backup to multiple destinations using different scripts, backups can do file exclusion, and there can be separate MD5 validation scripts. In addition I have today clarified that the Windows Add-On for Dissimilar Hardware Restore automatically adjusts drivers, and clarified for what kinds of tape library the Advanced Tape Support Add-On is required and for what kind it isn't. These edits have completed putting into the article all "Small-group features" that don't belong in the forthcoming "Enterprise Backup features" article; all of them happen to be features that—with the exception of Cloud Backup—were all present in Retrospect Windows before the release of Retrospect Mac 8. DovidBenAvraham (talk) 19:56, 28 October 2017 (UTC)

Please make sure that by adding detail you are not rendering the prose more technical and harder to follow. I'm not convinced, for example, that adding the sentence, "Using non-Desktop Editions of the backup server, multiple scripts can simultaneously backup to different destinations" really tells the reader anything that they need to know. Here as in so many other instances, detail is not good simply for its own sake. At some point it just becomes clutter. JohnInDC (talk) 21:46, 28 October 2017 (UTC)

Similarly, "clarifying that Windows Add-On for Dissimilar Hardware Restore extends the Emergency Recovery CD" is a matter for a user manual, not a high-level overview of a particular piece of backup software. This is not the place for that kind of stray - indeed trivial - observation. JohnInDC (talk) 21:50, 28 October 2017 (UTC)

I've since clarified the language of the last sentence of the Backup Destinations "Small-group features" item to say "If run on non-Desktop Editions of the backup server, multiple scripts can simultaneously backup to different destinations." That says in non-techncal language that those Editions are multi-threaded, which is an announced 2016 feature of the Arq personal backup app. (Actually multi-threading works in the Desktop Edition too, but you have to reset a Preference each time you start the Retrospect backup server to use it.)

The article used to have one item saying "Retrospect Emergency Recovery CD—a single generic boot disc that uses WinPE to provide bare metal recovery of most Windows computers ...." It used to have another item saying "'Dissimilar Hardware Restore Add-On'—giving Retrospect Windows the capability of restoring an entire machine to a completely different computer—including after-the-fact automatic adjustment of drivers to account for differences in the hardware." Imagine you are Sam the manager of a small office, and the hard drive on Suzy's ancient Windows computer suddenly dies. So you pull a new computer out of the office closet and put it on Suzy's desk, but it isn't even the same brand as Suzy's old Windows computer. Wouldn't you like to use a combination of these two features to re-establish a bootable hard drive with all of Suzy's files on it? DovidBenAvraham (talk) 19:13, 29 October 2017 (UTC)

Sure. And Retrospect can tell us all about it in their ads. JohnInDC (talk) 20:17, 29 October 2017 (UTC)

Is there any reason I shouldn't add the Windows Emergency Recovery CD as an item in "Small-group features"? It's certainly not an "Enterprise Backup feature", even though it's available only for Retrospect Windows, and as such it has just as much right to be listed as any other "Small-group feature". And is there any reason not to put the Windows Add-On for Dissimilar Hardware Restore back in "Editions and Add-Ons", where it used to be until you removed it, JohnInDC? I added the feature to the Add-On in "Editions and Add-Ons" because they are related, and it would save an item screen line in "Small-group features". The terse language I used was so as not to add an extra screen line to "Editions and Add-Ons". DovidBenAvraham (talk) 03:59, 30 October 2017 (UTC)

I'll make my last comment, as it has been 6 weeks to the day since the Rfc went in and I'm not committing any more time to this article. Lastly, no one uses the term GUI. The last time I heard GUI being used was about 2007. It is now UI and if it is mobile it is UX. Also console. The last time that word being used in action was the early 1990's. Anybody under 40, will think it is games console, your talking about. Truly. Your manual hasn't been updated for many years, perhaps 20+ years more It is woefully out of date. Anyway do what you want. Some other editor can deal with it. scope_creep (talk) 23:45, 20 October 2017 (UTC)

Ah, scope_creep, how kind of you to leave us with another of your justly-famous "nobody ..." statements (for which other editors can search the Talk page). You may be right; only one of the two doormen I have surveyed knew what GUI meant (a third doorman may—since he has a side business doing video recording and editing, but he's in the hospital with 5% kidney function). However GUI goes straight to the Wikipedia "Graphical user interface" article, and UI goes straight to a disambiguation page from which in two steps you can go to user interface types—which lists GUI as the second bulleted item. UX, on the other hand, goes to a disambiguation page which says "UX refers to user experience[my emphasis], a person's behaviors, attitudes, and emotions about using a product, system, or service" at the very top.

As for "console", all the enterprise backup applications whose reviews/documentation I have surveyed have what some of them more fully term an "administrative console". EMC introduced its Retrospect Console with Retrospect Mac 8, which was released in early 2009. I guess lead developers of enterprise backup applications in the late 2000s—many of whom must have been in their 40s by then—spent too much time imitating each other, and not enough time playing video games with their kids.

I'd love to know the sociological profile of the people you associate with in your job as a presumed system programmer. However, thanks for your help on the article. DovidBenAvraham (talk) 03:25, 21 October 2017 (UTC)

Preliminary discussion of "Enterprise Backup features" article

As I've explained above, creation of the new article would be a substantial amount of effort. The article would be built around a discussion of the 18 features that we eliminated from this article, but presented in a fashion that would not favor Retrospect or any other enterprise backup software application. It would be based on my realization that, starting around 2004, a number of makers of what were conceptually personal backup applications that had been expanded to handle small groups decided to expand them to be able to handle larger groups in enterprises. The expansion process required the developers to talk to potential users, and also impelled them to keep close tabs on what their competitors were developing. Thus a constellation of enterprise backup applications emerged in the mid-to-late 2000s that had essentially similar new features, even though in some cases the terminology for those features varied from seller to seller. If I'm to write an article discussing those new features that will be acceptable on Wikipedia, the article will have to mix refs to user manuals with refs to what reviews exist for these enterprise backup applications. Before starting to write the article, I need to get confirmation that—if written according to generally-accepted WP standards—the article would be acceptable to such editors as JohnInDC. That's why I've started a discussion on this Talk page. DovidBenAvraham (talk) 01:10, 22 October 2017 (UTC)

This really isn't the place to discuss this. Not too many editors follow this Talk page. I have no particular expertise in the area and haven't got much inclination to gain it; plus TBH I'm tired of trying to convey the same basic points again and again. You should go visit WP:HELP and follow the links about your first article, and where to ask for assistance. WP:Teahouse perhaps. You need a broader audience, who doesn't bring any preconceptions to the process, I think. JohnInDC (talk) 01:39, 22 October 2017 (UTC)

I do now understand "the same basic points again and again" that JohnInDC mentioned directly above. In particular, one thing I did yesterday is to look for reviews of other competitive enterprise backup applications. Of course the first problem was to identify them, since I haven't used Windows since I retired in 2004 and there doesn't seem to be another competitive enterprise backup application for Mac (Archiware seems oriented toward tape rather than disk, and Tolis BRU seems oriented toward media production enterprises). I decided that Retrospect Inc. must know who its main competitors are, and therefore looked at its "Competitive Analysis" Knowledge Base articles to identify those competitors. To my pleased surprise there are some real Web reviews (not just repeats of press releases) of the competitor products. Of course the information in those reviews will have to be supplemented with information from product user manuals (some of these manuals are behind "signup walls", so I guess I'll be deleting myself from e-mailing lists for a while), but that's the kind of thing I've already done in the Retrospect article. So I now think I'll just go ahead and create a sandbox version of the new article, and let other editors take a look at it at some point. I hope I don't have to put in too much work before getting a verdict. DovidBenAvraham (talk) 08:47, 22 October 2017 (UTC)

Just to be clear, I expect the new article to be revised and added to by other editors more familiar with primarily-Windows enterprise backup applications—indeed I welcome that. I just don't want the new article to be revised in the way that scope_creep tried to, with spurious interpretations of Wikipedia rules. All I want to be left with is an article I can link to from a brief new section in this article, saying something like "Retrospect has most of the Enterprise Backup features, excepting only ...." DovidBenAvraham (talk) 10:33, 22 October 2017 (UTC)

Just by way of reminder, again - whatever article you wind up writing can't be your own compendium of features you deem important to the class. Find articles that say, "these are the important features of Enterprise Backup Software" and summarize the lists they provide. And bear in mind, again, that Wikipedia articles aren't buyers' guides - it isn't the place to list a bunch of features alongside a list of products with checkboxes "yes" or "no". That's for Consumer Reports or Computerworld or whoever is helping consumers decide what software is best for them. And finally I'll state my original suggestion a little more strongly. The Talk page for Retrospect is not the right place to develop a related, but entirely separate, article. Please find a more suitable place for the discussion than here. You'll likely get better input at that spot too - you can always direct them back to these discussions for a primer. JohnInDC (talk) 12:05, 22 October 2017 (UTC)

Good point about articles that say "these are the important features of Enterprise Backup Software", JohnInDC. There's a 2016 Gartner report that sounds from the title as if it might be promising; there was a problem with the Gartner site as of this morning, but I've just looked at the 2017 version courtesy of a vendor that's mentioned in it. As for a buyers' guide, there's already one for backup software on Wikipedia (Retrospect is there; many of the competitors aren't) but I've no wish to enhance it. I took a brief look at WP:Teahouse last night, but I'm not sure I'd find the kind of editors I need frequenting it. Thanks to your past guidance on this article, for which I'm extremely grateful, I actually think I know what I've got to do to make the new article acceptable. Speaking of "articles that say, 'these are the important features of ... Backup Software'", the WP article Backup has proven extremely useful for links in writing this article,but (as I've said above) it seems to have been written based more on editors' knowledge than ref'd sources and probably wouldn't pass muster with you. DovidBenAvraham (talk) 16:22, 22 October 2017 (UTC)

Following a Teahouse discussion, I've followed a suggestion from Nick Moyes and replaced the former "In society" section at the end of the Backup article with a new "Enterprise client-server backup" section. As you can see from my initial comments in this section of that article's associated Talk page, I've been faithful to the lessons I've learned from JohnInDC and scope_creep. I've now added a one-line section to this article, linking to the new section in that article. Thanks, folks. DovidBenAvraham (talk) 01:14, 16 November 2017 (UTC)

Editions and Add-ons

I never tried to shorten that section but now that I'm paying attention to it, I realize that it's cumbersome and confusing and after reading it three times I still don't have an idea what an "Edition" is. It should be about 4 sentences long: "Retrospect also sells Editions and Add-ons, which are thus-and-so." I'm going to see about making this better. JohnInDC (talk) 19:39, 29 October 2017 (UTC)

Here's a possible new first paragraph: "Retrospect also sells Editions and Add-Ons. Editions are Retrospect Inc.'s 'soak the rich' strategy for making installations backing up any computers running Windows Server or macOS Server pay a great deal more for the same product. Add-Ons are Retrospect Inc.'s 'soak the rich' strategy for making either Windows installations that need to backup NTFS files for continuously running systems such as Quickbooks, or Windows installations that need to backup various server applications, or Windows installations that need to restore boot volumes to dissimilar hardware, or any installations that need to backup to a tape drive other than a single non-autoloader/non-library one pay a great deal more for the same product." Using this would enable you to eliminate the existing second paragraph in the section.

Here's a possible new second paragraph, replacing the existing first paragraph: "To avoid the need to distribute many versions of the executable for the backup server, activation of Editions and additional Add-Ons is governed by license codes. There is only one server executable distributed for the Macintosh variant and one distributed for the Windows variant. In addition, one client executable is distributed for each applicable combination of machine architecture and operating system."

Let me note in passing that there is an error in your existing second paragraph in the section. To backup to a single-slot tape library, usually known as an autoloader, requires either at least the Single Server Edition or the Advanced Tape Support Add-On. Let me also note in passing that, by deleting the former next-to-last paragraph in the section, you have omitted all the Windows Add-Ons for backing up three kinds of servers, and the Windows Add-On for Dissimilar Hardware Restore. DovidBenAvraham (talk) 04:56, 30 October 2017 (UTC)

If my reworking the section has introduced an error then the solution is fewer words and a higher level of generality, not more words and more specificity. I'm beginning to wonder why we need to describe "Editions" at all, if in the end it's just Retrospect pricing strategy. JohnInDC (talk) 10:43, 30 October 2017 (UTC)

Actually, given the elimination of any specific discussion of the Retrospect Mac Console and the elimination of the "Documentation" section, we could entirely do away with even mentioning Editions. There would be only minor costs in accuracy: in the last sentence of the Backup destinations item in the "Small-group features" section, and in any proper discussion of the Advanced Tape Support Add-On. There would, however, be a substantial cost in human lives: the deaths from "sticker shock" of readers going from this section in the article to the online Product Configurator its last paragraph references.

What I really suggest is: An abbreviated three-sentence version of the first paragraph I suggested in my "04:56, 30 October 2017 (UTC)" comment. It would have a NPOV-friendly substitute for "soak the rich", and would not enumerate the Add-Ons. This would be followed by the three short Add-On description paragraphs that were in the section prior to your 29 October edits. Those would be followed by what was the third paragraph prior to your 29 October edits, which starts out "Each Edition marketed ...". Following that would be the second paragraph I suggested in my "04:56, 30 October 2017 (UTC)" comment. The section would end with the final paragraph as you shortened it in your 29 October edits. DovidBenAvraham (talk) 12:51, 30 October 2017 (UTC)

Knowing what you would say, JohnInDC, I even found a third-party review that mentions Retrospect Editions and who the various Editions are designed for. Obviously the review doesn't discuss Retrospect Inc.'s motivation. DovidBenAvraham (talk) 18:31, 30 October 2017 (UTC)

Instructions and / or guides to what editions, etc. are appropriate to a particular user are the province of PC magazines, not Wikipedia. "A single Retrospect Edition supports a specified number of workstations. Retrospect also sells "Add-ons", which provide additional functionality such as X or Y". That's all. JohnInDC (talk) 11:45, 31 October 2017 (UTC)

I think the reason you don't understand the concept of Retrospect Editions, JohnInDC, is that you don't know what a "server OS" is or was. "macOS Server, formerly Mac OS X Server and OS X Server, is a separately sold operating system add-on [my emphasis] which provides additional server programs along with management and administration tools for macOS....A separate 'server' operating system is no longer sold [my emphasis]; the server-specific server applications and work group management and administration software tools from Mac OS X Server are now offered as macOS Server, an add-on package for macOS sold through the Mac App Store .... These tools simplify access to key network services, including a mail transfer agent, AFP and SMB servers, an LDAP server, a domain name server, and others." By contrast "Windows Server is [still] a brand name for a group of server operating systems [my emphasis] released by Microsoft."

Apple now sells the macOS Server add-on for a whopping $20; I think this is primarily because many of its most-widely-needed capabilities can be provided instead by a NAS—which many Mac installations have bought because it's less trouble. I can't say what Microsoft charges for a Windows Server OS—which varies depending on the "member of the family" you buy, but the price is probably falling for the same reason.

Meanwhile, if you use Retrospect Inc.'s Product Configurator, you'll find that the price of Retrospect Mac jumps from $119 for the Desktop Edition to $659 for the Single Server Edition—whether or not that single macOS Server add-on is on your "backup server" or a client machine. For Retrospect Windows the same Edition jump is only to $559, even though Windows Server is a distinct OS rather than an add-on. And it gets much more expensive if you're running more than one "server OS" machine on your LAN.

It's pretty obvious that, whatever the "server OS" development effort was many years ago, Retrospect Inc. is continuing a long-standing policy of having one low price for presumed personal/tiny-enterprise customers while having a much-higher range of prices for presumed better-heeled SME customers. That's what I was referring to as "soak the rich" in the partially-tongue-in-cheek first paragraph of my "04:56, 30 October 2017 (UTC)" comment. The policy is in stark contrast to that for the Arq backup product, which is $50 whether you're running it on multiple LAN workstations or on a LAN from a single "server OS" machine. I strongly believe the article should have just enough explanation of Editions to prepare potential Retrospect customers for the shock. DovidBenAvraham (talk) 15:39, 31 October 2017 (UTC)

Here's what I now seriously propose as the new first paragraph: "The backup server Edition is dictated—and priced—by the number of macOS Server or Windows Server computers being backed up in the installation. If there are no such "server OS" computers being backed up, the installation can use the much-cheaper Desktop Edition. Add-Ons are backup server features that are mostly used by larger installations; they are separately priced." Each of the two uses of the word "priced" in the paragraph would be directly followed by a ref to the Product Configurator on Retrospect Inc.'s website; thus the current final paragraph in the section, beginning "The combinations of Editions and Add-Ons marketed ...", could be eliminated. The other paragraphs in the section would be as I proposed in the second paragraph of my "12:51, 30 October 2017 (UTC)" comment, but with the final paragraph eliminated as proposed in the preceding sentence. DovidBenAvraham (talk) 07:20, 2 November 2017 (UTC)

"The backup server Edition is dictated and priced by the number of server computers being backed up. The less-expensive Desktop Edition can be used where desktop units, and not servers, are not being backed up. "Add-Ons" are separately priced additional server backup features." JohnInDC (talk) 11:25, 2 November 2017 (UTC)

I'm glad you like my proposal, JohnInDC, but your version of the first paragraph has a problem because it's not precise enough. As the last sentence of the first paragraph in the article lead says, "The company's backup server application runs on either a macOS or a Windows computer, but there are also versions of the client application that run on Linux [my emphasis] or classic Mac OS." If you go to this page you can download Retrospect Client for Linux x86 or Retrospect Client for Linux x64. Now there may be a few enthusiasts running a Linux computer strictly as a desktop unit, but almost everyone who runs a Linux computer is using it as some kind of server. (There never was any "server OS" version of Classic Mac OS.) Nevertheless Retrospect Inc. and its predecessors have never made any attempt to identify such Linux servers so as to charge more for them. Remember, as I have said above, the Edition pricing differential has always been designed to "soak the rich"—"the rich" being arbitrarily defined as any installation that is backing up one or more computers that run macOS Server or Windows Server. That's why my version of the first sentence in first paragraph specifically names those "serverOSes", and the second sentence says you can use Desktop Edition at any installation that doesn't run them. Also, your proposed third sentence has a double "not" in it, which I assume you didn't intend. DovidBenAvraham (talk) 12:35, 2 November 2017 (UTC)

Again - not a buying guide. This bit of information barely warrants inclusion in the article at all; so solutions should lean toward less detail, not more. Accordingly: "The backup server Edition is dictated and priced by the number of servers being backed up. The less-expensive Desktop Edition can be used where desktop units (or Linux servers) are being backed up. 'Add-Ons' are separately priced additional server backup features."

Sorry, the resultant first paragraph is going to get a bit klunkier. For one thing, does "number of servers being backed up" mean Retrospect backup servers or "server OS" machines? I used to put "backup server" in quotes, but scope_creep insisted on taking the quotes out. For another thing, "desktop servers" is going to have to be expanded to include "mobile computers", because backing those up has been a feature of Retrospect since 1996 when Dantz applied for a patent on Proactive scripts. I think you'll end up preferring my proposed version, even though it provides desperately-needed brand-name advertising for Tim Cook and Satya Nadella instead of Linus Torvalds. BTW, wasn't it scope_creep who came up with the idea that merely mentioning a brand name in a WP article is Advertising, marketing or public relations ? As you can see, it's not—especially in a paragraph that comes as close to saying "soak the rich" as is compatible with NPOV.

Nevertheless I'll now proceed to rewrite the section as we have agreed. The easiest way for me to do that is to start by reverting JohnInDC's 19:45, 29 October 2017‎ edit, and then proceed from there. So don't panic. DovidBenAvraham (talk) 16:13, 2 November 2017 (UTC)

I don't know why we need to say anything more than the Desktop edition backs up desktops and Linux servers, and the Edition versions are for OS X and Windows servers and are priced based on number of machines being backed up. Again this level of detail is stupefying. JohnInDC (talk) 17:57, 2 November 2017 (UTC)

I don't understand. Are you going to leave it in the cumbersome, multi-paragraph format, or reduce it back to the two or three sentences I'd whittled it down to? JohnInDC (talk) 19:36, 2 November 2017 (UTC)

I've reduced it, again, to the essential definitions of "Edition" and "Add-on". We don't need more. JohnInDC (talk) 19:40, 2 November 2017 (UTC)

(This is a comment I was about to save, whose timing overlapped with JohnInDC's 19:36 and 19:40 comments.) No, JohnInDC, the Desktop edition backs only desktops and mobile computers and Linux servers; I have now put into the article an "only" that you left out and I missed. IMHO the level of detail in the first paragraph was much less stupefying with the version I proposed in the first paragraph of my "07:20, 2 November 2017 (UTC)" comment; try reading it again. But you didn't like that version because its first sentence explicitly named the two "server OSes" that require more than the Desktop Edition to be backed up by Retrospect. You preferred a version of the paragraph whose second sentence specifies (mostly by implication) the OSes that don't require more than the Desktop Edition, purely to satisfy some cockamamie interpretation of the WP Advertising rule apparently thought up by scope_creep. If you want I'll substitute my version of the paragraph; with "that are mostly used by larger installations" removed from the third sentence, it actually takes the same three screen lines that your version—with necessary additions I made for accuracy—takes. DovidBenAvraham (talk) 19:56, 2 November 2017 (UTC)

No, I don't like it because these are simple concepts that can be conveyed simply, without breaking it down to OSes, pricing, the number of servers, "license codes", the various flavors of add-ons or any of that. These are very simple concepts. Retrospect has a consumer (and Linux) version, and a more expensive one for enterprise that is priced based (generally) on the number of machines or servers being backed up. That, and "add-ons" add function. This section on pricing and Retrospect's price discrimination model doesn't need to be here at all; and if it is, then it should be limited to describing, in general and quickly grasped language, what these terms mean. If people want to know more they can go to the Retrospect website. JohnInDC (talk) 20:12, 2 November 2017 (UTC)

You could even say (if the sources support it) that Retrospect markets both consumer and enterprise editions, with the latter's pricing related to the number of machines being backed up. I don't much care. What I do care about is an article that obscures simple concepts with Retrospect's own weird terminology and arcane pricing - it's not helpful, it's not clear, and it's not necessary. JohnInDC (talk) 20:19, 2 November 2017 (UTC)

To answer your 19:36, 2 November 2017 question, JohnInDC, I thought I did have implicit permission to implement my multi-paragraph 07:20, 2 November 2017 (UTC) proposal—since all you were questioning was the wording of the first paragraph. I'm sorry I made that assumption; I should have explicitly asked for permission.

A key question I should have asked before, JohnInDC, is what do you think is the knowledge level of a reader who makes it all the way down to the "Editions and Add-ons" section? The first barrier that reader would have to have enough knowledge to get past is the last sentence in the lead (as rewritten by scope_creep to insert a System Programming term): "The product is used for GUI-scripted backup in a heterogeneous network, primarily by small and medium-sized businesses." Anybody who can get past that can get past the "History" section, but would then have to get past the "Small-group features" section. Even with the links that scope_creep insisted that I insert, the density and terseness of that section would deter anyone who doesn't have a fairly-good overall knowledge of small-group computer technology and a fairly-good grasp of computer backup procedures. So we shouldn't assume that the "Editions and Add-ons" section has to be written so that every Wikipedia reader could understand it, any more than we have to assume (as scope_creep did) that the reader would have enough mathematical knowledge to understand (as I no longer completely do)—or a need to understand—the Checksum article he linked to in a "Small-group features" item.

To be frank, JohnInDC, I don't think you have that knowledge level—as is indicated by your past editing errors on this article. I tried to give you some of that knowledge in my "15:39, 31 October 2017 (UTC)" comment, but you then demonstrated that it wasn't sufficient in your "11:25, 2 November 2017 (UTC)" proposal for the first paragraph of the "Editions and Add-ons" article section—which lacked enough precision to show you now had enough knowledge to understand and rephrase my 07:20, 2 November 2017 proposal. Thank you for accepting my enhancements to your rephrasing. However I also think you lack the knowledge level to understand the basic concepts in some of the Add-Ons, particularly the 4 you omitted from your latest edit. For the first 3 that's not totally surprising; as an application programmer working in a Windows installation from 1999 through 2004, I used Microsoft Exchange constantly and at one point considered using Microsoft SQL Server—but I have only a slight theoretical knowledge of VMware because it became popular about the time I retired. However if you had used Windows computers for more than one hardware generation I'm sure you would have some idea what Windows boot volume drivers, which are adjusted by the Add-On for Dissimilar Hardware Restore, are.

I think any reader who makes it all the way down to the "Editions and Add-ons" section would have enough knowledge to understand my 6-paragraph 15-screen-line version. In the first two of the three paragraphs describing Add-Ons, I added one or two sentences of explanation in which there is no "weird Retrospect terminology" ("autoloader" is a standard IT term I linked to in a WP article)—but I could delete those if you insist. I thought a reader might be puzzled at the end of reading the section by whether Retrospect Inc. has to distribute an exponential number of backup server executables, so I put in a three-screen-line paragraph at the end saying they eliminate the need for that with license codes—but that too could be deleted. If—on the other hand—I crammed a mention of the four unmentioned Add-Ons into the last sentence of your first paragraph, my 8th-grade English teacher would rise from the grave (if he's not still alive at 105 years old) to smite me for creating the mother of all run-on sentences.

So please let me put back in my extra 9 screen lines, JohnInDC. The section will be the better for it. As for a supporting ref, the article I externally linked to in my "18:31, 30 October 2017 (UTC)" comment provides the support you asked for. DovidBenAvraham (talk) 01:54, 3 November 2017 (UTC)

TBH many of your comments I don't read with an eye to detail. When an edit requires paragraphs of explanation to justify it, it is - in my estimation - likely not a sound edit. Tell me. What precise information is to be conveyed in the "Editions and Add-ons" section? What information are you trying to impart to the reader there? Not to the "systems admin who wants to know how Retrospect is priced", but, someone who doesn't know much about the software and wants to learn more? Because, again and again, this is not a user manual, it's not a buyer's guide; it's a high-level article about the software, its functions, use and purpose. The question is not whether more can be said about "Editions & Add-ons" - surely there is - but whether any of this "more" is necessary, or improves the article. This section (if we bother with it at all), IMHO, needs to answer two quick questions: 1) What does Retrospect mean by "Edition"?; and 2) what does Retrospect mean by "Add-on"? Those answers take two sentences, and any reader who after seeing that wonders whether Retrospect is a suitable "software solution" for their small business needs can just go and click on the company web page linked in the article and learn all about it. JohnInDC (talk) 02:34, 3 November 2017 (UTC)

OK, I've solved this problem in a way that should satisfy both of us. I've added mention of the remaining three (see next paragraph to learn which one has disappeared) Add-Ons to the last sentence of the first paragraph (while cleaning up the punctuation and grammar), which only adds one screen line. I've also added a sentence to the second paragraph saying that Editions and Add-Ons are also activated by license codes; this sentence is so cut-down that it doesn't even add an extra screen line.

The Add-On that has disappeared in the 2017 version of Retrospect Inc.'s Product Configurator is the one for VMware servers. Evidently there are now so many home and tiny-enterprise customers using VMware servers, especially under Windows, that Retrospect Inc. decided requiring an Add-On for them boosted the price in a way that reduced total sales revenue. That they didn't announce this is in line with their corporate sensitivity on such subjects. I see Retrospect Inc. as having become financially "hooked" many years ago on the "soak the rich" Editions and Add-Ons pricing policy, and painfully having to "kick the habit" now. DovidBenAvraham (talk) 12:12, 3 November 2017 (UTC)

Better, but we still don't need to (all?) 4 Add-ons to convey the concept; and readers can go to the website to find out whether Retrospect with or without the extra cost add-on will meet their particular needs. I'm inclined to remove the list and reduce it to an example - if any. Which do you like best? JohnInDC (talk) 17:46, 3 November 2017 (UTC)

I cannot believe, JohnInDC, that we are having this dispute over mentions of important Add-On features that, including the features you left in, total only 2.5 screen lines of space as part of a single sentence. It surely cannot be because the Add-Ons are extra-cost items; the NetBackup article explains—not just mentions—an Auto Image Replication (AIR) feature that—if you read page 7 of the Veritas document linked to in its reference—turns out to be an extra-cost add-on.

Indeed Retrospect's Emergency Recovery CD mentioned in that sentence is a non-Add-On feature. It uses the WinPE released by Microsoft a few years before for Windows XP/2003; according to the Ullman reference "the WinPE method will allow for a single, generic boot disc to provide bare metal recovery for any computer supported by WinPE." I discussed it in the last paragraph of my "19:13, 29 October 2017 (UTC)" comment as surely being worthy of inclusion in "Small-group features", from which you deleted it at "01:31, 27 September 2017 (UTC)"‎ "to remove excessive detail, operating tips, etc.". If you reread that comment, you will surely realize that the feature is just as useful if Sam and Suzy are married with computers at their home, and if Sam runs down to his local Best Buy instead of pulling a new computer for Suzy out of the office closet.

Therefore please state for the record, JohnInDC, what WP rule allows you to delete the Add-On items that take up only 1.5 screen lines additional space beyond what you yourself put in. DovidBenAvraham (talk) 11:59, 4 November 2017 (UTC)

It's not a matter of a rule, or screen space, or word count, but just simple English prose. Copy editing. You need a single short sentence to say what Add-ons are, and a single example helps convey the concept. Punto. Identifying and describing them all is cluttery and list-y and makes even the very short paragraph a slog - and unnecessary. That's why I asked you above - what is so important beyond the simple notion of Add-on that you are trying to convey with the exhaustive list - bearing in mind again and again and again, that the article is not a Feature List or a User Guide or a Manual or Usage Tips or a Marketing Brochure. Less can be more, and this is one instance. JohnInDC (talk) 12:39, 4 November 2017 (UTC)

Now you're complaining about a run-on sentence, but that's exactly what I predicted we'd get in the fourth paragraph of my "01:54, 3 November 2017 (UTC)" comment. In 3 minutes (I've already tried it to the point of doing a Show Preview) I can convert that sentence, which takes 3.2 screen lines, into a 5-line paragraph of one "heading" line followed by 4 one-line bulleted items. I'll even delete the mention of QuickBooks, although I put it in because it's an extremely-common case of a continuously-running (hence open NTFS files) app on a Windows system—so I considered mentioning it a helpful "heads up" to a likely reader of the "Editions and Add-Ons" section (as I described him/her in the second paragraph of my "01:54, 3 November 2017 (UTC)" comment).

As far as "what is so important beyond the simple notion of Add-on that you are trying to convey with the exhaustive list", you should ask that question of Glst2, who at 15:08 on 20 February 2017 added the 2017 entry to the table of Acronis True Image Versions—a table that you and scope_creep pointed out to me in several comments above as a sterling example of how I should list the Retrospect features. And what we see in that 2017 entry is an "exhaustive list" of all the features in the Acronis True Image 2017 "premium version", and only those features. We see that Acronis True Image's Premium Subscription is just a name for a collection of all the extra-cost features that would individually be labeled Add-Ons in Retrospect. I believe that a major feature of an app is a major feature, regardless of whether it is bundled in or extra-cost. Separating the Retrospect Add-Ons into their own section in the article—which the Acronis True Image article has not done—is just a way of distinguishing features that all users would want from features that only some users would want. I can see no reason why the latter group of major features should not be enumerated just because they are extra-cost, and I'd like to see an WP-rules citation of what you believe is any such reason—especially since it does not seem to have been applied to an article that you and scope_creep have cited to me as a model. DovidBenAvraham (talk) 22:40, 4 November 2017 (UTC)

It isn’t a “major feature”. It is literally an add-on, an extra that only a subset of users may need. And I asked you what you hope to convey with an exhaustive list, not what some other editor might have had in mind in some other article. See WP:OTHERSTUFFEXISTS. JohnInDC (talk) 03:09, 5 November 2017 (UTC)

The same "only a subset of users may need" can just as well be said for the features in the Premium Subscription version of Acronis True Image, which are described in the third and fourth paragraphs of the "Design and features" section of the latest review (and yes, they are precisely the same features described in the 2017 "Versions" entry in the WP article). Moreover, the fifth paragraph of that same review section says "All True Image versions provide bootable recovery media with the ability to restore to dissimilar hardware, i.e., not the same type of hardware that the backup was created on.", so Retrospect Inc.'s Dissimilar Hardware Restore Add-On is Acronis' essential feature.

I would certainly argue that all five Retrospect Add-Ons, especially Dissimilar Hardware Restore and backing up Microsoft Exchange servers and Microsoft SQL servers (all of which you omitted) are important to a rather large subset of Retrospect Windows users (many of whom also have tape autoloaders or tape libraries). And permit me to cite the last paragraph in this section this section of Wikipedia:Other stuff exists; "arguing in favor of consistency among Wikipedia articles is not inherently wrong–it is to be preferred." Do you want me to invoke "Whether a given instance of something can serve as a precedent for some other instance must be decided by way of consensus" for a 1.8-screen-line addition to the article? DovidBenAvraham (talk) 05:24, 5 November 2017 (UTC)

Let me say it again: I don't care what the other article says. I don't think it should be there either, but I also don't feel like embarking on what might be another laborious, endless series of Talk page discussions at yet another of these pages. A Wikipedia reader doesn't need to know the function of every single Retrospect add-on to understand that they exist or what they are. A potential buyer - sure. But they're not the audience. See WP:DIRECTORY, WP:NOT. You don't need to describe every single one, and it reads poorly - and if we are going to look to the other article, it does it much better: "Added active ransomware countermeasures, blockchain-based notary services, and electronic signing (premium version only)." A simple, single English sentence. JohnInDC (talk) 12:10, 5 November 2017 (UTC)

I see that you've decided to stop talking and simply add back in the bulleted list of each of Retrospect's add-ons. I think it makes the article worse, not better, and without any benefit to the general reader (i.e. one who is not comparison shopping in backup software). I'm not going to go to the mat on this because in the end it is trivial; but really now, the time has come for you to stop adding detail and material to this article, and move on. The article remains too technical, too detailed, and too much like a User Manual or Shopping Guide but I'm sick of talking about it. You should turn your attention either to the article that you have proposed to write; or to one of the tens of thousands other articles in the encyclopedia that could use more attention. JohnInDC (talk) 17:24, 5 November 2017 (UTC)

But before that, you have to do better than this sentence: "The backup server Edition is dictated and priced by the number of 'server OS' computers being backed up." This is the first sentence under "Editions" and it doesn't tell you what an "Edition" is, but rather just what "dictates" it. Is an Edition a version of the software? By "dictated", do you mean, "determined"? Does this translate to, "Retrospect sells different 'Editions' of its software, which vary on the number of 'server OS' computers being backed up."? I intended to rewrite this and then realized that I still have no idea what an Edition is. (If that's accurate then is it really necessary to say that it's "separately priced" - doesn't it go without saying that "more computers will cost more"? JohnInDC (talk) 17:39, 5 November 2017 (UTC)

First, I'm sorry that I actually sprang adding back in the bulleted list without giving you a near-simultaneous warning. The idea for re-arranging the two paragraphs in the section came upon me just as I was doing research for a further discussion with you. I tried the re-arrangement out as an edit, and looked at it in Show Preview; it looked pretty good and only took up one more screen line than what we already had. At that point I got a phonecall from a walk-in customer of my little business, saying he was going to arrive in 5 minutes—earlier than expected. I realized I wouldn't have time to write the comment, much less post it, so I hit Save Changes using an Edit Summary I had already written. The business with the customer and its ramifications took much longer than I had expected, so I'm only writing this now.

Second, the result of the research is the WP article Obscure does not mean not notable, and the sections within it. In particular, "In circumstances where using layperson's terms and fully satisfying professional readers' needs are impossible or nearly impossible, editors should only [article author's bolding] meet professional readers' needs and just let general readers know the significance of the field to which the topic belongs." I think I established in the second paragraph of my "01:54, 3 November 2017 (UTC)" comment that any reader getting down to the "Editions and Add-Ons" section is not going to be "one who is not comparison shopping in backup software".

Third, the first page of my passport says "Nationality: United States of America". That's an arbitrary label (see this section of the "Citizenship" article for background) which carries with it certain privileges in certain places. Likewise "Edition" is an arbitrary label, conferred by a license code, which gives a particular copy of the Retrospect executable certain privileges when the code is run. "Edition" is a term thought up by the predecessors of Retrospect Inc. well before 2004 as part of their "soak the rich" concept of differential pricing. I think "license codes, beyond the one that dictates the Edition [my emphasis]" in the first sentence of the second paragraph really clarified that "arbitrariness" concept while staying NPOV, but you didn't like my using the term "license code" in the section. If you think you can convey the "arbitrariness" concept in a better way than that, you are of course free to try. In more specific answer to your question, a backup server's Edition dictates several things it has privileges to do—all of which I tried to describe in the section.

Although I hate to say it, the "license code" elimination is another example of your hastily changing what I have written without taking time to understand it—and then blaming me for the result. Another example of your hasty changing is your cutting-down the last item in the second paragraph. I have changed it back to indicate that the particular Add-On is for additional client computers beyond the maximum the backup server's Edition allows; read the second sentence in the article lead if you still don't understand what "client" means in Retrospect's client-server backup context.

If—in your fundamentally well-intentioned way—you don't mess up the article further, I think I'm done editing it. Although we have had our differences, thank you for your assistance. DovidBenAvraham (talk) 22:17, 5 November 2017 (UTC)

Four paragraphs but I am still left not knowing what a Retrospect Edition is. It's like saying, I don't know, "a mammal is defined by its skin covering, metabolism, and birth and nurturing characteristics". It tells you what defines a mammal but not what a mammal is - a warmed blooded, furred animal that gives live birth and nurses its young. What is an "Edition"? Just a simple declarative sentence - maybe two. If the answer is, "Retrospect sells tiered versions of its software, which offer varying capabilities. Retrospect calls these versions 'Editions'." then can we just say that please? Or if it's the same version but with capabilities that can be unlocked by a code, then say that. "Retrospect is sold at a variety of performance tiers, with functions unlockable through the purchase of license codes. Retrospect calls these versions 'Editions'". The first sentence of "Editions and Add-ons" should say what an Edition is. Please do that, because I can't. JohnInDC (talk) 23:40, 5 November 2017 (UTC)

Or: "Retrospect is marketed in different "Editions", which provide different features and performance depending on the needs of the user and the license code purchased". How about that? JohnInDC (talk) 23:50, 5 November 2017 (UTC)

Your wish is my command, JohnInDC. I chose the second of the suggestions in your "23:40, 5 November 2017 (UTC)" comment, but I scrunched it so it is only a single sentence that adds only one screen-line. I should note that I practically had a heart attack when I read that comment, because you were actually proposing adding a screen-line or more. And your original suggestion used the word "Retrospect" twice! Scope_creep will be running around the room screaming "Advertising! Kill it! Kill it!" when he/she sees that, even though I mercifully cut the sentence down to a single "Retrospect". Sorry, but it's been a strain dealing with the two of you since mid-September. DovidBenAvraham (talk) approximately 01:00, 6 November 2017 (UTC)

JohnInDC, why have you suddenly rejected first–paragraph wording that was based on your 5 November suggestion, and has been in the article since the day after that? The Desktop Edition darned well backs up Linux servers now, so I've put that capability back in while leaving out the mention that Retrospect Inc. intends to take the capability away in a future release. (My friend has verified with the head of Retrospect Inc. Sales that that's precisely what the Release Note means, and that the never-before-used yellow flagging of the Release Note is as close as their mixed-up Documentation Committee can come to an official statement of intentions—but you don't like it so out it comes.) As for the mention of license codes, and that they apply to the backup server, you agreed to those mentions 4.5 months ago as discussed ad nauseam directly above. I agree that one mention of price levels is sufficient to get the idea across.DovidBenAvraham (talk) 27:16, 22 March 2018 (UTC)

Let me explain more fully why I tried to put in mention of the Release Note saying Retrospect Inc. intends to take away the Desktop Edition's capability of backing up Linux clients running on "server-level" distributions, and why I insist on a mention of license codes for the "backup server". As I mentioned in my "02:56, 26 September 2017 (UTC)" comment, for many years the owners of the Retrospect software have had a "soak the rich" differential pricing policy for Server Editions. This has always applied to the "server OSs" macOS Server and Windows Server, but has never applied to clients running a Linux server. One salient fact is that "Starting with Lion, there is no separate Mac OS X Server operating system. Instead the server components are a separate download from the Mac App Store.", although there is still a separately-sold group of Windows_Server OSs. IMHO the probable current difficulty in identifying a macOS computer running the server components is a motivation for Retrospect Inc.'s trying to require a Server Edition for backing up "server level" Linux clients. It looks as if the Retrospect developers would have liked to implement this requirement for the Retrospect 15.0 release (the new version of Retrospect Windows is also 15.0 instead of 13.0, skipping two whole-number versions), but ran into a technical delay. That seems to be the only reasonable explanation for that yellow-flagged Release Note, which is the first one that has ever mentioned a future Retrospect feature. IMHO the engineers' motivation for leaving in that Release Note is rooted in the fact that a user's license codes apply to all minor versions ("dot releases") of a Retrospect major version. Thus if I upgraded to Retrospect Mac 15.0 and later downloaded Retrospect Mac 15.1 or 15.5, I might unexpectedly have my Desktop Edition "backup server" inform me that the new "dot release" does not allow me to back up my Linux client (which BTW I don't have). I wanted to give administrator readers of the WP article a "heads up", so that they could warn their bosses that they would sometime this year have to come up with the money for a Server Edition license code. DovidBenAvraham (talk) 01:29, 25 March 2018 (UTC)

Wikipedia isn’t a Buyers’ Guide or purchasing advice site, and we don’t report product changes or revisions that companies merely say they’re going to do. Please wait until it’s real. Thanks. JohnInDC (talk) 11:27, 25 March 2018 (UTC)

I understand that rule, which is why I didn't mention the Web Console that the new User's Guides' "What's New" chapters announce will be a preview release in May 2018 (which, when its control facilities are fully operational—which IMHO won't be before September and may not be until March 2019, will finally solve the problem plaguing Retrospect Windows discussed in the last sentence of the "History" section of the article). My underlying problem is that the "What's New" chapters of the User's Guides have been turned into cut-down regurgitations of Retrospect 15's marketing documents. They are so lacking in useful overview content that I couldn't use them as refs, and had to wait for Agen Schmitz's Tidbits article (which is so short it omits the significant "AI" improvement in the Proactive script feature—which fortunately has documentation with overview content in a Knowledge Base article) before I could put the Retrospect 15 features into the article. Thus Retrospect Inc. saw its way clear to announcing an upcoming good feature in a non-press-release document, but not to announcing what obviously will be an upcoming bad feature for many administrators. Evidently someone's conscience made them leave the latter announcement as a Release Note. You can at least see why I tried to publicize it for readers of the WP article. DovidBenAvraham (talk) 14:39, 25 March 2018 (UTC)

Perhaps; but if you’re ever inclined to include information in an article to “publicize” it, then odds are the edit doesn’t belong. It’s not what we’re here for. JohnInDC (talk) 16:43, 25 March 2018 (UTC)

Enterprise client-server features / self-referential text

I reduced this section to a simple statement that Retrospect provides most of the features set forth in the linked article. I changed the wording because as a general matter we don't refer to other articles as "other articles" as the original phrasing did. Rather we just wikilink appropriately to concepts discussed elsewhere in Wikipedia. I removed the sentence about the one feature that Retrospect doesn't support along with the comment that the omission wasn't that important anyhow, because we aren't a shopping guide or feature set resource; because Retrospect's comparative subset of supported features may change over time as the other article evolves, making this level of specificity a liability in this article; and because the observation that a different Retrospect capability more or less makes up for the omitted one is OR, and / or synthesis, and - again, this level of detail isn't necessary. If a reader wants to know more about enterprise client-server features, they can visit the linked article. If they want to know how Retrospect stacks up, feature for feature, with the things described there, they can visit Retrospect's website. JohnInDC (talk) 18:47, 27 November 2017 (UTC)

I've now met JohnInDC's objections by adding a sentence saying that the features Retrospect supports are shown by refs in items in the linked-to article section. This eliminates his concern about feature specificity requiring future changes to this section, because Retrospect's "comparative subset of supported features" will be shown only in the Enterprise client-server backup features section of the "Backup" article. In my added sentence I referred to that linked-to section obliquely, to meet JohnInDC's objection to my "other articles" wording. I also changed the word "Retrospect" to "The software" in JohnInDC's sentence, so that I could mention "Retrospect" once in my added sentence without being accused of advertising by scope_creep.

I just removed it. Just link to the other article. Please don't add editorial or other commentary about what the link is supposed to signify or how the other article is structured. JohnInDC (talk) 03:37, 28 November 2017 (UTC)

On top of which. The other article section isn't about "Retrospect" or "NetBackup" or any other particular software product. It's about the general idea of enterprise client-server backup software, and concepts that have emerged as important to the field. (Let's set aside the large OR problem right there, seeing as there really are no 3d party sources that say much of any of that.) So while right now that article relies heavily on the capabilities of those two programs, there's nothing at all about the general field that requires that to be the case going forward; and in the future the article may be written or revised entirely without reference to Retrospect, or NetBackup - making the promise here ("the other article will have a Retrospect cite if Retrospect supports a feature") impossible to guarantee. Articles stand by themselves. Don't create editorial cross-dependencies. Just link and be done. Thanks. JohnInDC (talk) 03:53, 28 November 2017 (UTC)

In the comment immediately above, I suspect JohnInDC violated the maxim "Do not move fingers on keyboard until brain is fully engaged" (that's an old joke from the days of having to learn to drive stick-shift). First, I don't think there's a Wikipedia requirement that any link to another article must be considered in terms of how that other article might change in the future. Second, the "Enterprise client-server backup features" section of the "Backup" article really does not rely heavily on the capabilities of Retrospect and NetBackup—other than as examples; all the features are also covered—as JohnInDC cautioned me to do in his "12:05, 22 October 2017 (UTC)" comment above—by refs to 10 third-party articles that state (sometimes interspersed with other material) the need for those features. Someone in the future could cut the examples, and the basic description of the features would still stand. Third, let's assume that someone does in the future do that cutting; there would be a simple fix to the Enterprise client-server features section of this article that would add (oh horrors!) about 6 or 7 screen lines: Look at the first two items in the Small-group features section of this article, and notice that they cover 10 Retrospect features in only 5 screen lines. I managed to do that by, following the guidance of scope_creep, making practically every mention of a feature also be a link to that feature in another article, thus eliminating the need for a description of the feature. If I simply named Retrospect's 11 enterprise client-server features that are not already mentioned as small-group features, with each name also a link to the still-extant description of that feature in the "Backup" article, I could do that in no more than 7 screen lines—even though the names of the features are longer. In that case I'd move the Retrospect documentation/review refs that had been cut from the "Backup" article into this article, but the footnoting for those refs wouldn't add appreciably to the 7 screen lines. DovidBenAvraham (talk) 03:52, 29 November 2017 (UTC)

It's not about the content of the other article but rather about express reliance on the existence and placement of specific references in that article that - as you note - aren't necessary to the article's content. It’s an easily foreseeable and easily avoided problem. JohnInDC (talk) 12:27, 29 November 2017 (UTC)

We've got a real problem here. JohnInDC seems to be saying that the only way for readers of the section in this article to find out which enterprise client-server features Retrospect implements is for them to do what IMHO is WP:OR in the corresponding section in the other article—but I'm not allowed to give them any help! If I had been allowed to leave in the sentence "Those features it supports have references to Retrospect documentation or reviews in their feature items, as linked-to in the first sentence of this section.", the reader would know that he/she could find out which features Retrospect implements by running a mouse pointer over the ref numbers. But JohnInDC is saying I'm not allowed to leave in any form of that sentence, so the reader will have to guess how to discover which features Retrospect Retrospect implements.

I'm reminded of an experience I had two years ago, while writing my initial Wikipedia article about a notable friend who had recently died. There was a key biographical fact that my friend had told me, but which had not been included in any articles published about him (the fact was trivial while he was alive, but significantly affected the course of his later life). I was first told by an editor that "as told to" is not an admissible source, but then told "Wikipedia is inadmissible as a ref source, see WP:CIRCULAR; and this whole paragraph reads like WP:OR" when I tried to imply the fact by dates and links to WP historical articles. Next I was told that "Whether it was likely that ... would have been sent overseas must be left to the imagination of the reader" was "adding unsourced opinion, see WP:NOR". Finally I had to create a 5-screen-line paragraph, in which all but the first sentence consisted of referenced-to-fairly-obscure-sources historical recitations implying why the date cited in the first sentence meant my dead friend wasn't sent overseas at the end of WWII.

It appears that the only way that allows a reader to find out—without doing what amounts to WP:OR—which enterprise client-server features Retrospect implements is for me to do now what I suggested in the last sentence of my "03:52, 29 November 2017 (UTC)" comment. Do I have your permission to add those 6-7 screen lines to the section, JohnInDC? Do you have an alternative suggestion? DovidBenAvraham (talk) 22:04, 29 November 2017 (UTC)

It would be better to have the material here than linked to footnotes elsewhere. And whether it adds 7 screen lines or 3 screen lines or 14 is beside the point if the material is cumbersome, or unclear, or repetitive - "excessive detail" isn't measured by how much screen acreage is taken up but rather by how well or poorly the article reads as a result of the addition. I suggest something like, "Retrospect also supports several enterprise client-server backup features, including..." and then listing the more important ones; and without going into whether it's "most" or "almost all" of the collection described in the other article. JohnInDC (talk) 22:41, 29 November 2017 (UTC)

Thanks, JohnInDC. The only reason I haven't already listed the names of the enterprise client-server features in this article is that I know your aversion to increasing the length of this article. Unfortunately, because I run only a dinky home installation of Retrospect, I don't actually use any of the enterprise features myself. Therefore I am totally unqualified to decide which are the "more important" features; I can only judge by what my friend has seen on the Retrospect Inc. forums. For example there is a senior IT tech at the Texas A&M College of Engineering (unusually, he posted his name and employer) who is really worried because Retrospect Windows takes 18 hours to do a full-volume scan of his large NAS; he would love to use pre-scanning, but it doesn't work because the Isilon NAS runs a FreeBSD-derived OS instead of a Retrospect client under Windows or macOS . For another example, there is a tech who has spent months writing her own Bash shell scripts using the Script Hooks feature—which she has helped debug—to output backup events for her Windows and Mac clients as a file from her Mac backup server; she doesn't want to use any of the three monitoring systems for which Retrospect Inc. has already written shell scripts in the appropriate languages. Thus I think I'll have to name all the 11 features that are left, once I exclude (because they're already mentioned as small-group features) "Multi-threaded backup server" and "E-mailing of notifications" and also "Avid production tool support" (because—as I've said in the "Backup" Talk page—I'm not yet sure whether Retrospect's adding of this feature is prescient for enterprise use or just a goodie for backup administrators at Mac video-producing installations). I will not include any detail—let the reader look that up in the links and refs, and I promise not to say whether these are "most" or "almost all" of the features. DovidBenAvraham (talk) 01:40, 30 November 2017 (UTC)

Well, just don't say which are the most important. Then you can just list a few ("including"), using your editorial judgment. No one will complain about a list like that! JohnInDC (talk) 02:12, 30 November 2017 (UTC)

I ended up listing the names of 12 features, which included the 11 I had planned plus advanced network client support—added because I decided support for multiple network interfaces is vital for an enterprise with multiple LANs or a WAN. However they only took 5.1 lines on my screen. BTW, in my "03:52, 29 November 2017 (UTC)" comment I didn't mean to disparage your "03:53, 28 November 2017 (UTC)" comment that "in the future [my emphasis] the article may be written or revised entirely without reference to Retrospect, or NetBackup." It's just a difference of opinion as to how soon that future will come. I agree that the third-party refs for that section of the "Backup" article are fairly meager; I couldn't find better ones. Therefore for some years I expect that references to first-party articles from multiple developers will also be needed to support the necessity of the features described in that article section. And finally their necessity will become "received wisdom", as I think happened by 8 years ago for the discussions in other sections of the "Backup" article—which IMHO is why those discussions are allowed to exist with so few references. DovidBenAvraham (talk) 03:48, 30 November 2017 (UTC)

And now an "Enterprise client-server features" section formatting question, JohnInDC: The link to the same-named sub-section of the new section in the "Backup" article is in the words that follow "For" and precede the comma at the beginning of each sentence, except for the first sentence, in the article section. I saw no point in repeating the link for each feature name within a sentence; the link would be to exactly the same "Backup" sub-section. It would make the section read a lot better if each sentence was formatted as a separate list item. That would add about 3 screen lines to the section, because I could eliminate the words "For" and "these include" in each sentence. The question is whether I should use an unordered or description list. Description lists, which are used in all sections and sub-sections of the "Backup" article, would add an additional 3 or 4 lines to the section because they bold each item name and put it on a separate line from the rest of the item. I now think I'm immediately going to format the section as a description list and see how you like it; if it's too many lines, it'll be about two minutes work to reformat the section as an unordered list. The immediately-preceding "Small-group features" section is an unordered list. DovidBenAvraham (talk) 09:06, 30 November 2017 (UTC)

I reformatted the features in the section as a description list. That added 4 screen lines, but actually decreased the number of words—as I predicted in my "09:06, 30 November 2017 (UTC)" comment. The section looks great; it's much less "busy" as the interior design mavens used to say. I hadn't heard from JohnInDC that he doesn't like it, so I did the same for the "Small-group features" section of the article. That added 5 screen lines, but again decreased the number of words. I considered doing the same for the Add-Ons list in the "Editions and Add-Ons" section, but I rejected that idea because the items in that list work better as an unordered (bulleted) list—which they are now. DovidBenAvraham (talk) 10:25, 1 December 2017 (UTC)

JohnInDC, this morning you reverted my edit, saying "Undid revision 813868251 by DovidBenAvraham (talk) - no, this is implicit. They're blue. People know to click on them." Well, when we're talking about description list headers—which are bolded, you're totally and completely wrong. If you can see any real color difference between the bolded headers in the "Small-group features" section—which do not have any links with the exception of "Success validation"—and the bolded headers in the "Enterprise client-server features" section—which do all have links, then it may not be too late for you to let your keen eyesight qualify you for a career as a major league baseball pitcher. Blue just isn't very visible over bolded black lettering.

Just to verify most people's inability to see blue over bolded black, we've already inadvertently run a one-person test. The test was run on 4 December 2017, in this section of the Talk page for the "Backup" article. The test subject should be very familiar; it's you yourself. At 11:53 you wrote "I also don't think that Avid production support needs to be singled out. First it's not clear what's problematic about that one application that it needs its own special backup processes." At 13:14 I replied "If you'd just clicked the first three words of the bolded item title on the "Avid production tool support" item, you would have been transported to Media Composer (I changed the link yesterday from Avid Technology to make it clear to the truly ignorant). There you would have been greeted with ...." At 14:23 I wrote "Your mistake this morning, which resulted in your replacing a clunky subsection title with an inappropriate one, shows that the idea of looking for a link in the bolded item header of a description list is not obvious to all readers. Therefore ... I intend later today to add a parenthetical clause that essentially says 'To see the description of the features in a particular list item below, click on the bolded item header.'"

My intention of 14:23, 4 December 2017 is the change you just reverted. Given that it is unquestionably proper to put links to the descriptions of the "Enterprise client-server features" into that section, I can see only two possible alternatives. One alternative, which I have heretofore rejected because it is an idiotic waste of WP disk space, would be to copy the links in each bolded header into the name of each feature underneath that header. That's idiotic because every link in a feature name underneath a header would be exactly the same, since WP doesn't allow me to create links to anything smaller than a sub-section in an article. The other alternative is the parenthetical clause you reverted, which would make the current non-duplicated links solely in the bolded headers work for people without extraordinary color vision. It's up to you, JohnInDC; which alternative do you want? DovidBenAvraham (talk) 23:21, 5 December 2017 (UTC)

I can see the difference pretty easily. If you think visibility is a problem, reformat the lists to show italic. Moreover, color isn't the only cue to a link: When the mouse pointer is hovered over a link, the pointer changes shape and the text shows as underscored. These are standard Wikipedia conventions that work find throughout the encyclopedia. Editors supply the links, and people reading the article in Wikipedia, on computers or phones, know to look for and to click those links. Mirrored sites may not connect the links to their targets, and hard copy printouts (yes it happens) can't use the links at all; and text referring people to links that they can't use is undesirable. Explicit reference to article hyperlinks, as such, just isn't done in article space, so please don't do it here. (As for the Avid example you offer, I saw the link, and I clicked the link to see if it told me anything more than what I already knew - namely what Avid is - which it didn't.) The section is fine the way it is. JohnInDC (talk) 00:36, 6 December 2017 (UTC)

I switched them to italics. If you don't like that look then they could also be plain text, or bulleted, or some other format that doesn't obfuscate the wikilinks. JohnInDC (talk) 00:42, 6 December 2017 (UTC)

It took a while but I found it: Wikipedia:Manual of Style/Self-references to avoid; also Mystery_meat_navigation#"Click_here". JohnInDC (talk) 00:55, 6 December 2017 (UTC)

Here are some ideas for list formats - not sure which might be suitable for lists with a bit of narrative included, but maybe some are: Wikipedia:Manual_of_Style/Lists#Bulleted_lists. JohnInDC (talk) 00:59, 6 December 2017 (UTC)

A funny thing happened late last night, JohnInDC. My regular desktop computer, a MacBook Pro laptop, really went on the fritz (the screen went all weird and I can't turn it off). I therefore proceeded to do a restore of its Retrospect backup this morning onto a portable hard drive. While I was waiting for the restore to complete (it has to systematically delete all files already on the drive, and then the restore takes about 2.5 hours), I fired up Firefox on my Mac Pro backup server and prepared to give you hell for turning the bolded list item headings into italic—since the ones with links were just as non-blue as before. But viewed on the Mac Pro they are blue, and so are the bolded ones when I go back to the old versions via View History. Therefore what I thought was a problem with Wikipedia actually seems to have been some kind of problem with viewing blue on my MacBook Pro. I will therefore, when I get the chance after taking in my MacBook Pro for repair, revert the article to what it was before I put in the parenthetical phrase you objected to—since that phrase is really not necessary for other readers. I'm sorry for my acerbic comment yesterday. DovidBenAvraham (talk) 06:17, 6 December 2017 (UTC)

Hah, that's pretty funny. Thanks. Apology accepted, no problem. JohnInDC (talk) 11:53, 6 December 2017 (UTC)

This has gone from funny to weird. On Wednesday 6 December I took my MacBook Pro to the repair shop; the logic board was dead (and the machine is so old that Apple has stopped providing replacement parts), so I bought a new MBP to replace it. Meanwhile I was booting my Mac Pro, which is normally my Retrospect backup server, from the portable HDD Retrospect restore of my old MBP to try to carry on business as usual. Using Firefox from there, the links on this article do not show as blue. However using Safari (Apple's built-into-macOS browser) from that same boot drive, the links show as blue. They also show as blue when I boot my Mac Pro from its normal drive and use Firefox or Safari from there. And, now that I've got my new MBP migrated from the portable HDD Retrospect restore of my old MBP, the links still do not show as blue from Firefox on my new MBP, but they do show as blue from Safari on the new MBP. So there seems to be something weird about my MBP's copy of Firefox, but not the same version of Firefox in general. I've looked for Firefox preferences, and even changed a couple as an experiment, but nothing seems to help. Suggestions, anyone? DovidBenAvraham (talk) 14:00, 8 December 2017 (UTC)

Preferences -> content -> colors didn't help? JohnInDC (talk) 15:16, 8 December 2017 (UTC)

The weirdness is exorcised. Two nights ago, when I looked at Firefox's General Preferences, I experimented with changing the default font under Fonts & Colors, but failed to notice the Colors button at the bottom right. Clicking that button brings up a dialog with color-chooser buttons for Unvisited Links and Visited Links. The Unvisited Links default to blue, but the Visited Links default to black. I've now changed Visited Links to purple, which distinguishes them from blue but doesn't confuse them with red links. The reason the links were showing as blue in my Mac Pro normal boot disk's copy of Firefox is that I don't do any WP reading on that machine (the main reason I installed Firefox on it was to check LAN connectivity—if it can see the Web then its connectivity through switches and MoCA adapters to the modem in the other room is OK—back when Safari wasn't so good), so I don't visit any links from it. Safari doesn't have any Links color Preferences at all. Thanks, JohnInDC. DovidBenAvraham (talk) 19:46, 8 December 2017 (UTC)

Edited JohnInDC's pseudo-lists that used italic headings back to description lists, now that I'm satisfied those will show Visited Links in a color other than black in my main computer's copy of Firefox. I used the term "revert" in describing the edits, but I actually manually copied the item headings back from an old version so I wouldn't lose GünniX's Space Patrol changes of underscores to spaces on the pre-vertical-bar side of links. BTW I knew what unordered (bulleted) lists were last year, used them in the old versions of the article, and still use one in the "Editions and Add-Ons" section. But description lists are more suited to the preceding sections. DovidBenAvraham (talk) 03:33, 9 December 2017 (UTC)

Saturday is my big backup day, when I do full Retrospect backups of all 6 drives. I took that opportunity this afternoon to do a bit more testing of the article using my Mac Pro backup server's normal boot drive copy of Firefox. I have not yet changed that copy's Visited Links color preference from black to purple, so that when I clicked a link in the article and then went back to it its color had changed from blue to black. The underscoring still shows on a visited link, but the underscoring is then quite difficult to see on a visited link that has bolded text—such as the heading for a description list. I therefore feel more justified about my "23:21, 5 December 2017 (UTC)" comment; I of course had visited all of the description list heading links to test them as soon as I created them, and I frequently don't bother to switch to my computer glasses when doing editing. So the problem I uncovered here, with JohnInDC's advice, is really a human-factors conflict between Wikipedia's description links facility and Firefox's default for its Visited Links coloring. DovidBenAvraham (talk) 21:10, 9 December 2017 (UTC)

New features 2018-2019

The first new feature I can write about, other than backup/restore and migration/syncing of email accounts and GDPR-compliant "grooming", is the availability of multithreading in the Desktop Edition of Retrospect 15. Because there is one unified code base below the UI level, a multithreading capability has existed in the Engine for that Edition since 2009—but it has been disabled by snapping-back of the maximum allowable activity threads (termed execution units in the Windows variant) Preference to 1 whenever the Engine starts to execute what the license code indicates is the Desktop Edition. Retrospect Inc. has now deleted the (probably two lines of) code that did the disabling, but they haven't had the chutzpah to actually announce that improvement—IMHO because it would show the underside of their "soak the rich" Editions policy. Therefore the only ref I could use is a bug fix note in the cumulative Release Notes for the Windows variant, and I've consequently had to add a note—directly following the cite of that ref—explaining that "execution units" (the term in the Release Note) is the Windows-variant term for what the Mac variant terms "activity threads".

Because many backup administrators are still running an earlier version of Retrospect Desktop Edition, I've sought to avoid confusion by briefly mentioning that the disabling of multithreading still applies to those earlier versions. Also, because "multithreaded backup server" is listed in the "Backup" article as a Performance feature of enterprise client-server backup applications, I've avoided adding an extra screen line doubly-mentioning the feature in this article by adding the mention as a parenthetical addition to the appropriate paragraph header. DovidBenAvraham (talk) 17:41, 3 November 2018 (UTC)

The second new feature I have written about is the Web-based Management Console—with optional deployment of shared scripts. The marketing material—which I won't use as a reference—says that option requires an Add-On, but licensing of that Add-On is not yet available in the Configurator for online purchase. More than two weeks ago two senior Retrospect Inc. salespeople told my friend that the Add-On would soon be included in the Configurator—but that hasn't happened yet. The likely reason is that the Knowledge Base article for Shared Scripts says "Note that as of March 5, 2019, deployment options are limited to ProactiveAI scripts with standard source containers ... to cloud destinations with simple scheduling options. Support for local sources, local destinations including disk, scheduled scripts, and more extensive scheduling options will be available soon." So the basic capability for deploying shared scripts is officially released, but in Retrospect 16.0 it's so minimal as to be in practice useless for ordinary backup administrators. OTOH aggregation-drilldown within organizations, which also requires the same Add-On, is immediately useful to Partners—the Retrospect Inc. term for consultants who market the backup software to organizations. So it looks to us as if Retrospect Inc. is currently only marketing the license code Add-On to Partners, although a non-Partner administrator can buy it by phoning Retrospect Sales and saying "pretty please with cherries on top". I had considered removing Shared Scripts from the article, but enhanced deployment options are likely to be made available with the release of Retrospect 16.1—which past history shows is likely to happen around 15 May 2019. The only factor causing a longer delay would be if the developers cannot resolve the question of whether the Management Console GUI for defining Shared Scripts should look like the equivalent GUI in the Mac variant or the Windows variant (what they released on 5 March seems to us more like the Mac variant). DovidBenAvraham (talk) 05:00, 13 April 2019 (UTC)

I added a link to a marketing article[1] in the "User interface" paragraph of Enterprise client-server features. It's the only reference that shows screenshots of the enhanced 2019 version of the Dashboard, and I put an at=screenshots parameter in the ref. I'll replace the ref as soon as Retrospect Inc. publishes the screenshots in an updated version of the User's Guides or a Knowledge Base article. I added the same link to the "High-level/medium-term reports" paragraph of the "User interface" section of the Backup article . DovidBenAvraham (talk) 00:34, 15 April 2019 (UTC)

To make it clear, the Web-based Management Console as of March 2019 is—unless the Add-On is licensed—just a Web-server-based implementation of the enhanced Dashboard that is is in all Editions of both variants of Retrospect. In Retrospect Windows the Dashboard is a separate non-Web-based program. In Retrospect Macintosh the Dashboard is the initially-presented panel of the non-Web-based Administration Console. That's why, in the Macintosh screenshot in the linked marketing article, a sidebar is shown on the left with clickable categories of backup server information for the selected backup server. No such sidebar is shown in the Windows screenshot, and it's not clear how an administrator selects the particular backup server the Web-based Management Console is supposed to display. As of March 2019, Retrospect Inc. has been remarkably reluctant to show—rather than just describe—the version 16.0 enhancements in any Knowledge Base article or in the updated User's Guides. I get the feeling that their engineers were really rushing to get Shared Scripts implemented at all, and therefore relied on someone from Product Management—which is of course marketing-oriented—to document what they were implementing. That would explain why the referenced marketing article really documents features that were already in beta for later releases of version 15. My friend intends to threaten Retrospect Sales with my deleting the Retrospect 16 features from this WP article, to try to get Sales to put the screws on the developers to document better—which would of course benefit administrator users. DovidBenAvraham (talk) 02:30, 18 April 2019 (UTC)

Retrospect 16.1 is now released, but for Shared Scripts in the Management Console it doesn't implement "Support for local sources, local destinations including disk, scheduled scripts, and more extensive scheduling options ...." However it does include "Retrospect Management Console: Pause/Unpause/Stop support" and "Retrospect Management Console: Disable deployment for an existing shared script", which IMHO are important steps in giving—with the Add-On—the Management Console 2-way capability. Therefore I'm inclined to leave the mention of Shared Scripts in the article until at least the beginning of September, when past history indicates there will be a Retrospect 16.5 release with significant feature additions that didn't make it into 16.0. Probably for the 16.1 release the engineers were giving priority to fixing 9 significant bugs, of which 4 for Retrospect Mac (only 2 of which affected Retrospect Windows) involved Storage Groups—which are a higher-priority enhancement to the "Proactive scripts" feature used by enterprises that need to evade their "backup window". DovidBenAvraham (talk) 05:04, 16 May 2019 (UTC)

Is it true that the article "Relies too much on references to primary sources" for articles on software applications?

On 21 August 2019 JzG put the {{Primary sources}} tag on the article. When I asked on his personal Talk page "What must I do to justify removing the tag?", Guy (that's how he prefers to be known) replied "Remove anything that's sourced to their own websites or to press releases", and later "It's really quite simple. Only include reliable independent secondary soruces. Don't include anything that independent commentators haven't thought significant enotgh [sic] to cover. Don't inlcude [sic]sources that are obviously based on press releases (aka churnalism). Don't include WP:HOWTO or other manual-like content."

I won't re-hash that discussion here, but the argument I initially made is that "The basic problem is that this is an article about a client-server backup software application with a 30-year history" and that "Secondary-source reviews simply don't mention all of the features of a software application; for those a Wikipedia editor must fall back on the primary-source application manuals and knowledge-base articles." Guy's "haven't thought significant enough" argument has since been rather punctured by my discovery that the same Dennis Publishing reviewer who reviewed Retrospect Windows 7.5 on 20 April 2006 just did a review of Retrospect Windows 16 on 24 July 2019. No doubt because of the 13-year-gap between reviews, Dave Mitchell's new review does in fact briefly allude to many long-time Retrospect features.

Motivated by that discovery, I have embarked on a project of replacing cites of Retrospect primary-source refs with cites of the Mitchell 2019 ref and the Kissell 2007 ref and the Engst 2009 ref—all of which I have enhanced with additional annotated page numbers and/or quotes. The secondary-source refs aren't in most cases as informative as the primary-source refs they replace, so I'm leaving those applicable annotated page numbers in the primary-source refs. DovidBenAvraham (talk) 17:48, 31 August 2019 (UTC)

Having found a few more not-terribly-meaty reviews that mention non-"significant" features, I managed to reduce the cites of primary-source references from 52 to 27, out of a count of about 101. A new problem cropped up, which is that some genius(es) in WP's support staff instituted an error message for any reference that doesn't have the website= parameter. I fixed one of those in this article, substituting website=YouTube for via=YouTube in the "Changing paths Cloud Mac" ref. I then went on to update links I had previously put into the "NetBackup" article that now need to go to precise subsections in the split-off "Enterprise client-server backup" article. Since I didn't write that article, a lot of its refs didn't have the website= parameter; because I assume (probably naively) that website= should be followed by the name of a website rather than just the lead portion of its URL, I had to follow the URLs—some of which had to be fixed because they were dead. Finally I started doing the same thing for the "Backup Exec" article, only to discover that the WP genius(es) had returned to sanity and re-eliminated the error message for a missing website= parameter. DovidBenAvraham (talk) 14:04, 4 September 2019 (UTC)

• "It's really quite simple. Only include reliable independent secondary soruces. ", is wrong. There is no policy to justify that.

Guy is conflating two things here: the use of primary sources (and broadly, any source not meeting the letter of WP:RS) with relying on such sources, for the purposes of WP:V.

We have a policy that challenged content must be verifiable (WP:V) by use of WP:RS, i.e. sources other than primary. However there is no exclusion for non-RS sources beyond this, which often means PRIMARY or SELFPUB sources. You can still use those sources, and you can use those sources in addition, provided that anything else which has been challenged (and much simply won't have been challenged, it may even be BLUESKY and self-evident) is also supported by RS. Andy Dingley (talk) 16:00, 10 September 2019 (UTC)

Rewrite of 19:11, 12 September 2019 version of article, replacing Guy's greatly cut-down version

I've eliminated all 14 remaining primary-source reference, which Guy had been complaining about since 09:41, 21 August 2019. I've also cut the total article length by about 0.5 screen-pages, which he and Scope_creep and Pavlor were also complaining about starting in this archived section of the Reliable Sources Noticeboard. Finally I've combined the "Small-group features" section and the "Enterprise client-server features" section into a "Standard features" section, responding to a complaint of Pavlor in in this continuation section of the RSN, which—along with the elimination of a couple of less-widely-used features—resulted in shortening the features discussion length by 0.3 screen-pages.

I've expanded the "History" section to nearly what it was before, because IMHO what happened in 2009 and subsequently is essential to understanding the enduring bad reputation of Retrospect among those who had previously been fairly happy with the application—which I think includes Guy. However I've had to delete the sentences explaining why Retrospect Windows has been stuck for the last 10 years with a klunky GUI compared to Retrospect Mac, because the only available reference for those sentences is a Retrospect Inc. Knowledge Base article—a primary source. I've pointed out that those deleted sentences really functioned as "anti-marketing", so Guy has achieved unintended consequences.

As I pointed out in the first RSN section, Guy considered TidBITS a blog because he didn't take the two minutes to find the WP article detailing its 29-year history as "an electronic newsletter and web site dealing primarily with Apple Inc. and Macintosh-related topics.". IMHO all my other secondary sources are now equally above reproach, with the possible exception of this ChannelPro article. The reason I have been forced to cite that ref a single time is Retrospect Inc.'s recent abandonment of its prior policy of copying and expanding the descriptions of new features in "What's New" chapters of the User's Guides into other chapters in later-major-version UGs. Instead it decided to do the expansion of "facilitating reconfiguration for cloud seeding and large-scale recovery" as two three-minute YouTube videos. Only the video for Retrospect Mac used to explain "large-scale recovery", and its 15-second explanation has been edited out of the latest Drobo-oriented version of the video. Obviously both these videos are first-party references, and the TidBITS editor who wrote the review of the Retrospect Mac version that introduced this pair of features didn't manage to find that video so he could write about the pair of features. DovidBenAvraham (talk) 07:00, 21 October 2019 (UTC)

Oh look, you bloated it out with unencyclopaedic marketing and HOWTO stuff again. I didn't expect that. Actually I did. Your monomaniacal focus on this is becoming rather wearing. Guy (help!) 07:10, 21 October 2019 (UTC)

Just to pile on: the 'standard features' and 'editions and add-ons' in this version just do not show why they have to be in the article, verifiable existence is not an inclusion standard, it should be relevant (and for thát you need independent, reliable sources showing that relevance). As it stands in that version, it is just an indiscriminate collection of information. You will have to show specifically why a certain feature is so special, not just that it exists. Listing features and add-ons here is just promotional material (even when not written with a reason to promote). (and yes, I do note that a lot of other similar articles have the same indiscriminate material, and that should also be removed). --Dirk Beetstra ^{T C} 07:42, 21 October 2019 (UTC)

To both you gentlemen, my general response is: "Show me the Wikipedia article(s) where your alleged rules are stated as official rules." To Guy my specific response is: "In what rules article is it stated that simple statements of particular application features automatically become 'unencyclopedic marketing' and 'HOWTO stuff'?" To Dirk Beetstra my specific response is: "In what rules article is it stated that 'You will have to show specifically why a certain feature is so special, not just that it exists. Listing features and add-ons here is just promotional material ....'?"

To Dirk I would further add that Retrospect's non-enterprise features go well beyond those normally found in personal backup applications (see this January 2019 comparison addendum maintained by the independent backup book author Joe Kissell) , and that its enterprise features—which are also present in the Desktop Edition—make it the equal of much-more-expensive Backup Exec and NetBackup—both of which no longer have the capability of of backing up Macs as of a couple of years ago. You'll have to find your own comparison matrix for "make it the equal"; I haven't been able to find one, and your enterprise client-server backup application is normally dictated by whatever consultant your enterprise happens to hire. DovidBenAvraham (talk) 05:39, 22 October 2019 (UTC)

DovidBenAvraham, I have just one answer: I challenge the statement that the features are special, it is to you to show that they are indeed special. That is what is reflected in all our sourcing policies and guidelines. Mere existence alone, even if verifiable, is not worth mentioning. I can believe that some of the features are 'special', but you'll have to show that they are special. That also likely boils the list down to a smaller size, and that is probably better worked out in prose than in a list-like format. Dirk Beetstra ^{T C} 05:56, 22 October 2019 (UTC)

And with that, some of the text needs to be seriously toned down, it sounds rather promotional in some cases. --Dirk Beetstra ^{T C} 05:59, 22 October 2019 (UTC)

Dirk Beetstra, I never made "the statement that the features are special". Where are "our sourcing policies and guidelines", that say "it is to you to show that they are indeed special"? What does "special" mean—distinct from any other application of the same type? It is my understanding that I need merely show a feature is verifiable and also not "trivial", and that is done by giving an independent second-party reference that mentions the feature. And while you're showing me those "sourcing policies and guidelines", please quote any text from the last version of the article I wrote before Guy started editing it again that "sounds rather promotional". I just wrote a brief history of the key moments in the application's development, plus short descriptions of non-trivial features; how can any of that (other than my copying Guy's previously-added mention that Retrospect Mac 9 was well-received) be promotional? DovidBenAvraham (talk) 05:56, 23 October 2019 (UTC)

@DovidBenAvraham: You never made that statement, but you keep on pushing these features into the article which strongly suggests that they are special. WP:V: "... and any material challenged or likely to be challenged must be supported by inline citations" .. our challenge is that they are not worth mentioning, if they are to be included one needs an independent, reliable source to show that they are somehow special. That is also in line with WP:LSC's "criteria for inclusion should factor in encyclopedic and topical relevance, not just verifiable existence." We do not just list features just because they exist. All spreadsheets do calculations, most cars have 4 wheels, many backup programs backup to tape drives, removable storage, external hard drives, cloud locations and spanning over multiple volumes (I was using backup software in the 90s that did that ..).

To give one example (besides lists of features showing 'look what it does', etc.): "Retrospect is sold with varying backup server capability levels, called "Editions", at non-expiring license–code prices". "Retrospect is available with varying backup server capability levels<full stop>". That ref on Tidbits from 2009 and used 17 times is for version 8 and ending in a lot of pricing information, and Tidbits sponsoring features (which may have been different 10 years ago ...) suggest that sponsors receive some benefits including an article. --Dirk Beetstra ^{T C} 06:33, 23 October 2019 (UTC)

DovidBenAvraham, at this point you are effectively a WP:SPA devoted to buffing up this article to the most promotional state possible. The usual outcome in such cases, is a topic ban. Guy (help!) 10:46, 22 October 2019 (UTC)

Guy, please explain "buffing up this article to the most promotional state possible" with reference to the last-sentence question in my 05:56, 23 October 2019 (UTC) comment directly above your 10:46, 22 October 2019 (UTC) comment. I think you will have a tough time showing that I am a WP:SPA whos merits a topic ban, because I've made it a practice to write comprehensive Edit Summaries. They show the following: All my additions to the article between February 2018 and May 2019 were minor tweaks of less than 300 bytes, with the exception of adding mentions of non-trivial new Retrospect 15 features on 22 March 2018 and a mention of Retrospect Virtual—since deleted—on 29 October 2018. My only additions to the article between June 2019 and 25 August 2019 of more than 300 bytes were either dealing with link changes required by the split-off of the "Enterprise client-server backup" section of the Backup article into a separate article, or reporting the 25 June 2019 merger of Retrospect Inc. into StorCentric. My edits since 25 August 2019 have been exclusively to comply with your adding the{{Primary sources}} tag to the article.

In any case, since March 2018 I have taken primary responsibility for editing all sections of the Backup article and the Continuous Data Protection article, both of which long predate my becoming a Wikipedia editor—and neither of which promote Retrospect in any way. With regard to WP:SPA, I think that shows I am one of those "well-intentioned editors with a niche interest" in backup, rather than one who appears to "edit for the purposes of promotion or showcasing their favored point of view". DovidBenAvraham (talk) 07:01, 23 October 2019 (UTC)

Your inabiolity to see the promotional nature of your own edits is not a surprise. Guy (help!) 08:06, 23 October 2019 (UTC)

I guess that 95% of your last 500 edits to mainspace are to .. what .. at most 10 articles? --Dirk Beetstra ^{T C} 08:44, 23 October 2019 (UTC)

(Un-indenting this so as to respond to both of you) Want to know what makes Retrospect's features "somehow special"? This 2012 Macworld article says "Retrospect 9.0 is designed to fill the space between personal backup products such as Time Machine and enterprise-grade software that protect servers and workstations by the thousands. A server running Retrospect can manage backups for a hundred or more computers .... No other product offers Retrospect’s capabilities in a multi-platform setting for anywhere near the price." I intend to quote that in the article's lead. Two properly-referenced features lists will prove Retrospect can "fill the space" between personal and enterprise backup products.

The "for anywhere near the price" statement is especially true if an installation installs the Desktop Edition, which—even with Add-On licensing of more "client" computers than the 5 permitted by that Edition—costs a great deal less than Backup Exec or NetBackup. An installation would pay a good deal more for a Retrospect license if it runs macOS Server or Windows Server, but most SMEs are now instead running no-Retrospect-charge Linux servers—which seems to be a major reason why Retrospect Inc. arranged in June 2019 to be acquired by Drobo-owner StorCentric. The point of this paragraph is that—especially since Retrospect Inc. quietly eliminated its Desktop Edition no-multi-threading restriction about a year ago—I can now eliminate the discussion of Editions that Dirk Beetstra considers "promotional" and just have an "Add-Ons" feature section.

As for the 2009 TidBITS article "EMC Ships Modernized Retrospect 8" that Dirk objects to because its "ending in a lot of pricing information" suggests "that sponsors receive some benefits [for] including an article", the 2012 Macworld article linked to in the first paragraph of this comment ends with "Note: When you purchase something after clicking links in our articles, we may earn a small commission. Read our affiliate link policy for more details." I need to use older reviews for refs, and I can't help it if reviewer publications 7 to 10 years ago were allowed a bit more leeway in financial relationships with developers—so long as they disclosed them as these reviewers do.

However I now find that I can replace all the 13 features cites of that 2009 TidBITS article with cites of either the 2012 Macworld article or this October 2012 Ronver Systems "EMC Retrospect 8 for Macintosh" Web page, leaving only 4 history cites.. Ronver Systems is a Belgian company that "specializes in XL data storage solutions for the 'content/imaging' industry", but I have definite proof that they didn't simply copy that Web page from EMC Insignia's (owner preceding Retrospect Inc.) PR material; it includes a "Terminology Changes" section that I've never been able to find in first-party documentation. (The 2009 split in Retrospect terminology between the Windows and Mac variants, which I mentioned in a second paragraph of the "History" section that Guy deleted, was a byproduct of the EMC management foul-up that explains a lot of the bad reputation that Retrospect has had since 2009—a reputation which I suspect Guy is aware of.)

So how about it, gentlemen? Do I have your reluctant consent to proceed with the further-shortened revision of my 04:30, 21 October 2019‎ version of the article? DovidBenAvraham (talk) 07:05, 24 October 2019 (UTC)

"Retrospect 9.0 is designed to fill the space between personal backup products such as Time Machine and enterprise-grade software.", done, full stop, over. Put a ref on it, and you are fine. No need to sell it further. The rest of that is making the story promotional. We are not writing an article on a blog or a magazine article. That they are cheaper than similar software is not encyclopedic. That is work for price comparison websites out there.

We are not here to sell it, the pricing argument is not necessary, that is a promotional statement.

I am not fully objecting against the TidBits article, but I do notice that they write articles on subjects when they have paid for advertising on TidBits. That brings the full neutrality of all articles in question. They are not necessarily fully independent of the subject. And that reference was used 17 times.

So no, I don't think I would agree with you editing the article at this point. We seem to have serious disagreements about what language and what material belongs in this article. --Dirk Beetstra ^{T C} 07:42, 24 October 2019 (UTC)

I have recently looked more closely at this page of Online Appendixes to Joe Kissell's latest book on Mac backup. I could use two cites of the first table in that as a detailed replacement for all but a single line each in the "Backup destination" and "Backups" paragraphs and eliminate the "Cloud Backup" paragraph. That, along with the collapsing the "Success validation" paragraph into the "Enterprise client-server User Interface" paragraph and eliminating the paragraph discussing Editions (as discussed in my 07:05, 24 October 2019 (UTC) comment), would eliminate 0.25 screen-pages—leaving only a total of 0.75 screen-pages for both features sections that would only name "special" features.

Dirk Beetstra, as to the 2009 TidBITS article being cited 17 times, I don't think I made it sufficiently clear in my (before later clarification with cite counts) 07:05, 24 October 2019 (UTC) comment that citing the 2012 Macworld and 2012 Ronver Systems articles instead for features would eliminate 13 of those 17 cites—which would remain eliminated in what I have proposed in the first paragraph of this comment. That would still have left 4 history cites, but—unless I put back the second paragraph of the History section that Guy has lately deleted again—those cites would also disappear leaving no need for a ref to the 2009 TidBITS article.

Guy, I think you'd better link to a specific Wikipedia rule explaining why any of the features paragraphs left in what I have proposed in the first paragraph of this comment would be a violation of WP:HOWTO. Why is naming an application feature ipso facto describing how to use the feature? DovidBenAvraham (talk) 12:24, 25 October 2019 (UTC)

Dirk Beetstra and Guy, here's a diff in my Sandbox of the revised article version I proposed in my 12:24, 25 October 2019 (UTC) comment. DovidBenAvraham (talk) 17:40, 26 October 2019 (UTC)

Lovely advert, useless for Wikipedia though. Guy (help!) 17:41, 26 October 2019 (UTC)

Because the bottom of the diff showed the other working items in my personal Sandbox, Guy proposed to delete my personal Sandbox because it's "where this user has been padding the article and pushing back against all attempts to prune it back to something less like an advertisement for months". IMHO that's a pretty cheesy use of what at most was an attempt to be helpful—by using the diff to show the feature cut-downs. Anyway, here's purely my latest proposed revision to the article. I could just copy it into the article, but then Guy would have to again strain his fingers deleting the cut-down features sections. Note the 2 features sections total around 0.66 screen-pages, even with a two-line explanation of Editions. DovidBenAvraham (talk) 05:56, 27 October 2019 (UTC)

DovidBenAvraham, sigh, I have given above an example of a sentence that is clearly promotional, and you insist to keep that one in there. The article has more of those. I have told that most of those features are nothing special, and you insist to keep them in there. The list of features is larger than the actual article. The list of features contains features which are absolutely completely trivial. If there is any feature that is so completely different from anything else out there then that can easily be expressed in a sentence of prose with proper references. The existence of features is not encyclopedic information. Dirk Beetstra ^{T C} 06:35, 27 October 2019 (UTC)

Oh my, Dirk Beetstra, you've apparently failed to thoroughly understand my 05:56, 27 October 2019 (UTC) comment directly above your 06:35, 27 October 2019 (UTC) comment; maybe my English was too complicated for even your excellent skills. 😞 Here, here, here is my latest proposed revision to the article. I didn't copy it, didn't copy it, didn't copy it into the article because—based on past experience—Guy would have immediately deleted the two features paragraphs. If you read that proposed revision, you'll see that I've (1) deleted the pricing sentence that you correctly said was promotional, and (2) eliminated all mention of the usual "completely trivial" features by simply citing the Joe Kissell 2019 table that describes them in two single-screen-line sentences—and those specifically mention other backup features that are not at all usual in consumer backup applications. DovidBenAvraham (talk) 07:22, 27 October 2019 (UTC)

"Retrospect is sold with varying backup server capability levels, called "Editions", with non-expiring license–codes[13] that cover one major version." .. 'is sold with' .. 'non-expiring license-codes' .. all promotional. 'Termed Media Sets — can be on any of the usual consumer storage media, tapes or WORM tapes—with barcoding, or CD/DVD discs.' .. I've been backing up to tape drives in the 90s. I'm backing up to CD/DVD discs for years. And mentioning those features is in itself again promotional. 'Kissel describes them in two single-screen-line sentences' .. so done, two sentences. You make a table which is longer than the remainder of the article. It does not belong. Leave it out. I don't understand why you insist in listing all these features and blowing them up, all it does is turning this article into a promotional piece of text. --Dirk Beetstra ^{T C} 07:34, 27 October 2019 (UTC)

And what you should do, instead of making your spam-fork in userspace, is make the edit and IMMEDIATELY self-revert. Then discuss that revision. But you still fail to understand the source of our contention. --Dirk Beetstra ^{T C} 07:35, 27 October 2019 (UTC)

DovidBenAvraham, Yes, I would have reverted it, because you are basically adding back material that has been removed, discussed, and been agreed to be inappropriate. I fully understand that you do not accept that Wikipedia is not the place for what is in effect a marketing document. I have suggested an alternative venue - Wikibooks - where you can include as much HOWTO and PR detail as you like, but you seem very reluctant to accept this. What you need to understand is that however hard Retrospect try to spin it, there is pretty much nothing unique, or even distinctive, about their product. I was a long-time user of it when I ran Mac networks and when I worked for an Apple reseller. I deal with backup software in my daily life. I know the product landscape. Restrospect is not seen as a significant player, and Wikipedia is not the place to fix that. We have bigger articles on more significant things that are discussed more widely, and smaller articles on minor things. That is how it goes. The current article focuses on analytical non-paid-for sources that address the company and its history. You're including churnalism, paid-for analyst work and the like. I have been there when these things are written, I know some of the people who write them They are marketing documents, not independent work. The company pays the analyst then places the story in the trade press. If you search, you can find me quoted in some articles of that type. I have seen the sausage being made, and the marketing people were in the room at the time. Guy (help!) 10:03, 27 October 2019 (UTC)