Talk:Microsoft Visual SourceSafe

Clarity in Naming

I am aware of Visual SourceSafe 6.0. This article discusses Visual SourceSafe 2005. If there is a relationship between those two, it should be spelled out explicitly. If there are other versions (which I suspect there are), they should be listed. This is a plea for Really Basic information.

--This article is poorly written and structured.

Suggested format:

• Introduction - what does it do, how do you use it
• History
• Alternatives —Preceding unsigned comment added by 74.174.17.194 (talk) 20:44, 2 May 2009 (UTC)
• Criticisms - explain how sourcesafe works - filesharing, say that it is used by many hundreds of thousands of developers, but has been criticised for lack of robustness: explain problems, quotes, external links to criticisms
• Sourcesafe 2005 - what's new? Does it fix the issues it was criticised for?

I agree with your assessment. While I don't currently have the time to improve the article, I just want to note that SourceSafe 2005 does not fix the issues SS was criticised for. Unless Microsoft managed a quantum leap in the few weeks between Whidbey RC (which I've been testing) and Whidbey Final, SourceSafe is still extremely sluggish - sluggish enough over LAN already but a real horror to use over UMTS -, behaves erratically, and the GUI of SourceSafe Explorer is still sorely lacking and often confusing. For example, when you diff a file against the repository version in the main view, the local version is displayed to the right and new passages appear in green. When you perform a project diff and then select the very same file for comparison, the local version is on the left side and new passages appear in red, as if they had been deleted. There is still no "Annotate" command as provided by Subversion, and as CVS has offered for ages. This makes identifying the revision in which a certain change originated a major pain. Most interesting dialogs (diffs, file histories etc.) are still modal which can be a real nuisance. The application seems to be single-threaded as well and often doesn't respond at all during a lengthy operation. Other more fundamental problems haven't been fixed, and if for the sole reason that this would have required breaking changes to the repository format. For example, the branching support is still a joke and moving a directory ("project") means to break all previous revisions, which can then no longer be "faithfully reconstructed" as MS themselves put it. The repository access protocol is still vulnerable to connection failures, and corruption can ensue. The most visible changes from VS 6 are that Microsoft changed the splash screen and icons. Twice. At least.

In short, not only does much of SourceSafe appear unfixable, but Microsoft also does not even seem to be willing to fix the easier to fix problems. A more straightforward solution than trying to cure this terminally ill patient, would be to develop and promote an entirely new source control system, and this, of course, is exactly what Microsoft is doing right now. The good news for me personally is, first, that our team did not experience (or notice, that is) data loss so far, and second, that we're migrating our projects to Subversion in a few weeks time. Aragorn2 15:34, 25 November 2005 (UTC)

OK, I rearranged the page. What needs to be rewritten? This article still needs some work. Pinano 23:47, 27 November 2005 (UTC)

I added some more info about VSS 2005, Team System, and Team Foundation Server. If you read between (and sometimes in) the lines in the documentation and chat transcripts, MS is pushing people towards Team System big time. Too bad it's so expensive. I also updated the advantages section ... VSS isn't as affordable as most people think. Ben 17:01, 4 December 2005 (UTC)

Re the MainSoft thing, citation needed. Googling Mainsoft + "visual source safe" resulted in several press releases from 1998-1999 (example), but the MainSoft site itself contains only one reference to Visual Source Safe, and that's in a forum question whose answer is "Source control integration is not supported in the developer edition. However, this feature is supported in the enterprise edition, see: http://dev.mainsoft.com/Default.aspx?tabid=32&src=FeatureMatrix.html" Anonymous 23:12, 23 March 2006 (UTC)

I suggest that "SCM" should be defined and linked. I assume it means "Software Configuration Management". In the context of this article (Visual SourceSafe), readers are likely to guess it means something like "Source Control Management"...or does it?--67.69.19.58 15:19, 1 June 2006 (UTC)

• It means Smith Corona Marchant, of course. (And if you have any idea what I'm talking about, you're showing your age). Not-gonna-sign-this-no-way-no-how 20:03, 14 August 2006 (UTC)

--Relative cheapness? It's most popular competitors are free.

Storage of binaries not within the intended scope

A recent change says:

the storage of binaries is not within the intended scope of the software.

I'm marking this as needing a citation, because as it happens I was just installing SourceSafe v 6.0 yesterday and the cheerful little marketing message that flash by as you install it specifically said that one of its advantages was that you could use it for storing all your binaries, Word documents, etc. etc.

Where does Microsoft say this is not an intended use of SourceSafe? Dpbsmith (talk) 12:58, 14 August 2006 (UTC)

A quick run through Google finds many web sites that support your statement. SourceSafe has always had special code to handle binary files in a different way than text files. I've heard the statement before in the real world, but I'm very much of the opinion that it's an after the fact legend made up by apologists when confronted about bugs in the software. The phrase of the day is "The Henny Youngman Solution": "Doctor, it hurts when I do this!" "Don't do that." - Richfife 18:19, 14 August 2006 (UTC)

p.s. It's not a recent change. It just looks that way because I restored a vandalized paragraph - Richfife 20:08, 14 August 2006 (UTC)

I see that Microsoft's Henny Youngman practices article does not say anything at all about not using it for binaries. Dpbsmith (talk) 20:10, 14 August 2006 (UTC)

A bit late to this discussion, but VSS does have problems with corrupting binary files. See http://support.microsoft.com/kb/930282/ - 203.134.166.100 (talk) 05:24, 29 July 2008 (UTC)

Criticism are both too short and too weak for the reality of VSS

I've used both SVN, CVS, TFS, and Source Safe with different clients, and I can tell you without a doubt that Visual source safe was a step backwards in source control technology. I wish I could say that you are probably better off not using source control at all rather than VSS, but I can't in all honesty go that far, though many braver experienced and skilled developers will make such a claim. It really seems like it was microsoft's both cynical and successful attempt to gain control of source control market share, rather than to produce anything worth using. Humanity as a whole has probably been hindered by this monstrosity, because of the negative impact it's had on software development. Too many project leads that don't know better have set their company up on this POS. For this reason I believe that most of this article should probably be about how it completely sucks and why, because that is where the true story of VSS lies. I wouldn't be surprised if the reason that this article hasn't gone that way already is a combination of people having a vested interest in VSS or just technical ignorance. M4bwav (talk) 19:28, 15 January 2009 (UTC)

I agree, but... You have to back that up with reliable sources. - Richfife (talk) 17:01, 16 April 2009 (UTC)

Obviously, dudeM4bwav (talk) 16:56, 21 April 2009 (UTC)

SVN and CVS "better and easier"

Hopefully I'm formatting this right. Talk always confuses me. Some one added stuff to the article that said CVS and SVN have more features and are just as easier or some such. While I don't deny that SVN and CVS have features that VSS lacks, VSS also has features that those lack, like single checkout mode. As for comparing the VSS Explorer and VS Integration to Tortoise and Ank, not by any stretch of the imagination. VSSE and Tortoise are completely different, one based on Windows explorer integration, and the other based on a proprietary tree/explorer type interface. They are not drop in replacements for each other. Personally I have found many, many situations that toroise is just simply not able to perform the same actions as VSSE. Ankh is okay, but again, not "just as good as" VS integration. It just doesn't have all the same features, for example you cannot rename files in certain situations using Ank.

In my experience, SVN + tortoise + Ank is far, far less stable than VSS 6.0c. I have had many local file corruptions with SVN, and I haven't had a serious problem with VSS in years. I do not like VSS 2005. VSS's biggest flaw is in its changeset/merging capabillity, where while it does it well, it does it one file at a time. SVN does handle this better than VSS. I am not going to say that VSS is best and SVN should go away, however this is an article about VSS, and unqualified statements that SVN has more features and is just as easy to use should not be allowed to remain.

Dmprantz 19:16, 19 December 2006 (UTC)

Mainsoft wrote to me to say they no longer make a VSS client for Unix. Hello James, Mainsoft stopped supporting Visual Source Safe on for UNIX platforms several years ago. Regards, Ohad. Ohad Eilon, Product Director, Mainsoft Corp.

VSS also has features that those lack, like single checkout mode. -- Single checkout mode is a "feature"? Hahahahaha. Marnanel 18:58, 26 July 2007 (UTC)

Why is there no mention of alternatives like SVN and Git? —Preceding unsigned comment added by 74.174.17.194 (talk) 20:41, 2 May 2009 (UTC)

That would be like writing about Usain Bolt in the article about Ben Johnson. Qed (talk) 17:36, 8 September 2009 (UTC)

Except that Microsoft offers several alternatives on MSDN page about VSS including SVN and Vault. Leeor net (talk) 05:10, 6 July 2010 (UTC)

Purchase of Source Safe

The article states that source safe was purchased from one tree software, but it does not state a purchase price. I cannot find one anywhere on the web. Does anyone know if the price was ever released? —Preceding unsigned comment added by 24.136.172.12 (talk) 00:15, 29 April 2009 (UTC)

Rewriting Source Safe history

The article trivializes the OneTree product and states that first version actually released was 3.1. This was the Microsoft rebranding of the last OneTree version (2.x, I think 2.4?). OneTree had shipped several versions of SS on half a dozen platforms, including DOS, OS/2, Unix and Mac. The current article reads like it was written by Microsoft PR. Jeff Dickey (talk) 11:22, 7 March 2010 (UTC) To add - it seems that according to all I found, the version Visual SourceSafe 6.0 was released in 1998 and sometimes you can also find it mentioned as Visual SourceSafe 98. Athae (talk) 14:34, 25 August 2011 (UTC)

Using Source Depot, eating own dogfood

Just in case y'all wanted to know: both Office and Windows products are maintained with source depot, not VSS. VSS is not scalable to the number of developers required for such systems. Source depot was derived from an early version of Perforce, which Microsoft licensed. In the case of source control, Microsoft is not really willing or able to play the "eat your own dogfood" game. —Preceding unsigned comment added by Qed (talk • contribs) 22:14, 29 August 2009 (UTC)

When I was working at Microsoft in the UK last year on their consultancy and services arm, they were using Team Foundation Server for new projects. Tasted like horse offel too. -- bitplane — Preceding unsigned comment added by 141.92.129.45 (talk) 09:38, 29 May 2012 (UTC)

shredder

Google hits don't bear out the comment about source-shredder (I get 750 today, which given that much of the search is based on this page, is down in the noise level). Tedickey (talk) 23:08, 29 March 2010 (UTC)

Pure lies

"Version 4.0 and later versions made significant architectural changes, such as replacing the previously-used database layer with the "Jet" database engine used by Microsoft Access."

This is, I can't put it any other way, pure bullshit. I tried to find the one claiming this from the history, but possibly due to not knowing the interface well enough I didn't.

Anyway, Jet was never used for (V)SS. In fact Jet refers to the very DB engine used in access.

(V)SS still uses the same basic on-disk layout as it used when Microsoft bough OneTree (this can be verified by any user).

In no way or form has the Access database engine (Jet, also known as ESI Blue) ever been used. I can not imagine anyone stating these things as a fact simply being misinformed, but must assume malice. Deal with it as you see fit. —Preceding unsigned comment added by 87.227.18.143 (talk) 23:43, 16 August 2010 (UTC)

If it's wrong, or even unsourced, it should be removed. But it seems pretty harsh to ascribe this statement to malice rather than some kind of misunderstanding. As currently written, there's nothing about the statement that seems disparaging, so malicious intent seems quite implausible.

Anyone else know more about the DB used by VSS? I've never used it...

Moxfyre (ǝɹʎℲxoɯ | contrib) 04:18, 17 August 2010 (UTC)

Malice seems unjustified to assume. The article is probably incorrect in stating that Jet was used (http://support.microsoft.com/kb/q152807/ suggests a special file based format (although that doesn't in itself preclude that Jet could be involved), the special file based format theory is also supported by http://www.ezds.com/html/ss_internals_index.html). Very few users today are likely to have access to versions of Source Safe prior to it being acquired by Microsoft in 1994 to verify the claim that the file format is the same, which could very well be the case though given that it still seems to be a special file based format. Although, if the article is correct in that more database corruptions occurs after version 3 then the special file format could have changed somehow.

D00mst (talk) 20:52, 17 August 2010 (UTC)

While it could be seen as jumping to conclusions, I stand by my statement that it's pure lies and can be nothing but malice - as it is stated as a fact "replacing the previously-used database layer with the "Jet" database engine used by Microsoft Access".

First of all, using whatever current or non-current version of VSS, anyone can easily verify that it's not using any MS Access style database.

VSS doesn't use any established database whatsoever. It uses a directory "data" in which it has subdirectories "a"-"z" where it stores files checked in - incl. history and metadata, such as "project" (the subdirectories in VSS), and any potential "security" information (what user can access what, and with what privileges - but this is not to be confused with actual security as it's only data in files that everyone accessing a VSS store anyway needs access to). The name of those files (and therefore the directory they are put in, as the first character of the filename equals the directory name it's put in) is simply a reversed radix-26 string (using a-z) for the running unique ID of the file checked in (1,2,3...). First file checked in will be named "aaaaaaaa". Second "baaaaaaa" and so on. There is no metadata description whatsoever in VSS. It's all hard-coded into the client code that access the files in the repository directly.

The database format used in MS Access on the other hand is, while still requiring "clients" to have filesystem access to it, a single file (.mdb) containing not only the data but also among other things metadata about the tables in it.

Having extensive (low-level, on-disk) experience with VSS from the last version OneTree released, up to Microsoft's 2005 release, I'd be happy to answer any questions that could put dubts to rest. The reason I didn't simply remove that erroneous comment myself was exactly this - there seems to be dubt about the correctness.

87.227.18.143 (talk) 03:52, 27 August 2010 (UTC)

I've removed some of the garbage from that paragraph. I'd go further (a lot further), but I know that wikipedia is based on popular consensus (aka urban myth), so there doesn't seem to be much point. We thought VSS was a great step up from PVCS, and I don't see anyone comparing it to VCS either. It was a good product in its time (1994-1999), and most of the people who repeat complaints about it's later history never worked with a version control system that was so complex it had it's own dba. I'd take most of the corruption stuff out because it's fundamentally misleading (its not "corruption" when the user destroys the data: it's bad design), except that somewhere out there, there must be at least one VSS user who was bitten by the data destroying SMB (server message protocol) bug in pre-SP1 Windows 2K Server 203.206.162.148 (talk) 08:25, 25 November 2010 (UTC).

External links modified

Hello fellow Wikipedians,

I have just added archive links to one external link on Microsoft Visual SourceSafe. Please take a moment to review my edit. If necessary, add {{cbignore}} after the link to keep me from modifying it. Alternatively, you can add {{nobots|deny=InternetArchiveBot}} to keep me off the page altogether. I made the following changes:

• Added archive https://web.archive.org/20070103081938/http://msevents.microsoft.com:80/cui/eventdetail.aspx?eventid=1032254284&culture=en-us to http://msevents.microsoft.com/cui/eventdetail.aspx?eventid=1032254284&culture=en-us

When you have finished reviewing my changes, please set the checked parameter below to true to let others know.

Y An editor has reviewed this edit and fixed any errors that were found.

• If you have discovered URLs which were erroneously considered dead by the bot, you can report them with this tool.
• If you found an error with any archives or the URLs themselves, you can fix them with this tool.

Cheers.—^{cyberbot II}_{Talk to my owner:Online} 14:12, 14 February 2016 (UTC)

Checked. The archived version is no good. The essence of the link is a streaming video that hasn't got archived. The link is dead for good. Best regards, Codename Lisa (talk) 17:23, 14 February 2016 (UTC)

External links modified (January 2018)

Hello fellow Wikipedians,

I have just modified one external link on Microsoft Visual SourceSafe. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:

• Added archive https://web.archive.org/web/20080906173447/http://blogs.msdn.com/checkitout/default.aspx to http://blogs.msdn.com/checkitout/default.aspx

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

This message was posted before February 2018. After February 2018, "External links modified" talk page sections are no longer generated or monitored by InternetArchiveBot. No special action is required regarding these talk page notices, other than regular verification using the archive tool instructions below. Editors have permission to delete these "External links modified" talk page sections if they want to de-clutter talk pages, but see the RfC before doing mass systematic removals. This message is updated dynamically through the template {{source check}} (last update: 5 June 2024).

• If you have discovered URLs which were erroneously considered dead by the bot, you can report them with this tool.
• If you found an error with any archives or the URLs themselves, you can fix them with this tool.

Cheers.—InternetArchiveBot (Report bug) 11:33, 29 January 2018 (UTC)

Version 6.0c Release Date

Hello. I recently came into possession of a boxed copy of Visual Studio .NET 2002 Enterprise Developer Edition. Inside the box was a CD-ROM for Visual SourceSafe 6.0c. The CD is copyrighted 1995-2001, but that is as close to a confirmed release date that I can get from the CD's contents. The readme (Readme.htm) at the root of the CD is written for 6.0c and has a copyright year of 2001 and a last-modified date of 2001-09-17. Most of the files in VSS/Win32/ss.exe have last-modified dates of 2001-09-11, but some (ssgui.dll and ssscc.dll) have a last-modified date of 2001-12-17. In any case, I haven't seen anything on the CD from 2002 or later. Since there's multiple dates to go by, and none of them are an official release date, I just put the year of 2001 in the release table. Hopefully my tidbit here can help someone who has better information. Huntertur (talk) 05:08, 9 February 2021 (UTC)