Jump to content

Talk:Wolfram Mathematica/Archive 5

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

Why is there no discussion of criticism whatsoever?

[edit]

Virtually any subject of a Wikipedia article for which there exist any published opinions will have some mention of criticism of that subject in the article.

Why is there nothing whatsoever regarding criticism of Mathematica in the article?

My guess: Wolfram has a team of employees whose job description is to remove any such discussion of criticism as soon as possible. (But this is only a guess.)

This is a serious omission in the article.Daqu (talk) 03:29, 7 February 2015 (UTC)[reply]

Agreed. This is a serious omission; and a suspicious one given the number of serious bugs involving basic features (e.g. with selection) and the lack of support for system-standard integration of the notebook (e.g., lack of usable clipboard support, poor support for printing, Finder/Spotlight integration on OS X), poor and unresponsive product support, and the painfully slow pace of bug repair. This article feels like a sanitize product promotion. AldaronT/C 17:35, 25 December 2015 (UTC)[reply]
I don't see criticism sections or lists of bugs on pages for similar software. The advert tag is for active advertising content, not the lack of negative content. Perhaps the template you were looking for was NPOV? If you think the tag is correct, please give examples of other software pages where it is used. Your point seems to be a rehash of discussion from several years ago without new arguments. --Pleasantville (talk) 18:06, 31 December 2015 (UTC)[reply]
"I don't see criticism sections or lists of bugs on pages for similar software." But other software doesn't have the same combination of expense and lack of basic functionality (e.g. it in my case, it is both the most expensive software I own, and the only software that doesn't support normal clipboard behavior). AldaronT/C 20:04, 31 December 2015 (UTC)[reply]
Years ago I made a similar argument about GoDaddy and was in essence told that individual grievances with a company or product are not considered encyclopedic on Wikipedia.Pleasantville (talk) 19:05, 1 January 2016 (UTC)[reply]
When such "individual grievances" accumulate and become serious usability issues, they're worth mentioning. AldaronT/C 21:55, 1 January 2016 (UTC)[reply]
For those new to this ritual. User Alderon regularly posts (usually) unsubstantiated claims that each upgrade makes Mathematica unusable. At first glance the cited bug here appears to be at least partly incorrect (copy paste of non-math text on OSX to Word or TextPad both behave as expected). He has also asked for his complaints (how it's too expensive, and how awful the tech support is etc) and feature requests to be added to this article several times before, which have been argued out at length. Since we have been here before, I am going to remove the tag until he can gain some consensus for adding bugs lists etc to the article, rather than leave the tag there until it is clear that there wasn't consensus, as in the past. Briefly my comments: 1) other software articles do not contain bugs lists (go look at a few). 2) You know wikipedia well and know that anything contentious is going to need a decent citation, and linking to your own questions on stackexchange isn't going to meet that standard. JonMcLoone (talk) 18:21, 5 January 2016 (UTC)[reply]
Are you disputing the accuracy of the behavior described in the linked question? If not I'm going to reject point (2). Point (1) can be easily addressed by simply listing such behaviors as something more neutral like "non-standard" behaviors (which are freely discussed in other software articles). And in any case, neither of these points directly address the relevance of the template (which I'm restoring): this article is written, maintained, and heavily monitored by Wolfram employees like you. That's not what Wikipedia is about; and certainly something readers of the page should be made aware of. AldaronT/C 22:05, 5 January 2016 (UTC)[reply]
Linked question suggests that copying "abcde" corrupts the text to appear on multiple lines. I copied that text from Mathematica and pasted into TextPad and saw "abcde". That was as far as I investigated. The claim "which is entirely useless almost everywhere." appeared to be innacurate. I just clicked through the first 50 software articles listed at the bottom of the page (all the Computer Algebra and Numerical links) and only two have any criticism sections OpenFOAM which has an advantages/disadvantages section which is mostly subjective and LabVIEW where almost every item is currently being challenged with Citation Needed tags. 48 articles do not have any criticism sections- this is not a normal feature of a WP article and not a requirement to avoid the Advertizing tag. Look at the policy that you have linked to and you will see "Information about companies and products must be written in an objective and unbiased style, free of puffery. All article topics must be verifiable with independent, third-party sources". Highlight the sections you think contain puffery or bias so we can discuss them, or where citations are needed, and we can look for them. As for my involvment, lets not talk in inferences - cite the edits you disagree with so we discuss them. JonMcLoone (talk) 09:53, 6 January 2016 (UTC)[reply]

OK, since you have switched your justification for the Advertizing tag from the need to list bugs to the fact that I edit the page, let me help wrap this up quickly by categorizing my recent edits for you. Since the start of 2014 there have been 105 edits to the article. I have made 16 of those: 3 formatting and text edits [1] [2] [3], 5 version number maintenance [4] [5] [6] [7] [8], 3 facts changed by the release of new versions [9] [10] (reverted by you) [11], 1 platform addition [12], 2 tag removal (see Talk) [13] (reverted by you) [14], 2 Addition of third party Hadoop link to the "Links to ther applications section" [15] [16] I think the only ones that could conceivable be categorized as advertizing are the links to the Hadoop Link which is not a Wolfram product and to which I have no direct connection. Unless you have further justification of the Advertizing tag (in line with the policy it links to) I will make a 17th edit and remove it. JonMcLoone (talk) 13:41, 8 January 2016 (UTC)[reply]

I can't match your commitment to this issue (I'm not paid to care about it). Change it back if you wish; but it won't stay that way. My concern is only making clear to potential Mathematica users the kind of trouble they may be getting into (at considerable expense). (And again, your experiences with TexPad doesn't address the question: have you examined the formats Mathematica puts on the OS X clipboard?) AldaronT/C 16:40, 8 January 2016 (UTC)[reply]
I am motivated to defend myself against accusations. I am not paid to edit this page, it is not in my job description, no pay is attached to editing, I do not report my edits to my manager, I make them in my own name. But as a Wolfram employee I am an expert on the subject. And as a Wolfram employee have an interest in keeping it accurate and not allow it to become a forum for your hostility towards us. Perhaps I am also still sore after last time, when you wanted to add two non-bugs to the page, and you demanded my managers name to complain about me.
I think you confuse issues with your hyperbolic statements including the "entirely useless almost everywhere" when it sounds like you mean "is hard to use from AppleScript". I am sure others did what I did first and copy some simple text to TextPad, Firefox, Word and think "I can't see the problem". The internal clipboard format is not "the kind of trouble" most users get into. The copying of math content, while generally a good thing could be done better over math characters. Your other issues - seeing the print layout on screen is done using the menu Format/ScreenStyleEnviroment/Printout, and bad indexing in Spotlight (which probably is a Spotlight bug not a Mathematica one), is fixed in terminal with sudo mdutil -E /
The selection problem, you fixed yourself by deleting the cache and has no reproducible characteristics. So out of 6 bugs you want to put in a "bugs" section about a half of one is likely to impact the average user.
The advert tag may return, and I am happy to discuss specific points which may turn out to be valid. But the article is currently, I think, very neutral, and I wouldn't want it to become like the LabVIEW one I linked above which swings between unjustified cheerleading and unjustified complaining.

JonMcLoone (talk) 17:49, 8 January 2016 (UTC)[reply]

The article is far, far from neutral. It is an extensive list of features, many of which are trivial. There is no comparison with non-Wolfram alternatives. A substantial fraction of the sources are affiliated with the product. The tone is promotional, for example there is no mention of short-comings or limitations of the product. The list of "related products" exclusively lists wolfram products. There is a history of every minor version increment. There is no reference to any critique, such as the review which was published (by Fateman) in the Journal of Symbolic Computation.
JonMcLoone, try to imagine if employees of BigHomeopathy were to write the wiki article, and while trying not to use obviously biased language they produced an article consisting solely of an extensive list of every advantage of their product as claimed by their organisation: that wouldn't seem encyclopedic would it? In fact, it would be difficult to distinguish from an advertisement.. Cesiumfrog (talk) 12:00, 9 January 2016 (UTC)[reply]

Let me comment on that point by point...

  • "It is an extensive list of features, many of which are trivial."

Yes. Isn't that the first aim of an encyclopedic article? To answer the question "What is it". Presumably the article on "bicycle" says "two wheels" wheels may be a trivial bit of technology but that is what a bike is.

  • "There is no comparison with non-Wolfram alternatives." I do think this is very hard to do over such a large set of only partially intersection features over many product, as I have commented in the past (see talk archive). The article links to about 100 alternatives, their pages are there to describe what they do.Equally, I don't think we should be adding a "how X compares to Mathematica" to those 100 pages.
  • "A substantial fraction of the sources are affiliated with the product." Yes, this is a problem, I can try and find more citations, if that is an acceptable contribution from me. In many cases references to the Documentation are the clearest description of the scope, or lack of, for a feature. But others would be better to come from other places. But a quick scan suggests the worst section for this is the "related products" see my comments on that below. Also the "Design" section appears to exist purely to host an unhelpful documentation link. I would be happy to remove that whole section, that adds little to the article - if you support such a deletion.
  • " The tone is promotional, for example there is no mention of short-comings or limitations of the product." Please give an example of the promotional tone. Or is your point that the thing is inherantly promotional unless it makes negative comments? If so see my point above about 48 out of 50 checked articles being the same. Are they also overly promotional?
  • "The list of "related products" exclusively lists wolfram products." I agree, I don't think that belongs in the body of the article. This is a Mathematica article, not an article about Wolfram products in general. Some pages for companies have a collabsible footnote template for such information that keeps it off the body of the article. I would be happy to code that up, if you would support such a change.
  • "There is a history of every minor version increment." I agree. I argued previously in the talk pages that maintenance releases was too much detail. I have periodically deleted older 3rd digit relases in line with the discussion that happened then (see talk archive). I was going to wait until the next release, but since you bring it up, I will delete the 8.0.x releases now.
  • "There is no reference to any critique, such as the review which was published (by Fateman) in the Journal of Symbolic Computation." That is article is from 1992 about version 1.2. I can't imagine many software articles cite comments about 20+ versions previous, and 24 years ago. Maybe there are more relevant critiques?
  • "JonMcLoone, try to imagine if employees of BigHomeopathy were to write the wiki article, and while trying not to use obviously biased language they produced an article consisting solely of an extensive list of every advantage of their product as claimed by their organisation: that wouldn't seem encyclopedic would it? In fact, it would be difficult to distinguish from an advertisement." - there isn't any statement of "advantages" in the article only lists of features. Thats like a homeopathy article that says "contains 99.99% water, 0.001% mint". So the analogy you seem to be drawing is the way a BigHomeopathy article might claim that a disease can be cured when it has no such effect. Is there a feature described in this article that you think isn't true or implies significantly greater scope than is really supported? If so say which ones. Please give examples, because this whole conversation is hard without them.

JonMcLoone (talk) 21:22, 10 January 2016 (UTC)[reply]

I have added some of the various comparison pages which explore the capabilities of Mathematica/Wolfram Language as compared to similar systems to the See also section. --Pleasantville (talk) 16:37, 12 January 2016 (UTC)[reply]
The "comparison of computer algebra systems", and the "comparison of numerical analysis software" were already linked from the info-boxes further down the page. Though arguably those boxes should be in the "See Also section", not below "External Links" which would have made them more obvious JonMcLoone (talk) 18:43, 12 January 2016 (UTC)[reply]

"The internal clipboard format is not "the kind of trouble" most users get into." I think you misapprehend the issue. If I copy abcdefg from a notebook, I get

a\\
b\\
c\\
d\\
e\\
f\\
g

and if I copy αβγδϵ I get

α\
β\
γ\
δ\
ϵ

This is what Mathematica puts on the clipboard and labels plain text. The results are much worse for anything of any complexity. I'm pretty sure this is not what anyone expects or finds useful, and it's certainly doesn't contribute to a practice workflow. Its a bug, or at best an extreme oddity. Where should such things be mentioned?. Isn't that something that should be discussed in the article. AldaronT/C 14:07, 13 March 2016 (UTC)[reply]

Mathematica 10.1

[edit]

New features at http://reference.wolfram.com/language/guide/SummaryOfNewFeaturesIn101.html Summary at http://www.wolfram.com/mathematica/quick-revision-history.html JonMcLoone (talk) 18:25, 30 March 2015 (UTC)[reply]

Mathematica 10.2

[edit]

The full list of new features is at http://reference.wolfram.com/language/guide/SummaryOfNewFeaturesIn102.html I will leave it to others to decide which are notable. JonMcLoone (talk) 13:41, 14 July 2015 (UTC)[reply]

Mathematica 10.3

[edit]

New features listed at http://reference.wolfram.com/language/guide/SummaryOfNewFeaturesIn103.html in case anyone wants to update the summary. — Preceding unsigned comment added by JonMcLoone (talkcontribs) 07:52, 15 October 2015 (UTC)[reply]

Use of JAVA SE 6

[edit]

How long is this statement noteworthy: "Mathematica prior to version 10 for OS X required Java SE 6 which is a deprecated component of Mavericks."? Mathematica 9 is now 5 versions behind current and the deprecation happened in OSX Mavericks which is 2 versions beind current OSX. So this only affects users who do not upgrade their Mathematica and have not upgraded their OSX for 2 years, now choosing to upgrade OSX only and having to install an extra component. JonMcLoone (talk) 13:49, 8 January 2016 (UTC)[reply]

No one has commented in 2 weeks, so I am going to make this change.JonMcLoone (talk) 18:31, 22 January 2016 (UTC)[reply]

Abandoned platform support

[edit]

This also seems to be a rather dull historical fact "Mathematica up to 6.0.3 supported other operating systems, including Solaris, AIX, Convex, HP-UX, IRIX, MS-DOS, NeXTSTEP, OS/2, Ultrix and Windows Me.[70]" since none of those OSs are in any significant use now. JonMcLoone (talk) 14:02, 8 January 2016 (UTC)[reply]

No one has commented in 2 weeks. So I am going to make this change too. JonMcLoone (talk) 18:31, 22 January 2016 (UTC)[reply]
[edit]

Just to highlight two changes I plan to make that received no comment in the long discussion above.

I plan to remove the section on Design which serves only to host a link to a not-very-informative part of the documentation of Mathematica.

I plan to remove the Related Products, which serves only to host a list of Wolfram products, some of which are already linked in the article, and replace it with an info box at the foot of the article. A draft of this is [here] though it should appear collapsed in the final destination. JonMcLoone (talk) 18:42, 22 January 2016 (UTC)[reply]

No one has commented in two weeks so I will make this change JonMcLoone (talk) 09:11, 5 February 2016 (UTC)[reply]