MathGroup Archive 2009

[Date Index] [Thread Index] [Author Index]

Search the Archive

Re: Maintaining a Mathematica bug list

  • To: mathgroup at smc.vnet.net
  • Subject: [mg97390] Re: [mg97349] Maintaining a Mathematica bug list
  • From: Murray Eisenberg <murray at math.umass.edu>
  • Date: Thu, 12 Mar 2009 02:20:07 -0500 (EST)
  • Organization: Mathematics & Statistics, Univ. of Mass./Amherst
  • References: <200903110925.EAA01644@smc.vnet.net>
  • Reply-to: murray at math.umass.edu

There's some trouble with the whole idea of a user-maintained "bug list" 
for such a proprietary product: one person's bug is another person's 
feature; yet another person's bug is a failure to understand the 
documentation or the core language.

So such a bug list could be counterproductive unless a real expert 
served as referee.

Maris Ozols wrote:
> Hello,
> 
> Note: I moved the discussion on maintaining a Mathematica bug list to
> a new thread.
> 
> While the number of Mathematica functions is growing exponentially,
> the satisfaction among the users (especially those, who have been
> using Mathematica for years) does not seem to increase too rapidly.
> The old-fashioned way Mathematica users have to submit their bug
> reports is incompatible with the futuristic Wolfram Alpha project.
> Also the claim that Mathematica "delivers unprecedented reliability"
> does not fit with the words such as "improvement" and "enhancement"
> being used to describe the recent changes, instead of calling them
> "bug-fixes" and documenting in a proper way. Such descriptions are
> appropriate only for advertising a new version of laundry detergent,
> not a programming language.
> 
> Honestly, I think that maintaining our own list of bugs would be an
> act of desperation. I am not sure if that would have any impact on
> WRI, but it would definitely make the life much easier for Mathematica
> users. I do not have enough experience to set up such a thing (e.g.,
> bug tracking system, Wiki, database, etc.). However, I believe that
> there are users on this forum who have such an experience.
> 
> I would like to encourage Mathematica users to express their thoughts
> regarding whether such a step is really necessary. I am certain that
> WRI have enough resources to allocate for improving the communication
> with the Mathematica user community. I think that it is just a
> question of recognizing the problem and willing to solve it.
> Mathematica is really a great tool and we all are willing to
> contribute to make it even more better!
> 
>> Do you know of something like this that exists
>> already for any other software product?
> 
> Surely, most of the open source projects nowadays have their own bug
> tracking systems. However, it is not very common for users of
> proprietary software to maintain their own bug lists. Here is one
> example that I found on the web:
> 
>   The Delphi Bug List
>   http://buglist.jrsoftware.org/
> 
> If you read this page, you can see that Delphi users had a very
> similar experience as we do. It is not clear though whether they
> succeeded or not. On their "What's New" page
> 
>   http://buglist.jrsoftware.org/whatsnew.htm
> 
> you can see that their bug list is no longer maintained, since Borland
> has made available their own bug tracking system "QualityCentral".
> However, the supplied link
> 
>   http://qc.borland.com/
> 
> does not respond. Another example of a user-maintained bug list is
> 
>   ColdFusion Bug Hunt
>   http://www.cfbughunt.org/
> 
> There are more examples of companies that actively encourage feedback
> from the user community. For example, Adobe, see
> 
>   Adobe Public Bug Database and Issue Reporting System
>   http://bugs.adobe.com/
> 
> and
> 
>   Adobe Labs
>   http://labs.adobe.com/about/
> 
> Even Microsoft welcomes user feedback:
> 
>   Microsoft Connect
>   https://connect.microsoft.com/
> 
> ~Maris Ozols~
> 
> P.S. Here are the two most ridiculous Mathematica bugs that I know of.
> 
> 1. If you are using Mathematica 6, try to evaluate this number:
> 
> -1/Sqrt[(5-2Sqrt[2]-Sqrt[5-2Sqrt[2]])(5+2Sqrt[2]-Sqrt[5+2Sqrt[2]])/2]
> 
> Actually WRI guys are not that evil, since this is fixed in 7.0 (I
> reported it on March 2007, though). I was very pleased when I found
> that out. Thanks!
> 
> By the way, I don't know if that happened by an accident or not, but
> their reply to my bug report contained two interesting links:
> 
>> Hello,
>>
>> Thank you for the email.
>>
>> see
>>
>> https://bugs.wolfram.com/show?number=54568
>> and
>> https://bugs.wolfram.com/show?number=76559
> 
> Unfortunately, bugs.wolfram.com is not working, so I believe it is an
> internal system. So, WRI, please make it publicly available!
> Mathematica users would really appreciate that!
> 
> 2. Open the Documentation Center and press Ctrl+Enter or Ctrl+, in the
> search bar at the top. I reported this for 6.0.3. Since 7.0 was
> released very soon, probably they didn't have time to fix this for
> 7.0. So, I hope it has been fixed for 7.0.1, since it should be really
> trivial to do this.
> 
>> One thing that Mathematica users could do is to set up a Wiki where
>> they maintain their own list of known Mathematica bugs. Of course,
>> this is really what WRI should do, not us. But, here are a few reasons
>> why it might be an advantage for us to do so:
>>
>> * In this way Mathematica users would declare their position regarding
>> the way Mathematica is being developed.
>>
>> * Individuals and organizations that are considering buying
>> Mathematica would see that it actually contains lots of bugs and those
>> bugs don't get fixed. That would discourage them from buying
>> Mathematica and make WRI to rethink their strategy.
>>
>> * Mathematica users could check if their bugs are already known
>> without wasting their time to submit the bug reports.
>>
>> * It would be easier to monitor which bugs get fixed and which don't.
>>
>> ~Maris Ozols~
> 
>> Hi. I think this is a wonderful idea and would be willing to volunteer
>> time to help out. Do you know of something like this that exists
>> already for any other software product? If anyone else thinks this
>> would be of help please let your comments be known so the level of
>> interest could be gauged. I too, agree that Wolfram should already
>> have something like this, and would love to be able to look somewhere
>> and see if the bug/problem I am experiencing has already been reported
>> rather than waste time reporting something already known. Where would
>> be a good place to locate something like this? GREAT IDEA!!!!
>>
>> -Bob
> 
>> Maris Ozols has made a practical suggestion, which SHOULD be discussed
>> further about the possibility to really implement it.
>>
>> ADL
> 

-- 
Murray Eisenberg                     murray at math.umass.edu
Mathematics & Statistics Dept.
Lederle Graduate Research Tower      phone 413 549-1020 (H)
University of Massachusetts                413 545-2859 (W)
710 North Pleasant Street            fax   413 545-1801
Amherst, MA 01003-9305


  • Prev by Date: Re: Re: Piechart labels Mathematica 7
  • Next by Date: Return Ambiguity
  • Previous by thread: Maintaining a Mathematica bug list
  • Next by thread: Re: Maintaining a Mathematica bug list