MathGroup Archive 2005

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

Search the Archive

Re: VersionHistory[ commandName] would be nice (was Mathematica 5.2: The 64-bit and...)

Peltio wrote:

>"Pratik Desai" wrote
>>I think the help browser has such a feature
>>New in Version 5.1
>Yes, but I was talking about a command that would return, given a command
>name, at least a list of two numbers: one with the version of Mathematica
>where it was first introduced, and one with the version(s) of Mathematica
>that had its syntax or behavior modified or enhanced.
>Something like (I'm making this example up)
>    VersionHistory["N"]
>        {1.0, {2.0, 2.2, 3.0, 4.2, 5.1}}
>Which mean that the command N was there since the first version and was
>modified in subsequent versions. (More sophisticated information could be
>given - like a list of version where the syntax was modified, a list of
>version where the output was modified, the version where the command was
>superseded... - but this is the essential bit that's needed).
>*Once we have this*, it's easy to write procedures that, given a program o
>package written in a given version, will tell us what are the commands used
>that have been modified since the time of writing, with respect to the
>version we are running. It would be easier to upgrade one's packages, and it
>would be easier to identify the source of compatibility problems in packages
>that don't work on a different version of Mathematica.
>Sometimes the simple fact that N[0] gives 0 in a version and 0. in another
>one can prevent a package from functioning properly.
>Passing a package to the YetToBeBorn function
>    CompatibilityCheck[packageName_String, versA_->versB_]
>could spare hours of code re-reading by pointing out the causes of possible
>The compatibility table and VersionHistory command should be available as a
>package on the mathsource so that people with older versions could see why
>it is time for them to upgrade if they want to run the latest packages.
>It's a small step for WRI, yet a giant leap for mathkind. : )
I suspect the WRI developers already have something like this and they 
haven't made it public yet.

>Invalid address in reply-to. Crafy demunging required to mail me.

Pratik Desai
Graduate Student
Department of Mechanical Engineering
Phone: 410 455 8134

  • Prev by Date: Re: solve output "problem"
  • Next by Date: Re: Re: Setting gridlines thickness in Plot
  • Previous by thread: VersionHistory[ commandName] would be nice (was Mathematica 5.2: The 64-bit and...)
  • Next by thread: Re: Mathematica 5.2: The 64-bit and multicore release