Re: [Texmacs-dev] Interface for FriCAS

[Andrey G. Grozin]

On Sun, 27 Mar 2011, Alexander Solovets wrote:
> I used to write here about experimental FriCAS interface for TeXmacs
> and included plugin sources. Now, when new version of FriCAS has been
> released, I'm kindly asking you to include this plugin in the main
> TeXmacs repository in order to let users test it properly.
I second this suggestion. The plugin works with the recently released 
version, fricas-1.1.2 (and will work with future versions, of course).

While we are at it, I'd propose to clean up some old cruft in plugins.

If we remove support of paleolithic versions of maxima, the plugin will 
become much simpler (neolithic versions will be still supported). There 
will be no maxima_filter, and the scripts will become shorter and easier 
to maintain. There is a more practical reason for this, too: some time 
ago, it was reported that on MacOS, when maxima is installed from some 
standard repository, the maxima script does not contain the string 
'list-avail', and thus the current script classifies it as a paleolithic 
version. And the interface does not work.

MuPAD is dead for quite some time. And I doubt that the mupad plugin which 
is currently distributed works with the latest MuPAD which existed. It 
worked with some older version. Maybe, it's better just to remove this 
plugin? If not, we should fix a "security" problem in it: the script sets 
LD_LIBRARY_PATH in a wrong way, and if the user had an empty 
LD_LIBRARY_PATH before calling this script, there is a (theoretical) 
security risk. I'm sure nobody uses the mupad plugin, and the risk is 
purely theoretical. But if we want to keep it, it needs a 1-line patch 
which fixes the problem. I know this line.

I am surprised people still manage to use the mathematica plugin. Much has 
changed in mathematica, and now using it requires a lot of extra effort. I 
suppose it should be fixed, but I don't have mathematica on Linux, and I 
cannot do this. If some users still manage to run it, let it stay here.

Andrey