[Scilab-Dev] Re: and idea for the development of Scipad

[fvogelnew1 at free.fr]

Selon Torbjørn Pettersen <torbjorn.pettersen at broadpark.no>:

> Ok - enclosed is a new CECILL version of help_from_sci.

Very good.
I'll try it.

> I've included a Scilab-version-5 check, which returns an error on older
> versions of Scilab.

Why not, but I think it's useless though it doesn't harm.

However, such a check needs anyway to be included in Scipad itself in order to
prevent launching help_from_sci in a Scilab-4.x or Scilab-gtk environment (where
it is not included and it's pointless to include it).

Enrico and myself try to take care of keeping Scipad as backportable as
possible, this has proven to be very useful in the past (many years long
development cycle before reaching Scilab 5, therefore we released backported
versions, also to let users use the debugger - still broken in Scilab 5, and a
few other reasons).

I'll add this version check at the Scipad level. Sorry if you misunderstood what
I wrote in the SEP about version checking, this was mostly a reminder for
myself.


> The current version contains the option I mentioned in a previous mail
> with respect to generation of demo files based on the code from the
> Examples
> section in the help file. It seemed like a natural extension of the
> help_from_sci idea.
> Anyway I don't think it should interfere with your scipad implementation.

I don't think so either, but I will check in practice.

> A remaining point to do is to include a check for the existence of the
> target .xml or .dem.sce files.
> At this point help_from_sci will happily overwrite any existing target
> files, which is rather nasty...

No, not when called from Scipad. This is already handled (user is asked for
overwrite or open the existing target file).
If you want to add such a feature in your code, for the case it gets called from
Scilab directly, please feel free to do it (slow time approach, no rush).

Thanks again for your contribution.

Francois