Scicos bug

[ga]

Hi,

I'm testing Scilab 5.0.3 with Scicos 4.2 under Windows Vista o.s. with
a graphics board Intel 965.

The Scicos window appear as 'truncated' for all kind of Scicos files,
it is simply not possible to work.

In practice the blocks are visible only for a subset of the window
(top left corner) and disappear in the remaining part of it. I wil try
to attach an image.

any suggestion?

Best,
g.

[Delebecque Francois]

ga a écrit :

Have you tried Scicos 4.3 included in ScicosLab ?
http://www.scicoslab.org

FD

[ycollet]

You can also try the 5.1 nightly build for windows:

http://www.scilab.org/download/index_download.php?page=nightly_builds_5.1

YC

[ga]

thanks to both of you for the replies.

I'm a little bit disoriented by the recent born of different versions
of Scilab, is Digiteo going to keep the open source philosophy of
Scilab ? What about the development of Scicoslab? will it be somehow
related to Scilab?

Best,
g.

[ycollet]

As far as I know, Digiteo will leave Scilab opensource (this was the
main goal of the change of the licence).

For Scicoslab, I am a little bit disoriented too. Scicoslab is based
on Opensource code like Lapack, BLAS, GTK, etc...
And because GTK is GPL, some access to the sources of scicoslab (or
the gtk part) must be provided. For Lapack and Blas, this is not the
same thing because of the licences which are more permissive for the
binray distribution.

Can somebody give some explanations related to the source code of
scicoslab ?

YC

[ycollet]

If you look at the following web page: http://www.scicos.org/history.html
You can read the following statement: "Scicos adopted the GPL license
in 2008. A new GPL Modelica compiler is being developed by S. Furic
and R. Nikoukhah."

And the GPL forces the distribution of source code.

For Gtk, it's LGPL which is a licence relatively permissive wrt the
distribution of source code. Sorry, my mistake.

YC

[ga]

Thanks.

I'm still disoriented.

I started using Scilab for my classes 2 years ago. I have written a
manual in Italian (downloaded approx 4000 times) with the idea to help
my students in moving from Matlab to an open source version of it.

Scilab 4.1 is a nice software that could be strongly improved from the
user-friendly aspect (if someone wants it to gain popularity) and I
have installed it in the software classrooms.

I expected Scilab 5 for this improvement but:

- on my Linux laptop it does not run due to a problem with the graphic
cards;

- on my Vista laptop the Scicos window is bugged and can not be used
(the original reason for this thread);

- on XP Scicos does not open anymore the paltree window;

- I tried to install Scicoslab at least on my laptop but gave up when
apt-get asked for dozen of upgrade of packages installed (yes, I
downloaded the correct version);

and this list is incomplete.

In order to be widely used Scilab (Scicoslab) needs to be stable, if I
move to the new version I should not find a new bug (e.g. the paltree
problem), in few words, I need to know where the developers want to go
and I need to trust the code. I'm still with the Scilab 4 version
installed and still not sure if and what version I will install
further.

Best,
Gianluca

[ycollet]

Scilab 5 represents a huge work:
- modularization,
- java interface,
- localisation,
- 64 bits,
- etc ...

I think it was released a little bit soon, but, for me, it is a good
thing. You have the possibility to try this version and fill bug
reports.
Now, wait for scilab-5.1.

I can make a parallel with KDE4. KDE-4.0 was some kind of beta. But
because people don't use too much betas, the released 4.0 so as to
meet a lot of users and have a lot of bug reports. 4.1 is a lot more
useable, and 4.2 will be really fine.

You still have the possibility to use 4.1.2 for a while.

YC

[ga]

Thanks again for your reply.

First of all I'm not underestimating the huge work made from the
version 4 to the version 5, sorry if my former message was not clear
enough.

Still I have a rude question:

Scilab and Scicoslab will be two completely independent software? As
an example, will it be some kind of coordination between the syntaxes
of the future developments of the same commands?

Does it mean that the users of Scilab 4.1 now need to make a definite
selection between those two branches of it?

Best,
Gianluca

[ycollet]

> Scilab and Scicoslab will be two completely independent software?  As
> an example, will it be some kind of coordination between the syntaxes
> of the future developments of the same commands?

As far as I know, Scicoslab and Scilab will be 2 independant
softwares. But maybe, in a not so far futur, we can hope for a merge
of these separate projects (let's hope together :-)).

>
> Does it mean that the users of Scilab 4.1 now need to make a definite
> selection between those two branches of it?

Yes again. For example, the scicos version shipped in scilab is older
than the one shipped in scicoslab.

YC