[scilab-Users] SOS: Version with Xcos usable for academic teaching??

[Nick Rudnick]

Hello Sylvestre,

the moment seems to come where I have to look into the sources -- do you 
have other clues besides the version of HDF5?

Sylvestre Ledru wrote:
>> If I have a choice, I would prefer to assist in fixing bugs at the
>> Linux builds, which still suffer from problems that prohibit usage --
>> I inspected mostly the 64 bit version, but I expect the 32 bit version
>> to be equally affected, as the error message seemingly is thrown by
>> the Java frontend.
>>     
> We are currently focusing the development version. That means we are not
> testing a lot the binary. We will starting doing so in the next few days
> for the first beta of Scilab 5.3.0. 
>   
;-)) Of course, I would not assist on basis of the binaries... but I am 
not a build system expert and I need binaries to distribute to our 
students, so I would be happy to have the automatic build process 
activated in time after I have done a fix...  That's all.
>> Specifically, at scilab-master-1276266589.bin.linux-x86_64.tar.gz, I
>> would like to present the error message I receive -- as I suspect a
>> java.lang.UnsatisfiedLinkError to be possibly a problem of access to a
>> native library which might be very easy to solve. It is thrown
>> whenever I try to drag something out of the palette browser (didn't
>> find this at Bugzilla):
>>     
> Thanks for the bug report. 
> This might be related to an issue with hdf5 libraries.
>   
Unfortunately, I have version 1.8.4, as required, installed:
    nick at weg:~$ strings /usr/lib/libhdf5.so | grep "version: "
    HDF5 library version: %d.%d.%d
    HDF5 library version: 1.8.4
    HDF5 library version: 1.8.4
>> In regard of a tutorial, I am in fact writing at the moment something
>> which might be regarded as one, 20-30 pages, with economically
>> motivated examples which should be easy to integrate into lessons,
>> consuming about 2-6 hours of lecture, depending on the students.
>>     
> OK, great. Let me know how it goes!
>   
I am finishing it the next day, and translate it then, which -- thanks 
Google Translate -- should go quickly.
>> Once I can forward working binaries for Linux, too, I will be very
>> satisfied, and could assist the Tk to JGraphX transition -- here I
>> first would like to improve the branching of connections.
>>     
> I am not sure to understand what you are talking about here ?!
> AFAIK, there is no Tk into Xcos now.
>   
Excuse me, this was unclear: I wanted to address the branching, i.e. 
improving the reliability of marking a connection light green right to 
be able to drag a branch. I guessed this to be one of the problems 
people usually have when migrating from Tk, as Tk has features in 
graphics object interaction which are missing in most other GUI Toolkits 
and have to be implemented manually -- in this special case I thought I 
identified a line object selection problem.
>> Seemingly, some (O)Caml code is used inside Scilab... very interesting
>> stuff... :-)
>>     
> Indeed, we use it for the Modelica capabilities in Xcos.
>
> By the way, if you don't mind, we could switch to dev at lists.scilab.org
> (or private if you prefer) to chat about contribution aspects.
>   
I sent a request to dev at lists.scilab.org.

All the best,

    Nick
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.scilab.org/pipermail/users/attachments/20100622/b1690878/attachment.htm>