[scilab-Users] Mac OS 10.5 Compilation issue

[Eric E]

Hi Sylvestre,
  Here's the Makefile...  Dunno if this would work, but it occurred to
me that I could try using an alias in .bashrc to alias install -D to
install -d or some such.  I'll let you know if that works.

I also ran scilab with and with the -nw flag.  When I ran it without -nw, I got:
Localization: setlocale didn't exist on the computer used to compile
Scilab ! This is abnormal ! No localization will be working for this
distribution of Scilab.
and that is indeed the error that appears in all the error files from
make check.  Running scilab normally (i.e., without the -nw flag)
brings up the standard Scilab command screen, which appears to work
but won't accept any input from the keyboard.  Setlocale is present on
my machine (or at least configure thinks so:
checking for setlocale... yes
Some initial googling about setlocale indicates that there may be some
weirdness about Apple's X11 implementation and setlocale.

When I run scilab with the -nw flag, the command window doesn't
appear.  Am I correct or incorrect to guess that -nw is like -nw for
Emacs: don't use X windows or something like that?  In any event, with
the -nw flag, issuing the plot command causes a window to appear and
then the following exception appears:

terminate called after throwing an instance of
'GiwsException::JniCallMethodException'
  what():  Exception when calling Java method : no jogl in java.library.path
 at java.lang.ClassLoader.loadLibrary(ClassLoader.java:1753)
....

and the app crashes shortly thereafter.  I do have jogl in the same
directory as all the other jars.

Is any of this making sense to you, or pointing to a common problem?

Thanks again,

Eric


On Mon, Jan 5, 2009 at 4:52 AM, Sylvestre Ledru
<sylvestre.ledru at scilab.org> wrote:
>
>
> Le dimanche 04 janvier 2009 à 18:19 -0500, Eric E a écrit :
>> > Well, depends what you mean by "tests of the build", if you want to
>> > check the creation of the binary itself, "make macros" is calling Scilab
>> > to build our macros.
>> > If you want to test Scilab features, "make check" is your friend.
>
>> 315/1152 - [fileio]
>> dir..........................................failed  : error_output
>> not empty
>>
>> Is this standard, or should I look into the outputs of make check?
> The output are probably just warnings.
> Try to run Scilab like ./bin/scilab -nw. You should not get any output;
> if you do, it is what is causing the "error_output not empty".
>
>> > In the root directory of the source tree, what is the result of
>> > $ grep -E "(^INSTALL =|^INSTALL_DATA =)" Makefile
>> > ?
>> >Under Linux, I get:
>> >INSTALL = /usr/bin/install -c
>> >INSTALL_DATA = ${INSTALL} -m 644
>>
>> I get the same thing as you do:
>> INSTALL = /usr/bin/install -c
>> INSTALL_DATA = ${INSTALL} -m 644
> Funny, could you send the modules/core/Makefile file ?
>
> Thx.
> Sylvestre
> PS: if you are registered to the dev mailing list, we could switch on
> this mailing list since it is quite technical questions.
>
>
>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Makefile
Type: application/octet-stream
Size: 232403 bytes
Desc: not available
URL: <https://lists.scilab.org/pipermail/dev/attachments/20090105/7d809b9f/attachment.obj>