Re: Modules 3.1.6

[Roberto B. <Rob...@ge...>]

At 07:24 PM 10-1-2002, you wrote:
>         * A 'break' in a modulefile will cause the modulefile to
>           not be added to $LOADEDMODULES, but will be considered
>           a successful execution to not affect other modulefiles
>           being loaded concurrently (Patch due to Scott Gaskins).
>         * Tests and docs added for the above.

Hi,
         i just downloaded and installed the latest release of modules but 
when i compile it and run the make check command it fails on all tests 
testing this new feature! I compiled modules on a Sun workstation with 
Solaris 8 and Tcl 8.0.
The strange thing is that (i am not an expert of this test method) manually 
load the break (1.0, 2.0 and 3.0) modules by hand modules complains on 
screen that it could not load module break/2.0 but modules 1.0 and 3.0 are 
correctly loaded and also the environment variables are set. Therefore i 
suppose that this may be a "bug" in the test routine...
The test fails for all shells!
The only other test that fails is the module whatis test but i did not 
investigate it further...

Another problem i see is the following:
If i load a module file with its complete path (i.e. 
/opt/Modules/Modulefiles/Test/3.0) with the command
module load /opt/Modules/Modulefiles/Test/3.0
if i type module list i get the module file correctly displayed ( Test/3.0 
) but in order to remove it i have to issue the command:
module rm Test/3.0
If i instead use the following: module rm /opt/Modules/Modulefiles/Test/3.0 
modules does not complain but it does not even remove the module.

This may not be a bug but i think that a warning message should be issued 
in such cases.

Best regards,
                 Roberto Beccherle