[globus-python] RE: [AG-TECH] Building failure - strange behaviour with pyGlobus

[Keith Jackson]

>> So I looked closer at the error, this globus-makefile-header.
>> I have found a very strange behaviour : when I change the
>> line 230 in /usr/lib/globus/bin/globus-makefile-header the
>> behaviour of the build process is deeply affected even if a
>> just add the word "TEST" in the output string. In brief, I
>> just change the line 230 from :
>> 	printf("No packages were found that matched your
>> query!\n"); to :
>> 	printf("TEST No packages were found that matched your
>> query!\n");

This is NOT a good thing to do. We're using a regular expression to  
match the error string that globus-makefile-header returns. You've now  
convinced pyGlobus's setup.py that you have working globus packages,  
which you clearly do not. You should run globus-makefile-header by hand  
and see what output you get. Until you get valid output from  
globus-makefile-header pyGlobus will not build. This isn't a problem  
with pyGlobus, this is a problem with your globus installation.
--keith


On Sep 3, 2004, at 4:12 AM, Ivan R. Judson wrote:

>
> Hi Joseph,
>
> (I've cc'd the pyglobus mailing list since this is actually a
> pyglobus/globus problem, and since it's happening during the attempt to
> build packages it's not a "user" level error, but a developer problem.  
> The
> combination of those two facts are the reasons we closed the bug  
> yesterday.
> I'm sorry if that wasn't clearer from the email you got about it.
>
> But: it looks like there's a problem with either your globus install,  
> since
> globus-makefile-header is a script that's used to generate compilation  
> flags
> for programs that use globus. Are you absolutely sure you've run  
> whatever
> post-install scripts need to be run for globus to correctly configure
> itself?
>
> --Ivan
>
>> -----Original Message-----
>> From: owner-ag-tech at mcs.anl.gov
>> [mailto:owner-ag-tech at mcs.anl.gov] On Behalf Of Joseph
>> Sent: Friday, September 03, 2004 12:35 AM
>> To: 'ag-tech'
>> Subject: [AG-TECH] Building failure - strange behaviour with pyGlobus
>>
>> Hi
>>
>> In fact, since the Beta2 release, my build isn't working
>> anymore. The pyGlobus module isn't build. In my build output,
>> I find a lot of :
>> "Building pyGlobus
>> Checking for which modules to build
>> Checking the util module
>> Missing dependency to build the util package.
>> This package is required to build all of pyGlobus.
>> The following is the error message from globus-makefile-header.
>> ('No packages were found that matched your query!\n',) Now
>> exiting from the build."
>> And so on, then
>> "ImportError: No module named pyGlobus"
>>
>> Nothing had changed between the Beta1 and the Beta2 on my pc,
>> so I filed a bug report but apparently it wasn't one, the
>> trouble should be with my installation of environment variables.
>>
>> However, just to be sure, I have
>> - resintalled gpt and globus : as usual, only the
>> globus_trusted_ca_42864e48_setup-noflavor-pgm appears not to
>> be correctly configurated after a gpt-postinstall and then
>> gpt-verify, but I was told before it's all right,
>> - checked the GPT_LOCATION and GLOBUS_LOCATION (they are set).
>>
>> So I looked closer at the error, this globus-makefile-header.
>> I have found a very strange behaviour : when I change the
>> line 230 in /usr/lib/globus/bin/globus-makefile-header the
>> behaviour of the build process is deeply affected even if a
>> just add the word "TEST" in the output string. In brief, I
>> just change the line 230 from :
>> 	printf("No packages were found that matched your
>> query!\n"); to :
>> 	printf("TEST No packages were found that matched your
>> query!\n");
>>
>> The new behaviour is astonishing, I now have :
>> "Building pyGlobus
>> Checking for which modules to build
>> Checking the util module
>> The util module will be built
>> Checking the ftpClient module
>> The ftpClient module will be built
>> Checking the ftpControl module
>> The ftpControl module will be built
>> Checking the gassCache module
>> Traceback (most recent call last):
>>    File "setup.py", line 915, in ?
>>      pyModList, eModList, swigList = build_module_lists()
>>    File "setup.py", line 247, in build_module_lists
>>      check_gassCache(pyModList, eModList, swigList)
>>    File "setup.py", line 425, in check_gassCache
>>      major, minor = check_version(dep)
>> TypeError: unpack non-sequence
>> Checking for which modules to build
>> Checking the io module
>> using version 4.0
>> The io module will be built
>> Checking the security module
>> using version 3.5
>> The security module will be built
>> Checking the util module
>> The util module will be built
>> swig -python -noproxy -DHAVE_CONFIG_H
>> -I/usr/include/python2.3 -o ./src/io_wrap.c ./src/io-4.0.i
>> swig -python -noproxy -DHAVE_CONFIG_H
>> -I/usr/include/python2.3 -o ./src/gss_wrap.c ./src/gss.i swig
>> -python -noproxy -DHAVE_CONFIG_H -I/usr/include/python2.3 -o
>> ./src/gsi_wrap.c ./src/gsi-3.5.i swig -python -noproxy
>> -DHAVE_CONFIG_H -I/usr/include/python2.3 -o ./src/util_wrap.c
>> ./src/util.i running build"
>>
>> Do you have any clue of what I'm doing wrong ? I just don't
>> understand !
>>
>> Cheers,
>>   Joseph
>>
>>
>>
------------------------------------------------------------------------ 
------------------
Keith R. Jackson, Lawrence Berkeley National Laboratory
1 Cyclotron Rd. MS: 50B-2239, Berkeley, CA 94720
tel: 510-486-4401  fax: 510-486-6363
krjackson at lbl.gov http://www-itg.lbl.gov/~kjackson/
------------------------------------------------------------------------ 
-----------------
-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 2787 bytes
Desc: not available
URL: <http://lists.mcs.anl.gov/pipermail/ag-tech/attachments/20040903/b69bf5aa/attachment.bin>