On Tue, Feb 19, 2013 at 06:38:23AM -0700, Alan Robertson wrote:
| Hi Luis-Claudio:
|
| I think there's a way to exclude files from packaging, but I'll have
| to look into that.
You can probably do that on the 'install' target of cmake. These files are
there because you actually ran the tests before packaging the files.
Removing them after the test run would also work.
| I may know how to fix the description. [But I don't know where the
| unexpanded macros came from].
CPackSourceConfig.cmake uses the project README file as the package
description:
SET(CPACK_PACKAGE_DESCRIPTION_FILE "/home/lclaudio/code/hg/assimilation/README")
SET(CPACK_PACKAGE_RESOURCE_FILE_README
"/home/lclaudio/code/hg/assimilation/README")
I suggest a shorter description... and having a one-line description to use
with the 'Summary:' rpm parameter.
| I thought I did the architecture correctly. But obviously not ;-)
| I have some 64-bit machines to play with...
CPackSourceConfig.cmake has these lines:
SET(CPACK_COMPONENT_CMA-COMPONENT_GROUP "cma-0.1.0.i386")
SET(CPACK_COMPONENT_NANOPROBE-COMPONENT_GROUP "nanoprobe-0.1.0.i386")
And they get copied into CPackConfig.cmake, used by cpack. When I run 'cpack'
or 'cpack -G RPM', the specfiles it generates have these lines:
* assimilation-cma-0.1.0.i386.spec
%define _rpmfilename assimilation-cma-0.1.0.i386.rpm
* assimilation-nanoprobe-0.1.0.i386.spec
%define _rpmfilename assimilation-nanoprobe-0.1.0.i386.rpm
Thus forcing the package names to have 'i386' on it, no matter what arch is
the real target. Removing the 'i386' string from the
CPackSourceConfig.cmake just creates RPM packages that simply have no arch
reference in the name.
My cpack-fu is weak, but there may be parameters to tweak this behavior and
to set some other interesting parameters. If the kids behave, I may have
some time to squint through the cpack docs.
Cheers,
Luis
| Thanks much for your testing!
|
| Let me know what else you find out!
|
| -- Alan Robertson
| [email protected]
|
| On 2/19/2013 4:32 AM, Luis Claudio R. Goncalves wrote:
| >On Mon, Feb 18, 2013 at 09:50:00AM -0700, Alan Robertson wrote:
| >| Hi Luis,
| >|
| >| Could you try building the latest version and see if it compiles for you?
| >|
| >| And try to create packages for it as well?
| >|
| >| This is all described in the README.
| >
| >Hi Alan!
| >
| >Last night was quite an adventure. Even though Nicole slept the whole
| >night, Martin spent most of it awaken and crying out as loud as his little
| >lungs allowed him too (which was surprisingly loud). We suspect he had
| >colics because my wife had some grape juice during lunch... :(
| >
| >Anyway, I was able to work for about 15 minutes and built the code on both
| >Fedora 18 and Fedora 17. The initial result was the same on both: the build
| >worked fine and I was able to get RPM packages with 'cpack'. I haven't
| >been able to install/test the RPMs but I noticed three issues on them:
| >
| >* Hard-coded 'i386' arch on package names, even though they have been built
| > as x86_64 binaries.
| >
| >* The RPM description is too long and filled with macros that were not
| > expanded.
| >
| >* The .pyc and .pyo files were packaged with the cma python code. Removing
| > these files from assimilation-cma-0.1.0.i386.rpm would allow it to be a
| > 'noarch' package, as most python packages are.
| >
| >Just to be thorough, these are the files I believe shouldn't be packaged:
| >
| >/usr/lib64/python2.7/site-packages/assimilation/tests/__init__.pyc
| >/usr/lib64/python2.7/site-packages/assimilation/tests/__init__.pyo
|
>/usr/lib64/python2.7/site-packages/assimilation/tests/cclass_wrappers_test.pyc
|
>/usr/lib64/python2.7/site-packages/assimilation/tests/cclass_wrappers_test.pyo
| >/usr/lib64/python2.7/site-packages/assimilation/tests/cma_test.pyc
| >/usr/lib64/python2.7/site-packages/assimilation/tests/cma_test.pyo
| >
| >Later today, hopefully before you wake up, I may be able to devote some
| >time to fix the first issue I mentioned. The third one should be easy too.
| >The package description... well, that's surely a task for you :)
| >
| >Cheers,
| >Luis
---end quoted text---
--
[ Luis Claudio R. Goncalves Bass - Gospel - RT ]
[ Fingerprint: 4FDD B8C4 3C59 34BD 8BE9 2696 7203 D980 A448 C8F8 ]
_______________________________________________
Assimilation mailing list - Discovery-Driven Monitoring
[email protected]
http://lists.community.tummy.com/cgi-bin/mailman/listinfo/assimilation
http://assimmon.org/
