Isn't this bug filled against the wrong component (pysnmp)? Shouldn't it be python itself?

ping?

IMHO, nobody cares

Fedora-11:

[sergeil@homedesk ~]$ ls -1 /usr/lib/python2.6/distutils/command/*.exe
/usr/lib/python2.6/distutils/command/wininst-6.0.exe
/usr/lib/python2.6/distutils/command/wininst-7.1.exe
/usr/lib/python2.6/distutils/command/wininst-8.0.exe
/usr/lib/python2.6/distutils/command/wininst-9.0-amd64.exe
/usr/lib/python2.6/distutils/command/wininst-9.0.exe

$ rpm -qf /usr/lib/python2.6/distutils/command/*.exe | sort | uniq
python-2.6-11.fc11.i586



Fedora-12 (beta)

[root@eee901 ~]# ls -1 /usr/lib/python2.6/distutils/command/*.exe
/usr/lib/python2.6/distutils/command/wininst-6.0.exe
/usr/lib/python2.6/distutils/command/wininst-7.1.exe
/usr/lib/python2.6/distutils/command/wininst-8.0.exe
/usr/lib/python2.6/distutils/command/wininst-9.0-amd64.exe
/usr/lib/python2.6/distutils/command/wininst-9.0.exe

# rpm -qf /usr/lib/python2.6/distutils/command/*.exe | sort | uniq
python-2.6.2-2.fc12.i686

pysmnp is the wrong component. You are talking about python.

Thanks for filing this bug report.

From my reading of /usr/lib/python2.6/distutils/command/bdist_wininst.py the .exe files are templates that are used when constructing installers for python modules; the installer is formed from one of those files at the front, with the content glued to the back.

Attempting to run this on an F11 box to create a windows installer for a simple package gives this:
$ python setup.py bdist_wininst
running bdist_wininst
running build
running build_py
creating build
(...snip many lines...)
creating dist
Warning: Can't read registry to find the necessary compiler setting
Make sure that Python modules _winreg, win32api or win32con are installed.
removing 'build/bdist.linux-i686/wininst' (and everything under it)

However, the above warning is non-fatal, and it creates a .exe file in the dist subdir that appears to work (briefly tested in Wine).

So the .exe files are actually useful on Linux for some use cases.

$ du -h /usr/lib/python2.6/distutils/command/*.exe
60K	/usr/lib/python2.6/distutils/command/wininst-6.0.exe
64K	/usr/lib/python2.6/distutils/command/wininst-7.1.exe
60K	/usr/lib/python2.6/distutils/command/wininst-8.0.exe
220K	/usr/lib/python2.6/distutils/command/wininst-9.0-amd64.exe
192K	/usr/lib/python2.6/distutils/command/wininst-9.0.exe

...giving 596K of disk usage for a rarely-used feature in a package that's installed on most systems.

> $ du -h /usr/lib/python2.6/distutils/command/*.exe
> 60K /usr/lib/python2.6/distutils/command/wininst-6.0.exe
> 64K /usr/lib/python2.6/distutils/command/wininst-7.1.exe
> 60K /usr/lib/python2.6/distutils/command/wininst-8.0.exe
> 220K /usr/lib/python2.6/distutils/command/wininst-9.0-amd64.exe
> 192K /usr/lib/python2.6/distutils/command/wininst-9.0.exe
> 
> ...giving 596K of disk usage for a rarely-used feature in a package that's
> installed on most systems.  

mclasen: IIRC you're fighting a constant battle for disk space on LiveCDs.  How significant a saving would the above be?

I've taken smaller savings in the past, but we are not in immediate emergency mode atm.

This bug appears to have been reported against 'rawhide' during the Fedora 12 development cycle.
Changing version to '12'.

More information and reason for this action is here:
http://fedoraproject.org/wiki/BugZappers/HouseKeeping

Would it make sense to move these to the devel subpackage? I don't want to break being able to build for MS Windows from Fedora but afaik these will never be needed during runtime, only when doing build. This would make it so we don't lose this feature completely, but will give us the space savings for the python runtime package.

Good idea - I've moved them to the -devel subpackage in python-2.6.5-6.fc14 as part of this commit
http://cvs.fedoraproject.org/viewvc/rpms/python/devel/python.spec?r1=1.181&r2=1.182

*** Bug 1426250 has been marked as a duplicate of this bug. ***