Hello all, I maintain python-albatross, a python web toolkit (http://www.object-craft.com.au/projects/albatross). Following (the proposed) python policy, I've created a dummy package python-albatross, and two versioned packages, python2.2-albatross and python2.3-albatross. The dummy package depends on the 2.2 version. Now, I have some questions regarding this scheme. Firstly, it seems a little silly to duplicate all the code in both packages. I guess there's really no other way of doing it, since there are build-time test cases that should be run with each version of python, and the install-time bytecompiling and module optimizing that much be run with each version of python as well. Are there any ideas for how to reduce the size without making it all an unmaintainable mess? (python-albatross is quite small, but this may be useful for bigger packages.) Secondly, the above scheme makes it impossible to install both packages at the same time. This is because both packages provide the same initscript, defaults file and logrotate file. One "solution", which may be against policy, is to have the versioned packages conflict with each other. Another solution would be to split all the common parts into a separate package, which is then depended upon by both versioned packages. But I don't know if this is permitted by the policy, or if it will introduce any side-effects that are undesirable. Perhaps it can be done, but should use versioned dependencies somehow? Please tell me what you think. Cheers, -- Fabian Fagerholm <fabbe@paniq.net>
Attachment:
signature.asc
Description: This is a digitally signed message part