changelogs in packages and space use

[Eric Sandeen]

David Cantrell wrote:

> Arbitrarily deciding to remove all changelog entries older than 1 year is stupid.  For some packages you'll end up with one entry (or none!) and for some packages you probably won't make a dent (kernel?).

It seems reasonable to me, as a simple starting point, to ask
maintainers to trim changelogs to information that is relevant to that
%{VERSION}.  (maybe last 2 versions).  Ideally that changelog would
contain documentation for all patches still carried, and go back to the
changelog entry that bumped the version (or 2).  Perhaps if that version
bump caused patches to be removed, that should also be in the changelog,
with rationale ("patch FOO removed, now upstream").

If the changelog entries have the format:

* <date> <author> <version>

then presumably automated checks could be run for changelog entries
older than the current revision (or 2), and warnings or errors issued.

If you must know what happened before that, go get the older rpm and
take a peek.

I know this doesn't address the replicated data issue, but it might
address the size of the data being replicated, at least :)

-Eric