On Tue, 26 Jun 2007 15:17:24 -0700 Wes wrote:
WH> There are vendors that try to be more backwards compatible than we do,
WH> and I hear from them all the time that their users complain about our
WH> .so suffix numbers jumping upward because it breaks older application
WH> builds.  (and no, they don't want to rebuild as their applications
WH> aren't likely to be affected by new things).

Then they should provide 'compat' libraries. That's why there are versions in
the library names - so multiple versions can be installed.

It's also an artifact of our policy on supporting multiple old releases for so
long. If we were willing to have a much more restrictive policy on changes
allows in released branches, we wouldn't need the forced version jump on major
releases.

WH> Sigh.  Really, I hate to say it, we should be using different libtool
WH> versions per library.

Indeed we should.

WH> That'll only solve some of the problems though.  When the oldest
WH> structures known to the project haven't changed, then an application
WH> certainly won't be affected by a struct update to something new and
WH> fancy that they're not using.

Yes. And if we imposed a more restrictive policy on libnetsnmp, we could
probably get away without the forced jumps between releases (at the expense of
maybe having to _not_ be able to fix some bugs in older releases, and require
users to upgrade).

-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/
_______________________________________________
Net-snmp-coders mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/net-snmp-coders

Reply via email to