[Mageia-dev] FireFox ESR <= we should totally go for this wrt stable releases

Fri Feb 10 15:15:17 CET 2012

Am 09.02.2012 21:56, schrieb Maarten Vanraes:
+> Op donderdag 09 februari 2012 13:04:00 schreef Sander Lepik:
+>> 09.02.2012 13:35, AL13N kirjutas:
+>>> ihmo, the users who want the latest versions will not even wait one week,
+>>> and will grab them from upstream. If we really wanted the latest release,
+>>> someone could backport them.
+>> And backports work since what time?
+> that is irrelevant, and we should not assume it will not be ready.
+>
+>>> the end users will benefit too, because the modules/plugins will still
+>>> work and not being unstable at every update. as a long time stable user,
+>>> i prefer the ESR... we've been complaining about too short
+>>> releasecycles... this is the answer, we should use it.
+>> Most users don't know what to do with tarball. They only know that their
+>> browser doesn't have latest promoted changes. And that means
+>> distribution is doing something wrong for them.
+>>
+>> We were complaining because extensions didn't work. I don't know how
+>> many plugins there were with problems. Can't remember any myself.
+>> Nonbinary extensions are now compatible by default, so one problem less.
+>>
+>> ESR means QA testing anyway. It's only bugfixes but that doesn't mean
+>> you can update w/o testing.
+> iinm our update policy dictates that our updates should only have updates in 
+> it, and no new features... (unless there is no such supported version)
+Our update policy already has an exception for Firefox, and so far there were
+no bigger issues during updates from our initial 4.0 to 10.0. And if most other
+distros don't use ESR, we're kind of alienated, and users will ask why we do that,
+because we're not aimed at enterprise usage currently.
+
+>
+> well, now there is... Either the new versions with extra features go into 
+> backports, or not at all. if this becomes a problem, then it's one more reason 
+> to fix the backports.
+>
+> we have a clear policy, we should follow it.
+Hypothetically, it would be kinda useless to have two versions of our default
+browser,
+meaning twice the QA and support, and backports are surely not opened solely
+for firefox.
+