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

Fri Jan 13 14:27:51 CET 2012

Michael Scherer a écrit :
+> Le vendredi 13 janvier 2012 à 11:21 +0000, Claire Robinson a écrit :
+>    
+>> On 13/01/12 09:36, nicolas vigier wrote:
+>>      
+>>> On Fri, 13 Jan 2012, Sander Lepik wrote:
+>>>
+>>>        
+>>>> 13.01.2012 03:20, Maarten Vanraes kirjutas:
+>>>>          
+>>>>> see https://blog.mozilla.com/blog/2012/01/10/delivering-a-mozilla-firefox-
+>>>>> extended-support-release/
+>>>>> see https://wiki.mozilla.org/images/9/9d/Esr-release-overview.png
+>>>>>
+>>>>> ESR is a 1y extended supported release...
+>>>>>
+>>>>> looking at the image we'd be having supported versions for our 9month release
+>>>>> schedule every time... we should totally use this release and not go towards
+>>>>> FF11 for our release.
+>>>>>
+>>>>> We've been complaining about the too quick release schedule... this is our
+>>>>> chance!
+>>>>>
+>>>>> ( i think if the FF maintainer wishes, he could also do backports of the
+>>>>> regular releases... )
+>>>>>
+>>>>> i'm hoping everyone agrees? including FF maintainer?
+>>>>>            
+
+I'd say let's go for the long-term release, and let those wanting the 
+latest and greatest use the upstream version.
+
+1) Upstream will release probably at least a week before we could 
+release it as an update.  With a 6-week release cycle, we will always be 
+lagging.
+
+2) Mozilla (Firefox/Seamonkey/Thunderbird) is very conservative in their 
+requires, and their rpm for mdv (and thus mga) menus are well done, so 
+newer upstream versions are highly unlikely to have any problems on install.
+(Note that upstream Mozilla required libstdc++5 long after mdv had 
+migrated to libstdc++6 and had later dropped libstdc++5.)
+
+>>>> I don't agree. But i'm not the maintainer.
+>>>>
+>>>> Why not?
+>>>> * Since fx10 all non-binary extensions are compatible by default (so our
+>>>> main problem goes away).
+>>>> * fx10 in 6 months is dead old for users POV. Many unhappy users. Lower
+>>>> popularity for Mageia. (Ubuntu AFAIK is going with fast schedule).
+>>>> * We will miss too many new and cool features.
+>>>> * When we release
+>>>>          
+>>> We could say the same about any other software. Firefox was an exception
+>>> on updates policy because there was no other choice. But there's no
+>>> reason to keep it as an exception when they provide a supported version.
+>>>        
+
+Exactly.  The fewer such exceptions the better, where alternatives exist.
+
+>> With 12 months support more often than not it would need updating in the
+>> lifespan of the Mageia 9 month release anyway.
+>>      
+
+Sure, but not every 6 weeks whether there is a bug fix or not.
+On the average in the last few years, significant bug fixes in 
+Firefox/Seamonkey have been more like every 2 or 3 months.
+
+>> Firefox is one of those programs that people like to be bang up to date
+>> with.
+>>      
+> All softwares are one of those programs. The only one that some non
+> technical users do not want to be updated are those that they do not
+> know, like glibc, python, perl. But still, there is people that want it
+> up to date, so firefox is nothing special.
+>
+>    
+>> It is 'bragging rights' to ship with the latest and something
+>> reviewers always give version numbers of along with libreoffice, kde, gnome.
+>>      
+
+We will always be at least a week behind Firefox's latest release.
+So where do you think those really wanting the "latest and greatest" 
+will go ?
+
+> Sure, and we neither update libreoffice, kde, gnome or the linux kernel.
+> Some people do ( kde is upated by Fedora, as well as the linux kernel ).
+> So that's a consistency issue, about what we promise to users.
+>
+> Stability is just that, stuff that do not have interface changes every 6
+> weeks, stuff that do not have slight mistranslation everytime string
+> change, stuff that do not risk breaking software after every updates.
+>    
+
++1
+
+>> I understand the arguments to go with the 12 months support but I think
+>> for the reasons above we should stick with the normal release cycle or
+>> maybe even offer both?
+>>      
+> Offering both would mean to double our workload of supporting firefox,
+> and have no advantages by using the long supported release.
+>
+> And that's rather useless from my point of view, if the goal is to
+> reduce the workload. There is already enough work to support the
+> distribution.
+>    
+
+-- 
+
+André
+
+