Mozilla recently announced the planned implementation of an Extended Support Release (ESR) for enterprise users and other users who do not want to take part of the rapid release cycle. We touched on ESR briefly back in October 2011. Under the ESR plan, there will be one ES Firefox release per year (the first based of the upcoming Firefox 10 due out at the end of this month). Under the rapid release process, there is a new Firefox release every six weeks which is about 8 releases per year (which is fairly close to the amount under the old release process).
I do think this is a great idea, however I have issues with the ESR process. It does address the concerns raised by Enterprise users. However, once a year I just don’t think is frequent enough. Keep in mind, there will still be regular stability and security updates throughout the year, but only one major update per year. A couple issues I could see with doing ESR once a year:
- Technology (especially the Internet) changes way too much in a year. ESR users are going to be stuck with older (and possibly obsolete) browser technology and could possibly be shut-out from sites that have adapted newer technologies. This almost seems like a major step backwards. Looking back at the time between past major releases, Firefox 3.5 was released twelve months after Firefox 3.0; Firefox 3.6 was released only six months after Firefox 3.5; However, Firefox 4 was released fourteen months after Firefox 3.6. However, the situation the Firefox 4.0 was a bit unusual and skews these figures a bit. Further, Mozilla made several mistakes during the 14-month development process. These including scrapping the planned Firefox 3.7 release in May 2010; moving to Gecko 2.0 (instead of doing one more release under Gecko 1.9.x) and continuously delaying the final release of Firefox 4 in an attempt to get all the planned features working. However, I suppose Enterprise users may not be that adversely affected by this and some of this could be compensated with add-ons.
- New feature/change overload (much like we saw when Firefox 4 was rolled out last spring). One of the reasons I like rapid release is that there a couple new features rolled out with each release (and release are based on a fixed time schedule not a feature ready schedule). This allows users to get new features/changes sooner instead of having to wait for other planned features in that release to be completed (much as was the case with Firefox 4). Plus, with only a couple per release, this allows the user to get use to these features/changes without it seeming so overwhelming. I am concerned ESR users are going to be in for quite a shock when the next ESR comes out in early 2013.
Again, the first planned ESR is going to be based of the upcoming Firefox 10 release at end of the month. This would be about 10-months after the Firefox 4 release. This is a span of only 6 rapid release releases (a typical year will be around 8 or 9 releases). Now, consider all the changes/features that have happened in the span of these 6 releases (new web developer tools, tab loading, 3rd party add-ons handling, and various adaptations and handling of new web technologies (more so on the backend). Imagine if you had not updated since Firefox 4 and then made the move of Firefox 10…it would be quite a shock, much like moving from Firefox 3.6 to Firefox 4.0 was. This is the way it would be with an annual ESR. This is why I would think semi-annual would work better. There would be two major roll-outs a year instead of one (still would be easier on IT departments than 8 to 9 per year). Plus users would get more of the new changes/features sooner and without being overwhelmed.
I suppose much like Rapid Release there were kinks that needed to be worked out with the ESR process as well. It just seems like it is going to be more difficult with only doing these once a year. This should prove to be quite interesting.