I’m on record in a few blogs and a few places stating that I feel the rapid release program is wrong in it’s current implementation and I stand by that. I don’t have all the research at hand of someone who works in Firefox management but it goes without saying that the people punching the numbers are worried about dwindling market-share.
Firefox was really perceived to be dragging its heels in the time between the releases of Firefox 3.6 and Firefox 4 and in that time, where was a seemingly lightweight simpler option that came on the market in the form of Google Chrome. On top of that, Microsoft finally got their act together and managed to successfully put out IE9. Microsoft sent Mozilla a release cake and Mozilla sent Firefox one too. It’s a tradition they had for major releases and with the rapid release system, I guess it’s come to an end which is a shame.
As things stand, there will be a major version bump every so often, indiscriminate of features. And this is where I feel things have gone wrong. Chrome is all about rapid release and having reached a whopping version 14 at the time of writing this, they’ve seemingly put pressure on Mozilla to follow suit. Firefox was at the time on version 4 and despite belieivng that version isn’t anything but a number, decided to quickly inflate the number.
However, I’d have put this all of that Product Managers door. I’m not against having rapid releases but don’t see why they have to major releases too. If you’re on version 4, why can’t the major releases by default be .x releases? Why can’t the project manager simply have a look at all of the work that’s going on, put a couple extra people on one bugs to expedite work in certain circumstances and when he believes he has enough for a major bump. Bump the version number. In such cases, you have someone code efficient to be able to verify that developers aren’t stalling on bugs and have a simple rule in place where by a feature is only landed in an off-state twice. Have another rule in place where by you bump version every six months and such releases are perfect for landing binary compatibility breaking changes. Based on such a system, you’re likely to see at least two version bumps a year. But more importantly, you don’t land non-finished features and not have them performing logically four versions later.
I’ve gone on record saying that the version number is the greatest fan of the fanboy! It’s what gives them the ammunition to try again to talk up their browser, no to champion their favourite browser. It’s a shame that Mozilla wants to do away with it’s biggest tool. But what’s more of a shame is that it seems to late in the day to reverse this action.