Thanks for the calm cool reply. :) I freely admit that I’ve been very emotional about this given the frustrations we’ve run into recently. SO I appreciate your looking past that.

Based on what you’ve described, I am indeed looking forward to Firefox 10. I really do want to continue to recommend and officially support Firefox as our browser-of-choice for our users.

And I hope that we can also see some disentanglement of Thunderbird from the Firefox roadmap. T-bird can be on a cycle, but don’t make it rush to keep up just because it’s time for a Firefox release. Let Firefox be Firefox and Thunderbird be Thunderbird.

I realize that from a developer’s POV, rapid release has lots of benefits. But for the end user it’s generally disruptive – there’s no (at least from where I sit) disputing that – and doubly so for the IT groups who deploy it. And no, I’m not on the enterprise working group. I’ve never had real call to know of its existence until now. But if I can find the time, I’m certainly going to look into it. Thanks.

I’m going to wipe the slate, so to speak, on myself and a couple of users next week-ish and move them to 8.0.1 Firefox and T-bord and see how it goes. I’ll let you know how it goes.

PS – And yes, I’m one of the Boston SKeptics organizers and a regular TAM attendee. :)

Now it’s release, release, release and pray to bob the bug fixed in 5 doesn’t show up again in the ‘all new super hot off the press’ 8.

We’re actually pretty good at not reintroducing bugs, because most of our bug fixes include automated regression tests. What’s harder is not introducing new bugs in features that used to work correctly. (We use the term “regression” to refer to both categories of new bugs, which can be confusing.)

I wrote a bit about regressions shortly before we switched to rapid releases.

As you say, rapid releases mean more opportunities for shipping regressions. But rapid releases also prevent regressions, by establishing a beta audience and drastically reducing the number of crash landings of not-quite-ready features. My intuition is that these effects should about cancel each other out, but given your anecdotes, maybe I should research whether regressions have in fact become more common.

Our ability to catch regression should improve as more users join the beta channel, for what it’s worth. Which channel are you on?

Compare these two Kraken scores

The measured difference in JS speed is small enough to be a measurement error. Opera didn’t even have a release in the period, and its score “changed” more!

I can’t think of any drastic changes to Firefox’s JavaScript engine (the component tested by Kraken) between Firefox 4 and Firefox 8, fwiw. The biggest speed improvements were to startup and for computers with little RAM.

Firefox 9, on the other hand, will be the first release to include Type Inference. TI will improve the performance of most JavaScript (in particular, scripts that are statically typed in practice). I think it’s great that Firefox users will benefit from TI starting next month.

And boy is it RAM hungry.

In most measurements, Firefox uses the least memory of any browser, starting with Firefox 7, and it’s continuing to improve.

I’m glad we already shipped 7. Rapid releases are a key part of “making a better browser”, as you put it. We didn’t make the decision based on fear competition.

The rapid release cycle also tends to break plugin/add-ons, often for no other reason than the fact that this version, which isn’t much different, starts with a different number.

I think you’ll be happy to learn that add-ons will default to compatible in Firefox 10. The result should be better than what we had in the pre-rapid-release world.

I’m curious about your thoughts on my more recent post, Improving intranet compatibility, and my other blog posts about rapid releases.

Also, have you joined Mozilla’s enterprise working group? A large source of the delay in figuring out a “extended support release” strategy is lack of input from deployment admins.

Your assertion that the new release cycle is more “artificial” has been noted.

PS I see you’re a fellow Pharyngula reader :)

Thanks for any help in advance.

Debbie

I am also hoping that Apple simply comes out with 10.6.9 which would solve this issue. Not everyone wants or can go to Lion. I have an Intel Core Duo which can no longer be upgraded.

Off to try and will post back.