When It’s Ready
February 11th, 2008 by Asa
Last week, a number of articles and blog posts noted that we had added another Firefox Beta (the 4th) to our schedule. Most of the Beta 4 mentions were made in articles and posts that were more heavily focused on the upcoming release of Beta 3 but several did investigate (or at least speculate) and report on the reasons for the addition of Beta 4 to the Firefox schedule.
As several of the linked posts below accurately reported, we’ve decided to add one more beta in order to make further improvements in the areas of polish, performance, memory, and overall quality. That does beg the question, how does Mozilla decide when to stick to the prior schedule and when to adjust the schedule. That’s what motivates this post.
The Mozilla product cycle is actually quite a bit different from those of most software producers. I’ve asked Mike Beltzner, the man driving the Firefox 3 release, to help explain some of the differences and how those differences informed the decision about adding a fourth beta to the schedule.
The first and most important thing to state is that we, as a project, are quality driven, not date driven. We use dates to set targets for milestones, and we strive to put out the best milestones possible, but due to the changing nature of the web, we always judge each milestone against our basic criteria of quality, performance and usability. The other factor how we rely on community use and testing of our betas to help us judge their quality. Without our close-to-million strong beta audience, we couldn’t get the usage and testing data required to provide us with the confidence to move into a shippable product.
In the context of adding a fourth beta, the decision was between holding beta 3 for some new work being done on performance, usability, and memory use, versus shipping the beta sooner to get testing on the improvements (based on community feedback) from the past month and a half. We decided that the best thing for our beta audience would be to release beta 3 and add another milestone as opposed to holding for the late-braking improvement opportunities. Those will be the basis for our fourth beta.
We all want to ship the best browser possible on the fastest timescale, and we meet weekly to evaluate our current product against the potential improvements we could continue to make. When our quality, usability and performance are solid enough to be called Firefox 3 (as judged by everyone involved in making it) we’ll ship it.
Often times folks in the Mozilla community, when asked when the next version will be available, use the short-hand “When it’s ready.” If you’re looking for a more precise description for “When it’s ready”, feel free to quote Beltzner or to reference this article.
More reading:
Mozilla Developer News » Firefox 3 beta 3 code freeze tonight, and
Mozilla Wiki » Firefox3/StatusMeetings/2008-01-29
Around the Web: ZDNet.com » Getting closer: Firefox 3 Beta 3…, ComputerWorld » Mozilla freezes Firefox 3.0 Beta 3, InternetNews » Firefox 3 Beta 4?, CyberNet » Firefox 3 Beta 3 Coming with a Big New Feature, Softpedia » Firefox 3.0 Beta 4 Confirmed…, Download Squad » Mozilla updates Firefox…, Gadgetell » Firefox 3.0 Beta 4 expected shortly, D’Technology Weblog » Firefox 3.0 Beta 4 confirmed, The OpenSourceDeal Blog » Firefox 3 update, Communication Technology » Firefox 3, Technology Basement » Firefox 3.0 Beta 4 up in the Air…, Krishna Srinivasan’s Blog » Firefox 3 Beta 3 to be out Feb 11, Gadgets Control » Firefox 3.0 Beta 4 expected shortly, iloog » Firefox 3 B3 completed at least will have a Beta, Just Thinkin’ » Firefox 3 Beta 3 Update, Security Watch » Firefox Updates Imminent, Electronista » Mozilla locks Firefox 3.0 Beta 3
Thanks for the link Asa, much appreciated. It always pats to do your homework about things like this, otherwise any credibility you might have had tends to wear a bit thin. I’ve always respected the idea of “when it’s ready” rather than adhering to a strict release schedule. The old saw of getting it right the third time just doesn’t cut it any longer (now what major software/OS company might I be referring to I wonder?).
I’m mighty impressed with what the nightly builds have shown so far in my constant use and so are my friends and acquaintances who are also using the nightly builds as their regular browser. It’s not very often that nightly builds can actually take the place of the previous final release.
Keep up the fantastic work.
Of course I did make the mistake on quoting a release date from Mozilla Links that said beta 3 was being released today so I guess the day will prove whether I was accurate or not. :P
Given the number of remaining blockers, I almost hope there’ll be a 5th or even 6th beta release before Fx3 ships. I think we all recognize how significant the Firefox 3.0 release is, so to me it’s better to take the time to get it right rather than let little things here and there slide for a later release, especially given the current uncertainty about whether Gecko 1.9.1 or Gecko 2.0 is next.
It may be a long time before some of the features currently blocking Firefox 3.0 have another chance to get landed as they may fall outside the scope of a 3.0.x point release. I’m just worried that the temptation will be very strong to just mark some blocking bugs blocking1.9- as the current beta cycle winds down in order to get the release out “when it’s done”
Hey Ryan,
I expect that we’re going to ship with some blockers not fixed. We’re definitely going to ship with some bugs that someone (most notable us) finds irksome, and I’m sure they’ll let us know (in our open and public bug tracking database) how we’ve lost their respect and they’ll take their business elsewhere.
We need to make sure we’re not optimizing for perfect at the cost of good. Firefox 3 is currently *way* better than Firefox 2: better for users, better for developers, better for the web. Of course it can be better. It can always be better. The real question is understanding when it’s good enough for prime time use, and when we should turn our focus to the next major release.
Mike, ultimately I agree with what you’re saying. I guess I’m more just hoping that there ends up being a Gecko 1.9.1-based Fx3.x release so that blockers which do end up missing the cut don’t end up waiting until Moz2 to come to fruition.
“When it’s ready” is no longer an appropriate answer. That was an okay response when we had zero users.
Mozilla needs to give people a time frame so people can prepare and plan because it’s work for people to test/update their web sites/apps/extensions to work for Firefox 3.
Q4 2008? Q1 2009? Summer, Winter ‘08. That’s the minimum, no one is looking for the exact date and it’s ok to slip.
I don’t think having yet another beta would be too drastic in terms of sacrificing good over perfect. Having a well polished application is a real pleasure to use and I didn’t think I’d ever see a web browser like that till I started using some of the new builds (Beta 3 and onwards). Well, except of course adhering to ISO rules which the Firefox development team seem to flat out refuse to do *sigh*.
But there’s still some fairly obvious issues and I’m worried that many of them won’t be fixed for beta4, one more small beta to check for polishing and to fix any big issues thrown up by beta4 (still have to let the masses use new memory allocator and stuff) sounds good to me.
Its fine if it takes time. I would say fixing breaking things first before going for new features. There are plenty of blockers that affect daily use of the browser. We still see features being added to beta. FF3beta is definitely better than FF2 but many functionality that were missing in FF2 are still missing in FF3. And they are low prioritize…(Example, Bug of drag and drop in Places/Bookmarks). Image drag and drop in Composer (on linux)….
Conclusion is focus on Usability rather than features.
I just hope that this bug that crept into FF2 is fixed
Bug 383369 – Secure (encrypted, https) sites loading as being partially encrypted. Broken lock is present as is the white address bar
This presents much confusion to users as evidenced by the numerous threads on mozillazine. Having to explain that if they open a new tab and load again the secure lock will be there leads to extended discussions with regard to security.
This on the bug page bothers me
Flags:
dveditz: blocking1.8.1.13+
dveditz: wanted1.8.1.x+
beltzner: blocking‑firefox3+
Robert,
blocking+ means it has a high chance of being introduced to that build of Firefox, in this case 2.0.0.13 and Firefox 3.
Robert, to expand on what Damian said, those flags mean that the people in charge regard that bug as “blocking” those releases, and therefore a priority to get fixed.
Specifically, dan veditz (dev lead on firefox 2.0.0.x) wants that bug fixed before 2.0.0.13 ships, and mike beltzner (overall lead on firefox 3) wants it fixed before 3.0 ships.
That doesn’t mean it *will* be fixed, but it’s a pretty serious statement of intent.
I have to agree with Rafael. An exact date is not required, but it would be nice if we could have a guesstimate on a timeframe so we can prepare
ScottishDan, I honestly don’t think at the moment the developers could give a date if they even wanted too. It looks more and more likely we’re going to get a beta5 to add that final piece of polish (e.g some feedback all Vista icons as there’s no full implementation yet) and lots of other outstanding polish bugs.
QA looks like it’s going to go at least 2 weeks since code freeze (started on the 26th).
I don’t know we’ll get a serious estimate of launch till we get to at least preRC1 or RC1.
This is a crucial time in the life of Firefox. When Firefox came out, IE had no tabs and ran every malware it could find. With IE7, for many users IE is a passable alternative with tabs, and more focus on security than it used to have. Firefox now has a reputation as a memory hog. Firefox 3 must address this.
Nowadays more pages use Flash - and the sad truth is these pages often work better in IE. One of the most popular sites, Youtube, for many Firefox users plays without sound! Reliable support for Flash in Firefox is essential for its success. Mozilla should have regular meetings with the programmers at Adobe and beat some sense into them. Either that, or give money to the Gnash project. Without plugins, Firefox usually runs like a dream. But for users who depend on Flash, IE looks better all the time.
Take your time, get it right. Because if Firefox 3 is not excellent and much better with memory and plugins than Firefox 2, Firefox market share will start going back down.
My favorite bug 383369 appears to be alive and well on FF3b5. Other than that I find FF3b5 absolutely awesome. Am using it daily on OSX Tiger with no issues other than not being able to use the sort bookmark capability (the sort although working in the Library window doesn’t “stick”) which is not an issue for me.
Great News!
https://bugzilla.mozilla.org/show_bug.cgi?id=383369
Kai Engert (:kaie) changed:
What |Removed |Added
—————————————————————————-
Status|NEW |RESOLVED
Resolution| |FIXED
thnaks for adding my url in this post
I just hope that final release is happening in the near future as FF2 is now almost completely useless for me because of constant crashes on random web sites.
I also sencierly hope all the common plugins, FireBug, Better Gmail and similar will be available as common plugins soon after release day.
What I’ve seen so far on beta releases is a huge improvement in terms of speed, memory management and rendering.
Please don’t make us use IE7 again …. :)
My suspicious mind suggests that Firefox 3 was going head-to-head in the final release schedule with Opera 9.50, as both browsers have been in perpetual development since forever, but guess what! Opera 9.50 has been released at long last, so pull yer fingers out guys!!
Heyyy is about time to launch the new FF3…. i am waiting… in the oficial site there is nothing :S …
good mozilla
Can’t wait to check out the 15,000 enhancements/fixes lol… so where is it :) I’m on the edge of my seat here!