Firefox 2.0.0.11 Released 199
BrianAU writes "Firefox 2.0.0.11 has been released, the Release Notes show the only major change as a correction of a compatibility issue with some websites and extensions as discovered in Firefox 2.0.0.10."
Re:As if this is news? (Score:5, Interesting)
The interesting thing is that it was the fastest ever release of a browser update. John Resig [ejohn.org] gives most of the details: A security patch in Firefox 2.0.0.10 was incorrectly checked in, and introduced a bug which was not caught by the testing process. That was only discovered after the release, so the code was fixed and the whole release process had to start up again. Three days later, the 2.0.0.11 update is available for forty languages and three platforms.
So, it reflects badly on Mozilla's testing efforts, though that is an area where Firefox 3 has made significant improvements with automated testing. It reflects well on their release process, which can push out a critical update in just a few days.
Version numbering (Score:3, Interesting)
I know the reason for this is supposedly the extension system, but that is not a valid excuse. An internal technical detail should not exposed to users like that, and even so, the reason is not the extension system, but that the version checking for extensions was designed wrong from the start.
Now, can we please have a sane two-part version number for 3.0 and up?
Re:Some people may prefer to not auto-update (Score:2, Interesting)
TO ALL DEVELOPERS (Score:1, Interesting)
Do not use revision numbers on your software that look like IP addresses. ESPECIALLY please don't use them in the user agent string so that these numbers appear in web log files. Such numbers muck up many things.
Use dashes, parentheses, brackets, underscores, r.10, or ANYTHING except 4 numbers in value range [0-255] separated by 3 periods. There is no way to tell them apart from an ip address. Then we have to write blacklists, manually scan for errors, adjust counts, grep through log files with filters for the front of the line, etc etc etc. Bleachk!
Re:As if this is news? - Actually it is news (Score:3, Interesting)
We were actually one of the companies that found the bug shortly after the release of 2.0.0.10 and if you can't see why this is news then I'm really glad you don't work on my dev team.
Just so we're clear on what the bug ACTUALLY was, the bug specifically effected the canvas drawing capability in the browser. It's not something they test for and frankly, given our experience developing for IE, it's not one they test for either (if IE's random and aberrant behavior is any indication, hell MS can't even make a browser that displays content in a compliant manner given the HTML spec).
A number of sites and web applications use this functionality specifically for navigation, and when Firefox was updated to 2.0.0.10 on many client machines automatically, some business critical web applications were seriously effected. Because of this it was a pretty serious issue.
The reason this IS news is because after confirming the bug and determining the extent of the effect on the user base, the Mozilla folks had nightly builds in our hands just hours after a fix was checked in. This got most of the immediately effected back to work within hours.
A number of us then independently verified the fix against our code and then provided rapid feedback to the team so they could issue a release.
This resulted in an astonishingly fast turnaround. I think the Mozilla folks are to be commended for both not resisting requests for a new release, and the speed with which they were able to respond to a bug effecting business critical web applications. If this had been MS we would have spent 2 weeks navigating mindless support bureaucracy and then fought with management excuses as to why a fix just "can't be turned around overnight." We would have then been forced to contact all of our customers and go into long, boring explanations most of them would never have understood... it's all down hill from there.
Why this IS big news: It is a really bright and shining example of why this type of development is succeeding even in a situation where recursion testing fails (and if you think recursion testing can't fail then you just haven't been developing long enough).
The other good thing that came out of this is we now have a mechanism where developers can subscribe to a mailing list alerting then to pending releases.
Not only did Mozilla respond with a technical fix to the bug AND promptly issue a release which addressed the issue, but they were humble enough to recognize there was a process related problem that needed addressing as well; they fixed that too.
ER
Re:If only... (Score:2, Interesting)
Re:TO ALL DEVELOPERS (Score:2, Interesting)
When a counterexample appeared that broke this behaviour (the year 2000), did programmers kick up a fuss and call for a change to the function? Or did they just start using the correct method to achieve their goal?