4 ms·
Also 16b1 and 17a2 coming out sometime tomorrow. October 9th: Firefox 16 desktop/mobile, 17b1, 18a2 November 20th: Firefox 17 desktop/mobile, 18b1, 19a2 Sche
by ck2 14y ago
Also 16b1 and 17a2 coming out sometime tomorrow.
October 9th: Firefox 16 desktop/mobile, 17b1, 18a2
November 20th: Firefox 17 desktop/mobile, 18b1, 19a2
Schedule pace seems insane but getting used to it and I'm fine with the way it doesn't seem to break anything.
Firefox 99 in 2017?
- robin_reala 14y ago17a2 is out now (or at least my Aurora just updated itself).
- masklinn 14y ago> Firefox 99 in 2017? 8.6 releases/year (release every 6 weeks), 2013 will start at Firefox 18, so Firefox 99 will be released in the first half of 2022.
- melling 14y agoWhy does it seem insane? "Release early, release often" has been around for years. Both Chrome and Firefox both update every six weeks. That means over half of the Internet is now upgrading at a nice clip. http://gs.statcounter.com/#browser-ww-daily-20120828-20120828-bar http://gs.statcounter.com/#browser-ww-daily-20120828-2012082... The big question is how to get more people on a 6 week cycle.
- ck2 14y agoBecause the major version number is now the minor version. In theory we should be on Firefox 9.15 or something. Now big number changes mean small improvements.
- AndrewDucker 14y agoNo, big number changes mean that six weeks have passed. That's all. The scale of the change doesn't affect the number at all.
- lhnn 14y agoThat's his problem: The numbers in versions have traditionally (and more usefully) been used to denote levels of feature change and bugfix. A time-based arbitrary version number is pointless; it'd make more sense to use a timestamp for a version number if they're going to do it that way.
- Karunamon 14y agoWhich is kind of the problem. The standard way of doing it is major release.minor release.build If it wasn't for this whole "Let's start bumping our version number to match Chrome's insanity" thing, we'd still be on Firefox 9 something.
- joedrew 14y agoIt has nothing to do with Chrome, and everything to do with 6-weekly updates. Think of it: You have no way to know, going in, whether the changes in any given update will be big enough to "warrant" a major version change. Worse, what if those features aren't stable enough to release and are turned off? It's very difficult to go from version N to version (N-1). For these reasons (and others), it's a lot easier for us to unconditionally bump the major version every 6 weeks. Source: I am a Firefox developer and a Mozilla employee.
- pbiggar 14y agoSo what?
- asadotzler 14y agoThe big number change means we're breaking binary compatibility which does happen every 6 weeks now. The version isn't meaningful to most users these days but it does matter to developers who must adapt and adjust as Firefox progresses.
- ConstantineXVI 14y agoIMHO with this kind of steady release schedule, it'd make more sense to go to year.month version numbers.
- masklinn 14y agoWhy? What's even the point for the year.month version numbers? I know ubuntu uses that, but as far as I'm concerned it's a completely useless complexity, the first half increments half the time, and the second half repeats non-stop, parsing that is actually more annoying than a straightforward "is 22 bigger than 19?". Not to mention the ultimate goal (of Chrome and Firefox both) is probably to have "version number" fade away for everything but debugging/problem checking, users shouldn't have to care, the browser is "always up to date" and that's that.
- ConstantineXVI 14y ago> "version number" fade away for everything but debugging/problem checking That's exactly why. Doesn't even need to show an explicit number to users, just "You are running the August 2012 stable release" would do.