Firefox 8.0 is Released

Sorry Sanddancer but my experience as an add-on author is that writing free add-ons can quickly turn to a demanding job.

For example, if you look at the tab mix plus add-on source code you can find a very long list of workarounds to make it work in different contexts.

so it worked properly on Linux. there is no way to know whether an addon instaled 'for the whole' system was added by third party, or by package. only addons in Your home dir are considered as 'installed by you'

I am aware of the page and have written an Extension in the past. I am unclear about the following things a) Is Jetpack the way to go going forward? How much of the XPCom api is available through Jetpack? b) Is there an IDE I can setup that will allow me to get autocompletion for extension development? c) Is there an API browser for XPCom objects listing the objects, the methods available on them and the minimum version of Firefox that supports the particular api? Having a searchable index would be even better. d) When I was working on a Firefox extension I often got the feeling that I was doing something suboptimal because the development process seemed pretty broken. I was building extensions and had to restart the browser for them to take effect, getting a debugger going was more work than it should be, I wondered why isnt the Firefox ui itself written using HTML and css. I am hoping that more tutorials and workflow related posts will address many of my concerns.

Knowing the workflow of folks at mozilla will be a huge help to people like me who are working on writing extensions because the road to writing a meaningful extension is not as straightforward as it could be.

In the end I dont want the above points to act as a downer in a software release post and would like to congratulate folks at Mozilla about the great work that they do.

> I'd prefer that applications used my RAM since unused RAM is just wasted electricity.

I... but... er...

This is probably the most ridiculous thing I've ever seen on this site.

You could write a two line C program that allocates all the memory on your system and then sits in a loop doing nothing, but then you'd be wasting even more electricity because your hdd would be reading/writing to your swap space. :|

> I'd rather have it used than have to fetch something from disk or the network.

Modern operating systems do use your free memory for cache. Linux will regularly chill at 90%+ memory usage (the majority being cache that can be freed instantly when other apps need it) and Windows has SuperFetch which does basically the same thing as I understand.

Some of us actually use our RAM for other things...

If I have 6GB of RAM and feel like e.g. editing a bunch of very large photos or videos without closing my browser, it's annoying to have that browser using 25% of RAM.

I really hope the persistent sync bugs get worked out soon. I'm told that the final, and big, fix was in alpha, which makes me think that it's not in 8.0. Soon, soon.

This is great, thanks for posting it. I manage my passwords with 1Password, and while they're good about pushing updates, I've never seen their Firefox extension marked as compatible when a new version of Firefox is released.

Which would beg the question why would I need an addon to make my current addons compatible with Firefox? Everyone knew that 6 weeks release cycle is going to break addons but they've decided to go with it anyway. I'd expect from Mozilla to make such decision after fixing the potential problems. I've been hearing this will be fixed soon but that date is yet to come.

Unfortunately, there is little documentation. I have looked into adding this for my extensions, and didn't manage to do it. Same documentation issues I had when I first started to port my extensions for Firefox to Fennec: pieces of documentations here and there, but nothing that clearly explains the differences between the 2 platforms.

That said, their extension framework is by far the most powerful of all browsers.

I just saw this being discussed today.. watch out for it if you have Firebug installed. There could be a performance regression because debugging is on even if you're not using it.

https://bugzilla.mozilla.org/show_bug.cgi?id=699528

Considering the whole reason they're doing this crap is to make it look like they're keeping up to date with Chrome, yeah, it kinda does make a difference.

Methinks they don't understand the concept of "Major release.minor release"

> There will be a new release every six weeks.

The corrollary to this being that releases are no longer news.

The most important reason for it was to keep new features from being stuck in limbo just because they happened to be part of the same branch as a feature that was taking a long time to be production ready. This is what killed us with 4.0. There were big and nice features in 4 that were actually ready more than 6 months prior to release, but they couldn't go because they were "full release" features and they were waiting for other features that were considered to be "must have" for 4.0.

With the rapid release cycle, when features start development, they have the ability to be easily configured out of the release product if they aren't ready for a particular version. We have a reliable release schedule that we expect to hit every 6 weeks. Obviously, we do our best to have as many new features as are ready be part of each release, but we don't leave people wondering when it will finally happen because we keep to the schedule and release the features that are ready.

Some of the biggest parts of this release were actually groundwork improvements to make the add-ons ecosphere more comfortable in the context of rapid release.

Mozilla is dropping the idea that some features are point releases and others are major version numbers. As a web browser there's always only been one web, so if a feature is ready then it's ready and it doesn't matter. Internally there's an actual product that other people might interact with -- and Firefox more than other browsers at that, as the traditional way to extend Firefox is to monkeypatch your changes in. But it was never quite so clear what a point release was there either, and it was getting in the way of advancing the core product.

I think simple dated releases would be an improvement (e.g., Firefox 2011.1, 2011.2, etc., or tag them based on month, this one being 2011.10, or 2011v2 or whatever.) But that's basically just a PR move. There's no longer anything really like a point release except in the case of regressions that demand an immediate fix (as happened with Firefox 7).

This is a valid, on topic, cogent post. Why is it getting downvoted?

Here is a post about where Chrome and Firefox are probably heading http://www.codinghorror.com/blog/2011/05/the-infinite-versio....

The biggest problem with updates are regressions. Since Firefox uses four release channels (since the introduction of the Aurora channel), most features get a decent amount of testing before being introduced to stable.

the whole point of this scheme is that there shouldn't be point releases. having them throughout the year is just as bad as any other point release.

the reason behind going to full numbers is that if you don't have the latest version, you have the wrong version. if there is a 13.1 and a 13.2, it makes it seem like people using 13.1 are still using the latest version. they aren't.