Mozilla 1.4 RC3 Is Out

zzxc writes "Mozillazine reports that the third release candidate for Mozilla 1.4 has been released. It is available for download from mozilla.org. Testing is encouraged to fix any bugs before the final release. No new features have been added to this release, though many bugs have been fixed. For more information, see the release notes."

Comments

Wow!

[sould]

Another mozilla 1,4 rc story...

Don't you have any sco news?

Firebird

[barcodez]

Firebird a subproject of Mozilla is a light weight version of Mozilla seems a lot better bet to me. Opens faster, has all the same features (such as tabbed browsing and popup killing) and seems to be more or less big free. Uses less memory too (at least in my primative tests).

Re:Firebird

[Max Romantschuk]

Firebird a subproject of Mozilla is a light weight version of Mozilla seems a lot better bet to me.

Actually, the current Mozilla Roadmap clearly states this goal: Deliver a Mozilla 1.4 milestone that can replace the 1.0 branch as the stable development path, then move on to make riskier changes during 1.5 and 1.6. The major changes after 1.4 involve switching to Mozilla Firebird and Thunderbird, and working aggressively on the next two items.

So actually, that's where we're heading :)

Re:Firebird

[Surak]

I've been using the the nightly builds since 0.5 was released. Mozilla Firebird is definitely *not* bug free (the releases are of course pretty stable), but I'm sure they could really use your help in testing it. :) The Firebird team has been really, really responsive about fixing bugs, and most bugs are usually fixed in the next nightly build.

My personal favorite feature is the ability to customize the toolbars. Admittedly, it's been a feature in Konqueror (my other favorite browser) since the beginning, but Firebird's toolbar customization is cooler as it supports drag-and-drop, while in Konq you have to use a Microsoft Office-style dialog box.

Re:Firebird

[havardw]

Then what about Thunderbird?

The big question is

[toddhunter]

When do we get RC4? It's so hard to sit through these RC2 and RC3 stories waiting to get to the exciting stuff.

Re:The big question is

[Anonymous Coward]

Mozilla 1.4 RC3 is expected to be the final release candidate. It incorporates the fix for one final GDI leak bug, and some minor stuff, but unless something completely unexpected will crop up, this very same release candidate will be rebranded as 1.4 final by the end of the week.

Re:The big question is

[ComaVN]

Erm, I'd say ANY release candidade is expected to be the final. Otherwise, why's is it a candidate? In my book, RC means you've fixed all bugs you're aware of at that time, so unless some others crop up, that's going to be the final version.

Re:The big question is

[Anonymous Coward]

But that's how it SHOULD work. A release candidate is a build which is expected to become the release. If you expect that it won't be the final build, it's a beta version, not a RC. You can trick users into installing RC builds a few times, but when they realize that the builds are just more "public betas", they will stop installing RCs and that will remove this important step from the feedback loop. Calling a build more mature than it is is also a bad move publicity-wise. People expect that the release is mostly the same as or only slightly better than the release candidate, so if you know about bugs which you deem inacceptable for a release build, don't make a release candidate before fixing them.

Re:The big question is

[The_Laughing_God]

I agree that the term "release candidate" *sounds* like it's expected to be a final release, and that may be true of some RC1 releases. However, it is a misleading characteriation of how the process is often meant to work (as indicated by the very fact that they're called RC1, instead of just RC -- RC2 is not unexpected) If all RCs were considered final until proven otherwise, we'd never see an announced "final release candidate", would we?

An RC1 gets tested more than mere alpha/betas, and that higher level of testing is *necessary* for a reliable release. If they didn't release RC1s with a few known issues, the actual release would not be as "final" (e.g. programs, chipsets, etc. often don't have fully stable function or performance on all features until the "A" version. Real world RC strategy is an up-front recognition of the realities that toss minor monkey wrenches in the best intentions of the engineering and testing departments.

An RC1 may be released with "minor" known bugs to help debug them, assess severity in the universe of real users, choose workable trade-offs, and enlist user aid in creating fixes for specific configs. Often a well thought-out Release Candidates contains chunks of testing/debugging code that is not meant to be in the final release. No matter what you might expect from the term, a 'Release Candidate" is usually not identical to the final release, even if it passes user testing with flying colors. This fact kills the simplistic assumption of many end users (we've all seen the rare release problems when debug code is removed from a stable RC)

The more intense RC testing typically turns up a handful of issues (nothing is bug-free). Some can be fixed cleanly once noticed. Others require testing beyond the abilities of the staff. Intermediate versions may be needed to work out the intricacies of the fixes across, say, all hardware and software configs. If intermediate versions are relabeled as "mere" betas, they won't get the testing that an RC gets, and the debugging could be delayed by months. If an RC1 includes 3 subtle issues, would you insist they all be "fully fixed" before an RC2 in 12-16 wks, or would you be happier if a RC2 with 'testing code' for one of more of those subtle bugs led to a fully functional RC3 (or 4) in 4-8 weeks?

That's why you rarely see post-RC1 'non-RC betas', and why we often hit RC3 or more: it's not that completely new issues arise in RC1/2/3, it's that downgrading from RC to v1.39b3 would have a psychological effect on the amount of deployment and testing, and project leaders know it

Release Notes way too bloated

[djpig]

I wish they would only put in the release notes the changes between RC2 and RC3 (and not between 1.3 and 1.4). Every time I read the release notes for the different RC's I get a strong dejavu. Must I really begin to diff them?

Re:Release Notes way too bloated

[cerberusss]

Must I really begin to diff them?

No, just scroll down. After the release notes, there's a section "New additions to the Release Notes".

Re:Release Notes way too bloated

[spoonyfork]

No, just scroll down. After the release notes, there's a section "New additions to the Release Notes".

Here's the documented "changes". Very subtle.

Mozilla 1.4 RC 2

• Mozilla 1.4 for Linux requires Sun J2SE v 1.4.2 Beta to run Java applets
• JavaScript access to Flash does not work on Linux Mozilla 1.4

Mozilla 1.4 RC 3

• If you're using Linux binaries compiled by mozilla.org then you will need Sun J2SE v 1.4.2 Beta or the Blackdown JDK 1.4.1 compiled with GCC 3.2 to run Java applets.
• If you're using the Linux binaries compiled by mozilla.org then JavaScript access to Flash will not work.

Re:Release Notes way too bloated

[mlefevre]

That's just a clarification of the stuff about Java, it's not indicative of a change between releases.

As you might expect, the only changes between RC2 and RC3 were a bunch of bug fixes, and those don't get mentioned in the release notes - they're release notes, not a changelog.

For the sake of the database, I won't post bugzilla links, but the list of fixes since RC2 is as follows:

88393 (Mac) Check in a high-resolution application and document icon ...
140357 (All) Backspace deletes text formatting,TypeInState should be s...
189429 (All) strict javascript warning in mail3PaneWindowCommands.js
197379 (Mac) file:// URLs from CFM mozilla don't work with Mach-O mozilla
199443 (PC) leaking GDIs when table cell contains an image, and text...
205360 (Sun) libxpcom.so depends on non-existent libiconv.so
206271 (PC) News Messages being marked as read automatically
206668 (Mac) [Mac OS X classic theme] context menu only work on frontm...
208560 (PC) P3P summary only works once
209033 (Mac) FIXE (Shockwave, Flash, ?) all typed letters (from kbd) appear...
209354 (All) typeaheadfind causes major memory leaks

Re:Open source == command line

[Klimaxor]

Mozilla is command line? wow, i must be on something to see all those pretty colors.

RedHat 7.x RPMS?

[guacamole]

Will the mozilla project provide Mozilla 1.4 final RPMS for RedHat Linux 7.x? It seems like they have discontinued them for all of there 1.4 beta and RC releases.

Re:RedHat 7.x RPMS?

[whovian]

I had long noticed this also. You used to be able to go into their ftp directory and hunt down the RH7.x version.

Actually, it's nearly trivial to install their tar.gz packages. It involves unpacking in /usr/lib. Then copy your plugin modules and links to modules (read the latter: j2re) into the new mozilla directory. You might have to modify a couple of the permissions on the mozilla or its subdirectories, and maybe put a link in /usr/local/bin. If you are a little careful, it is not bad at all.

> cd /usr/lib
> ls -ld moz*
lrwxrwxrwx 1 root root 14 May 31 06:20 mozilla -> mozilla-1.4rc1
drwxr-xr-x 5 root root 4096 May 31 06:18 mozilla-1.3
drwxr-xr-x 10 root root 4096 May 31 06:24 mozilla-1.4rc1

Firebird

[khalido]

When Moz 1.4 final is released, will Firebird then be based on 1.4, or will it remain based on the Moz 1.3 codebase? Also Moz needs better default fonts still. I had to install the vera fonts to make it look decent. In IE the fonts looks so much better. I know, thats becasue its using the fonts in windows and what not, and moz just can't include anti aliased fonts that won't work on systems x,y and z, but there needs to a system with prebuilt decent fonts. Moz is now so much better than IE, but default Moz on linux looks like a POS. Yes yes I installed truetype fonts now its fine but a lot of people don't know how to do all that. All this is becasue I had installed linux for a non computer person, who updated mozilla and then was stuck with the default fonts.

Mozilla -- Who compiles every release?

[Anonymous Coward]

Ok, now, don't get me wrong here.. I like Moz, so this is not supposed to be a flamebait:

Just how many of you download and compile every single version of Mozilla that's mentioned on /.?
It takes time to download (due to the /. effect) and it takes time to compile. Very few bother to go through this process for every release.

Is it really neccesary to mention every RC's here, or am I just being picky?

Time for a New Releases Section?

[beders]

How about a new releases section to cream off most of these storys? Keep the main page for "stable" releases

Why, I do! ;)

[SalsaDoom]

Absolutely I do.

Why? Because installing and playing with new software turns out to be a pretty fun replacement for games. Games are pretty sweet, Linux has a growing little number of them. But I mean...

I can't be the only one who finds updates fun, can I? ;)

For Slackware Users...

[Lispy]

there is a great Gnome-Project wich adapts pretty fast to new Mozilla releases and ships with antialiased fonts (I didnÂt like them in the beginning but am an addict now ;-) called Dropline-Gnome. I keep installing this for my newbies along with Slackware wich I prefer as a Newbie-Distro for itÂs clear structure. From this day on I stopped worrying about fonts in Mozilla. Most Gnome Apps use these fonts so it provides a consistent look&feel, too.

But last time I tried Firebird I realised the problem was still there. The defaults are ok...but not a beauty. Well, if for non-slackware users I guess Ximian-Gnome ships with antialiased fonts as well...

343 bugs.

[leuk_he]

If i check bugzilla there are currently 343 bugs open that are:

blocker or critical
and
assigned. (i did not select new 1441 bugs because they still contain dupes, or bugs that need te be cleaned).

That is a LOT! and they want 1.4 to be the next stable release for a longer time. I think it is still time forsome bugsquasing before releasing is.

LotÂs of these bugs are cross platform bugs (example:it wont build on true64,aix)

One bad bug i want to note is:209896
Bug: mozilla crashes if upgraded from 1.3.1 to RC2.
workround: uninstall first.

Yeah right: so every bug somebody calls (on some generic internet forum) the response will be: delete you mozilla directory first, then reinstall.

why i won't switch to lightweight firebird

[heymjo]

in mozilla, type something in the adress bar , press down key and you get "search google for" , press enter and boom results are there.
I don't want to switch to a different search field or even set up parameterized keywords to do this.. Google search with 2 keys (down + enter) is for me the killer feature as i do this well over a hundred times per day

Re:why i won't switch to lightweight firebird

[Gleng]

There's an extension for Firebird which adds a google search box to your toolbar.

Just go to Tools->Options->Extensions->Get New Extensions.

Have a scroll through the page, there's quite a few handy extensions that you can download, and not many of them are over a few KB.

Re:why i won't switch to lightweight firebird

[riflemann]

Keywords are a problem for this?? I find it a heck of a lot faster than scrolling down lists...

Create a bookmark with the follwing URL:

http://www.google.com/search?&q=%s

In the bookmarks manager, go to the properties of this bookmark, and set the keyword to 'g'.

All you have to do now for a google search is to type "g [search term]" in the address bar and hit enter. (without quotes)

Plus the keyboard travel for typing "g " is much smaller than for the arrow keys.

Moz 1.4 problems on install. Moz crashing.

[Futurepower(R)]

I have not been able to get Mozilla 1.4 to install on one machine (with a lot of email). I installed Moz 1.4RC3 over 1.3.1, and I get a Windows program crash message, offering to send Microsoft data about the crash.

The release notes said to install 1.4 in a new directory, but I spent hours teaching Moz to store email in a folder other than the default. I don't want to go through that again. Moz gives the option to install in a folder other than the default, but does not make it easy.

I re-installed 1.3.1 over the bad 1.4, and it works, no problem. The version I had downloaded does not say 1.4 RC3, just 1.4.

On another machine, I had no problem installing 1.4. Both are running Windows XP, SP1.

I am anxious to begin using 1.4 because I've had many problems with 1.3.1 crashing after many instances and many tabs are opened, and some are closed. The crashing seems associated with Windows XP's limit of 21 programs open at the same time. (After that, the program list is displayed in a disordered fashion. That "feature" seems to have been put in by Microsoft to discourage people from opening a lot of programs.) Mozilla's crashing seems to corrupt Windows XP, too, so that a reboot is required to restore full functionality.

When either Moz or Firebird crashes, all instances and all tabs crash. It would be great if instances were completely separate from each other. I can buy more memory, if needed, much easier than I can repeatedly lose work.

I've seen the same crashing of Moz 1.3 under Linux with many instances and tabs open, when some tabs are closed. I reported the problem, and there was speculation that there was stack corruption. I hope this is fixed in 1.4.

Moz/Firebird are not perfect, but they are by far the best, in my opinion.

mozilla link

[minus_273]

here is an intersting website by a moz developer (which sadly is not a parody but the REAL thing.. this guy is dead serious). http://mithgol.pp.ru/Mozilla/

Both sad and funny at the same time.

Really...

[nhaze]

I don't think we get enough Mozilla RC updates. Maybe we can start getting updates letting us know the status of nightly builds.

Triple-clicking the location bar in Windows

[ortholattice]

First, there are two pref's I have set up in user.js so that the first click that brings focus to the location bar selects all. Without this I'd go nuts, since I like to press "s" and have slashdot immediately appear, "sc" for sciencenewsdaily, etc.. This solves half of my problem.

user_pref("browser.urlbar.clickSelectsAll", true);
user_pref("browser.urlbar.clickAtEndSelects", true);

Now for the heart of my complaint. In Mozilla 1.2 and before, once you had focus on the location bar, double-clicking the location bar selected all, just as it does in Internet Explorer and numerous other Windows apps that have boxes for file names and URLs.

In Mozilla 1.3, the behavior was changed to: double-click selects a "word", and triple click selects all. The philosophy being, the location bar is like a mini text editor, so it should work like an editor. See this Usenet thread. (Frankly, the "word" that is selected after double-clicking has never been of much use to me.)

The problem is, I think (this is my theory) there is something fundamental in Windows where "triple-click" is not a real operating system event, like double-clicking, so some other kludge is used to time the clicks. Or maybe Windows XP or the mouse driver is just broken, I don't know. But anyway if I have the mouse speed set for fast clicking, I can't get triple-click to work at all. If I set the mouse speed slow, I can triple-click as long as I click not too slow and not too fast, but you have to get the timing just right. Half the time it seems I get it wrong and have to try again. And I hate having to set the mouse speed slow because that screws up what I'm used to with other apps.

I know this isn't the right forum for bug reports - I've been meaning to study this problem in more detail, logging Windows events and times so I can make a convincing case and write up a useful bug report, but time has just been slipping by and I'm afraid the final release (an important one from what I hear) will happen before this can be properly addressed. I will try though, I promise. :)

Am I just being fanatically nitpicky, or does this bother anyone else? (Well, at least I got it off my chest...:)

Re:Triple-clicking the location bar in Windows

[Briareos]

Why click in the URL bar anyway? Just hit CTRL-L and type away, since you're going to use the keyboard anyway...

(CTRL-L also selects the whole URL bar, so you can start typing right away.)

np: Senor Coconut - Musica Moderna (El Gran Baile)

Spellcheck in text boxes?

[Joe Tie.]

I remember hearing talk quite a long time ago about plans to allow spellchecking of textboxes, such as on slashdot, from a menu in mozilla. Anyone know if this feature is still planned once the dictionary is offcially added in? I'm using Mozilla for the moment, as the spellcheck feature is broken in the cvs build of kde right now. And for the most part I actually prefer it to konqueror, but that spellcheck is to me the killer feature which dictates which I use.