Mozilla 0.9.4 Released

asa writes: "Lots of bug fixes (1,467 at last count) since 0.9.3 including the ability to disable the JavaScript window.open() method during page load and unload events. You can find more information on what's new at the release notes and mozillaZine."

Comments

Mirror

[pete-classic]

I've got the new release mirrored at ftp://nerf-herder.net/pub/mozilla

-Peter

Re:Mirror

[Gerv]

The mozilla folks really ought to put up MD5s with the release.

Why? If they can tamper with the releases, they can tamper with the MD5s.

Anyway, the standard disclaimer we put on all releases applies: "If it doesn't melt your hard drive and send your tax evasion plans to the IRS, consider yourself lucky."
Gerv

Re:Mirror

[pete-classic]

Why? If they can tamper with the releases, they can tamper with the MD5s.

For mirrors. You get the MD5 (AFAIR, 128bits, conceiveably double that when including the filename ;-) from the "official" site and use it to verify that the bins on the mirror haven't been altered.

-Peter

Re:Mirror

[jesser]

I think bugzilla runs into database speed problems before it runs out of bandwidth, but I could be wrong.

Re:Mirror

[cymen]

For mirrors. You get the MD5 (AFAIR, 128bits, conceiveably double that when including the filename ;-) from the "official" site and use it to verify that the bins on the mirror haven't been altered.

And ideally when /. posts stories like this they could include the MD5 in with the story. Of course you'd have to trust /. though :).

Re:fp

[Gerv]

Read the release notes. Arabic language support should work.

Gerv

Hmm...

[Scoria]

Just six more releases until they'll be forced to call it 1.0! Wait... I wonder if they can do 0.9.8.1, 0.9.8.2, etc. and delay it a couple of more years.

Bad thought there. Eep.

Oh well, thanks anyway to the Mozilla team for their hard work, especially for the work they've done this week after the tragedies in our country. I know that most people's productivity levels are null, lately.

Re:Hmm...

[Gerv]

Actually, it goes 0.9.9, 0.9.10, 0.9.11,...

Gerv

Re:Hmm...

[Salsaman]

Of course...they are reaching 1.0 exponentially :-)

Re:Hmm...

[mutantcamel]

Agree totally, they're doing a fantastic job. my first brush with linux was suse 6.2 and now I get updates courtesy of coverdisks, and I've been impressed with every update, well done, you chaps.

Re:Hmm...

[Eil]

Erm, wouldn't that mean that 1.0 should have been hit on the 5th iteration? :P

Re:Hmm...

[Genyin]

people laugh... but that's seriously how it works. The nice thing about the dots are that they cleanly separate the parts, so you can actually say 0.9.10.

Mozilla is getting quite close to being stable enough to call 1.0, though...

Re:hmm...

[zmooc]

libc-2.2.4.so

Looks like it...I will try to build it myself. Thank you for your advice.

Re:Hmm...

[jovlinger]

actually, if it were exponentially, it would be the zeroth generation.

And I suspect he meant monotonically asymptotic, but that is a bit of a mouthful.
</pendant mode>

Actually...

[bconway]

Disabling window.open has been around for a couple of releases now, it's just not the most straightforward thing to enable. I was most pleased to find that hitting enter after filling in a form will actually submit a request everyplace I tried it, assuming that's the intent of the form (i.e. a search engine). This seemed to be a hit-or-miss thing in previous releases.

Re:Actually...

[zachlipton]

This used to be a huge issue and something that prevented lots of people from using Mozilla before. Thank you to those who have fixed this!

Re:Actually...

[NutscrapeSucks]

The classic Netscape thing to do was only submit the form on Enter is when the form contained only 1 text field. That seemed to be how Moz worked up to 0.93.

IE generally submits all forms on Enter, including sometimes when you are trying to type in a textarea :P I haven't got this installed yet, but it would be really interesting if they went with IE's convention.

Re:Actually...

[jesser]

IE generally submits all forms on Enter, including sometimes when you are trying to type in a textarea :P I haven't got this installed yet, but it would be really interesting if they went with IE's convention.

One nightly Mozilla build actually did have this bug -- hitting enter in a textarea would submit the form. See the last few comments in bug 22526, "Enter in text input submits form only if there is exactly one text input".

Re:Actually...

[NutscrapeSucks]

I like the idea of making IE behavior an optional pref. I've had users complain that forms are broken because they don't submit on Enter in Netscape, and then you have to explain that it's a browser issue, not your form.

Re:Actually...

[jesser]

I've had users complain that forms are broken because they don't submit on Enter in Netscape, and then you have to explain that it's a browser issue, not your form.

And making it a pref would fix that how? If a user doesn't think to try hitting the "submit" button when enter doesn't work, they're even less likely to look in prefs for a way to make hitting enter work.

IE's behavior (enter submits a form except in a textarea) makes sense, and that's what Mozilla does now.

Looking good

[boaworm]

Mozilla is turning out to be a really good browser nowdays. I missed a few features in 0.9.3 though, mostly that it tends to crash while at java-intense pages, as well as encryption.

Hopefully these things have gotten better, it is quite annoying when the browser crashes :-(

If Mozilla is going to be able to compete with the major browsers, it (IMHO) has to be a lot more stable. I can cope with a page being rendered badly, but not with a browser crash. IE is still a lot more stable. Or.. perhaps it is just bad Java Runtime integration ?

Thanks anyway Mozilla team, i'm off to the download zone :-)

Re:Looking good

[maggard]

If Mozilla is going to be able to compete with the major browsers...

What other major browsers? Opera? Lynx? The legions of other 1%'ers?

As far as most webfolks are concerned there's IE for Wintel, IE for Mac (they've different code bases and behave very differently), Netscape et al v.4x, Netscape/Mozilla et al v.6x then generic text-browsers for ADA compatibility. That leaves Netscape/Mozilla as one of the two major names and the rest lost in the "other" catagory*.

*Yes lots of browser-partesians will howl at this but for most web sites the vast majority of browsers hitting them regularly are IE or NS. No comment on quality or anything else, just reading the logs.

Re:Looking good

[chabotc]

From my experiances experimenting with java, it is mostly due to the fact that mozilla uses a java2 envirioment. (the jre1.3.xpi and sun java and blackdown java plugins for mozilla are all 1.3+ based).

Most of the applets you will find on the web will still be java 1 based. (This is what IE ships, duh)

There are some 'known' problems, leaking resources, threads and not relaunching applets when a java 1 applet is loaded in a java2 VM

big miss feature if you ask me, but in both sun's bug DB, and mozilla's bugzilla, its gotten marked as 'solved/wontfix', so don't hold your breath to see it resolved ;-)

Re:Looking good

[Lord Ender]

You get that info from looking at the logs? The logs are WRONG! Opera by default identifies itself is IE. There is a huge percentage of people who use Opera primarily. For instance, four of my non-techie friends use it all the time. My girlfriend uses it. I just showed them the basics of how to use it, then the cool feature only Opera has (like gesture navigation) and they were hooked. When reading the logs, remember this: a large percentage of those IE hits are really Opera .

Re:Looking good

[afflatus_com]

Just as a mention: in log analysis, many site owners (and software--even top-of-line ones like MediaHouse) overlook the fact that since AOL uses a multi-proxy management system, there is usually 6-11 separate unique IP address entries in the log from the visitor using AOL. Therefore, each AOL browser entry (which gets grouped under MSIE-powered browsers) is marked as 6 to 11 times, instead of just once. If AOL does switch to Mozilla at a later date, it will make a hefty swing in the browser percentages, as reported by what browser unique visitors are using.

Re:Looking good

[the way]

> If Mozilla is going to be able to compete
> with the major browsers...

What other major browsers? Opera? Lynx? The legions of other 1%'ers?

As far as most webfolks are concerned there's IE for Wintel, IE for Mac (they've different code bases and behave very differently), Netscape et al v.4x, Netscape/Mozilla et al v.6x then generic text-browsers for ADA compatibility. That leaves Netscape/Mozilla as one of the two major names and the rest lost in the "other" catagory*.

I'm so glad you can speak for "most webfolks"...

Although unfortunately I can only speak for myself, I can certainly say that we see quite a bit of Konqueror usage at our site. Nothing like competing with IE, or course, but certainly up there with NS/Moz. It is the default browser on a number of major distributions nowadays and has a similar feature-set to Moz and IE, so I think it's fair to call it a 'major browser'.

Re:Looking good

[fault0]

Yeah same here.. we've see konqueror usage go up with every KDE release to the point where it is about the same as Netscape/Mozilla. But still, nothing beats IE in terms of usage.

Re:Looking good

[krmt]

Your mileage may vary on this one.

For me, Mozilla is a hell of a lot more stable than IE. IE crashes on me not infrequently, and it usually means a reboot.

The fact that I use Mozilla a ton under linux and IE in those relatively infrequent times I boot in to Windows really swings it for me.

Re:Looking good

[Malcontent]

You'd think they would at least alter the user agent so that MS does not get credit. Why not say AOL borwser 1.0 or something. It would act as an advertisement at least.

Re:Looking good

[i-sob]

I know that my new browser of choice (OmniWeb) by default identifies itself as IE as well...I use Opera quite a bit too.

I think it's very important for "alternative" browsers to identify themselves correctly -- it's important to keep webmasters from coding specifically for IE or NS. (I masquerade my OmniWeb as Mozilla)

Re:Looking good

[uid8472]

OmniWeb, by default, masquerades as Mozilla for HTTP, thusly:
GET / HTTP/1.0
Connection: Keep-Alive
User-Agent: Mozilla/4.5 (compatible; OmniWeb/4.0.5; Mac_PowerPC)
[...]

Then there's a preference for faking the browser info for JavaScripts, which seems to be separate.

Re:Looking good

[Yokaze]

That would make it hard to produce agent-specific webpages.

Re:Looking good

[BZ]

Evey browser I know has "Mozilla" in the UA string. This includes IE. That's not really masquerading as Mozilla....

Re:Looking good

[Homewrecker]

a large percentage of those IE hits are really Opera

Pfft, I can accept some percentage, but large? Now you're just being ridiculous in an attempt to downplay the dominance of IE.

Re:Looking good

[sciencewhiz]

Here are the appropriate lines from one of my webserver's logs with hits from Opera when it is identifying at Opera, Mozilla 5.0, Mozilla 4.76, Mozilla 3, and MSIE 5.0, respectivly:

24.169.XXX.XXX - - [15/Sep/2001:10:43:14 -0700] GET / HTTP/1.1" 200 511 "-" "Opera/5.12 (Windows 2000; U) [en]"

24.169.XXX.XXX - - [15/Sep/2001:10:43:44 -0700] "GET / HTTP/1.1" 200 511 "-" "Mozilla/5.0 (Windows 2000; U) Opera 5.12 [en]"

24.169.XXX.XXX - - [15/Sep/2001:10:43:54 -0700] "GET / HTTP/1.1" 200 511 "-" "Mozilla/4.76 (Windows 2000; U) Opera 5.12 [en]"

24.169.XXX.XXX - - [15/Sep/2001:10:44:03 -0700] "GET / HTTP/1.1" 200 511 "-" "Mozilla/3.0 (Windows 2000; U) Opera 5.12 [en]"

24.169.XXX.XXX - - [15/Sep/2001:10:44:12 -0700] "GET / HTTP/1.1" 200 511 "-" "Mozilla/4.0 (compatible; MSIE 5.0; Windows 2000) Opera 5.12 [en]"

Note that in all cases, the string "Opera 5.12" appears.

Re:Looking good

[Malcontent]

No it would not. You'd just have to accomodate one more browser. There are enough people on AOL to make a difference so you'd do it. Either that or you'd just stick to the standards which you ought to be doing anyways.

Really I think it's a good thing for AOL. Doing something seemingly minor while dramatically effecting the MS share of the browser market. Also it would allow them to change browsers easier too.

Re:Looking good

[Yokaze]

Sticking to the standards include the RFCs. Now take a look at RFC1945 especially
Section 10.15 User-Agent:

This is for statistical purposes, [...] , and automated recognition of user agents for the sake of tailoring responses to avoid particular user agent limitations.

Now take a look at the note to this recommendation.

Note: Some current proxy applications append their product information to the list in the User-Agent field. This is not recommended, since it makes machine interpretation of these fields ambiguous.

So this field is there, exactly for this reason, to provide user-agent tailored pages. Lastly, the main-purpose was not to differentiate between Netscape and IE, but between different HTML version capable browsers. If one intends to provide even more specialised pages, it'll be his/her decision and problem, not yours.

Re:Looking good

[dotmaudot]

There are some 'known' problems, leaking resources, threads and not relaunching applets when a java 1 applet is loaded in a java2 VM

Call me stupid (I am used), but neither with 0.9.3 nor with 0.9.4 on Windows 2000 I am able to see Java applets. And yes, in Preference/Advanced Java is enabled. Could someone give me a hint?

ciao, .mau.

Re:Looking good

[Malcontent]

Once again.

AOL should chage their agent string. If nothing else they can overnight recude the market share of IE browsers and the statistics would be used to advertise the popularity of AOL. It would be an excellant marketing opportunity.

As for the standards I meant the HTML standard, the DOM standard, the javascript standard. Stick to those and let the chips fall where they may.

Re:Looking good

[maggard]

As some folks have pointed out most browsers mis-identify themselves as one of the big two in order to avoid being condemned to the "other" catagory.

This is well understood by anyone who trackes browser visits. This is not news, this is not a suprise and it doesn't change the numbers. Often it's possible to identify a browser by exactly what string it does give, othertimes a simple javascript is run or other behaviors are checked.

Any way you cut it; reading logs, actual testing or running surveys the numbers all come out it's IE, NS, then everything else, with the everything else ones pretty much lost in the noise.

Finally for those who continue to bleat out "Just write to the spec!" that'd be great if the browsers actually met the specs but they don't. Furthermore as folks are paid to work in the real world and produce results that are actually usable on their client's desktops they must use the tools available and that means a limited collection of very ideosyncratic browsers.

All of the browsers have their own quirks and strengths and most folks just try to write generic pages most of the time. When something is needed and is browser-specific then work-arounds are made and usually they cover 90%+ of the actual users of a major site.

It's always sorta funny and sorta sad to see how many /.'ers think that everyone else is stoopid and that they have unique insights into every situation. Again - yes browser misreporting is well known and is pretty much compensated for and for sophisticated stuff writing to the specs will get you as much trouble as often as not doing so will. Most web folks would LOVE to write to the specs, if only everything implemented them properly & consistantly!

Wonderful release!

[�laC|n]

Mozilla's back and better than ever!

It's now even more stable than before.. Seems like they finally are getting near the magical 1.0..

Thank god

[pheph]

Seriously... For some reason 0.9.3 was considerably less stable than 0.9.2 on both my linux and windows systems... I'm glad to see so many bugs fixed. This is my primary browser and each improvement to it is an improvement to my LIFE.

Re:Thank god

[G00F]

yes, it was crashign for me all the time, but there was a few things that could have done it. Like not having aclean install . . . .

Mozilla?

[gordzilla]

I switched to 0.9.3 permenantly but does anyone know what to put in the "helper apps" to
get RealPlayer8 working automatically? I'm not having much luck! Anyhow looking forward to giving 0.9.4 a try.

Re:Mozilla?

[ILoveMandrake]

If you are using linux & have Netscape do this:

cp /usr/lib/netscape/plugins/* /home/username/mozilla/mozilla/plugins/

or wherever you installed mozilla too... good luck

speed

[moronic1]

I havnt used a mozilla release for sometime(been a little busy to grab them everyday..) and WOW! talk about a speed... seems much faster

Wow!

[zachlipton]

Wow, what a great release! I think that 0.9.3 really is a key step in the right direction for 1.0. See http://www.mozilla.org/roadmap.html for more details on the roadmap and plans for 1.0.

Also, as a mozilla developer, I would like to thank all those who have joined the project recently and done something to help. Even if you cannot code, there is still lots that you can do. I urge you to download 0.9.4 or even better, a nightly build, and to look at http://www.mozilla.org/start, http://www.mozilla.org/qa/help, and http://www.mozilla.org/get-involved.html. There are many things that you can do to help which will help get 1.0 out the door sooner and better.

Re:Wow!

[Gerv]

Wow, what a great release! I think that 0.9.3 really is a key step in the right direction for 1.0.

Has someone been cutting and pasting out of their "Slashdot comments" file? ;-)

Gerv

Re:Wow!

[zachlipton]

As you can see ladies and gentlemen, when someone triages as many bugs as I have, what number attached to a release has lost all meaning.

Oh and yes, I do think that 0.9.4 is also a key step in the right direction ;)

How to manage popup windows in the new Mozilla

[davidu]

Ok folks, here is a really cool feature: The Ability to manage, on a site by site basis, which sites can give you popups and which can't. A very effective way to manage pop up ads. Here's how:

No POPUPS whatsoever:
user_pref("capability.policy.default.Window.open", "noAccess");

But...if some sites need popups, make a zone for them like this:
user_pref("capability.policy.strict.sites", "http://www.evil.org http://www.annoying.com");
user_pref("capability.policy.strict.Window.alert", "noAccess");
user_pref("capability.policy.strict.Window.confirm ", "noAccess");
user_pref("capability.policy.strict.Window.prompt" , "noAccess");
... you get the idea....

It is very cool, and there is a lot of scripting and other trickery you can do with these prefrences.

Btw, this is all from: Configurable Security Policies

-David

Re:How to manage popup windows in the new Mozilla

[zachlipton]

There is a bug on this in bugzilla though I forgot the # offhand. I do know that it is in the works and will be done as soon as possible.

Zach

Re:How to manage popup windows in the new Mozilla

[iturbide]

Cool. Just what we needed.

But is it in the interface? I'll be damned if I'm going to talk my mother through this.

Re:How to manage popup windows in the new Mozilla

[abischof]

UI for controlling popups is bug 75371. Feel free to vote for the bug, if that issue is important to you.

Re:How to manage popup windows in the new Mozilla

[Spy Hunter]

Actually, what is in Mozilla now is much cooler yet, which is the ability to disable 99.9% of advertising popups while letting 99% of wanted popups through, with no user intervention necessary ! No need to maintain a list of sites that need popups to function. It disables popups during page load and unload, but lets through popups that happen due to an actual mouse click.

Of course, if this feature ever gets widespread use we'll just see javascript links that open up advertisements in addition to their targets, but that won't happen unless IE gets this feature, which is unlikely. So download Mozilla and free yourself from evil automatic popups!

Re:How to manage popup windows in the new Mozilla

[Liquor]

I presume the bugzilla report is for adding UI configuration to the existing capability? Or is it broad enough to cover the new configuration item?

Meanwhile, I'd also like to be able to prevent launching a new window in an attempt to display an image from a site that has already been configured as 'never load images from this site' :)

But the new configuration (when I get the phone line freed up for a few hours) should be an improvement in that it shouldn't NEED to have it's configuration tweaked for every new and annoying site.

Re:How to manage popup windows in the new Mozilla

[Spy Hunter]

Follow the link "disable" at the beginning of the article. You must be using the new version.

Re:How to manage popup windows in the new Mozilla

[iso]

I've put in my vote. Still, it's remarkable that people would even consider this a "feature" when you have to edit obscure configuration files to make it work. I've been a Linux user since 1995, but I think this is a good example of the major reason I've now switched to Mac OS X: Apple wouldn't dream of making a feature that didn't have a UI interface whereas with Linux it's the norm.

Apart from Linux, I've been using UNIX for over 15 years and quite frankly I'm done with text files. I've put all 10 of my votes for Mozilla bugs into UI bugs because in 2001 I shouldn't have to be editing text files if I don't want to. It's amazing that these features have been in Mozilla for months but still don't have even a rudimentary graphical interface.

- j

Re:How to manage popup windows in the new Mozilla

[Ms.Taken]

Here's how to eliminate popups in Mozilla 0.9.4:

Add this line to user.js:
user_pref("dom.disable_open_during_load", true);

If you don't have user.js, create it in the directory where prefs.js resides.

Re:How to manage popup windows in the new Mozilla

[drfireman]

"Apple wouldn't dream of making a feature that didn't have a UI interface whereas with Linux it's the norm."

I'm not a particularly savvy Mac user, but over the years I've made dozens of modifications to various system features via resedit and by downloading little hacks. I'm sure there have been hundreds if not thousands of such undocumented features. With OSX I'm sure there are easily thousands of features you can alter that don't have a nice interface. So I don't think your statement is well-supported. It would be much more correct to say that Apple wouldn't dream of documenting a feature that didn't have an interface. I don't think it's productive to fault Linux software for not taking this same shield-the-user attitude. People may be trying to make Linux more idiot-friendly, but I have yet to hear anyone suggest that it should also be made less geek-friendly.

Re:How to manage popup windows in the new Mozilla

[vrt3]

I don't really mind editing text files; my problem is finding the documentation what, where and how to change them. I've tried before to find some docs on Mozilla configuration, but unsuccesfully. Can anyone throw me some pointers?

Re:How to manage popup windows in the new Mozilla

[damiam]

http://www.mozilla.org/unix/customizing.html has good information (and most of it isn't UNIX-specific).

Re:How to manage popup windows in the new Mozilla

[vrt3]

Great, thanks!

Re:How to manage popup windows in the new Mozilla

[LordNimon]

This doesn't exactly work like intended. It also kills popups that are created when you click on a Javascript link. In other words, it completely kills the ability for Javascript to create a window. Unfortunately, the includes more than just "popups".

Re:How to manage popup windows in the new Mozilla

[edremy]

but I think this is a good example of the major reason I've now switched to Mac OS X: Apple wouldn't dream of making a feature that didn't have a UI interface whereas with Linux it's the norm.

So tell me: how do I move the Dock to the right side of the screen or make the disclosure triangles blue if the app is frontmost?

Both can be done, but only if you know the magic incantations to feed to a command prompt. Apple really needs to work on the Dock, although I understand pinning will at least be an option in 10.1

Eric

Re:How to manage popup windows in the new Mozilla

[LordNimon]

Can you tell me where this feature is available? I can't find it anywhere. The above prefs.js settings don't work, because they disable javascript-opened windows on a mouse click, which is a bad thing.

Re:How to manage popup windows in the new Mozilla

[Spy Hunter]

The article itself contains a link which apparently no one read that explains what you have to put in your prefs.js file.

user_pref("dom.disable_open_during_load", true);
That should do it for you, if you're using Moz 0.9.4.

I love it!

[Brazilian Geek]

I've been using Mozilla's daily builds as my standard browser since M18 and as my email client since 0.8 and I've got to say that I love it - yes, it is a memory hog but I have more than enough memory to give a fsck.

I've been trying to evangelize the users from my work place into using Mozilla since 0.9.2 and so far I've managed to get 10 out of 90 to switch (from Netscape 4.75 of course, IE is a no-no acording to company security policy).

Way to go Mozilla Team - it gets better every single day, congratulations!

Re:I love it!

[BrookHarty]

Not sure what company you work for, but IT is almost m$ bought and paid for. Exchange, IE, IIS, Office2k, etc..

But in Operations (aka we make the money), we use mostly Solaris, Apache, Netscape 4.77 and Mozilla. Some of our admin tools wont even work on IE5. (Havnt tested IE6 yet...)

Re:I love it!

[javacowboy]

I have managed to at least partially convert my System Admin, who just happens to sit beside me at work. He runs FreeBSD and stubbornly stuck with Netscape, before giving it a try today :)

I use the browser AND mail client as my default applications, and they just keep getting better and better.

IE may have slightly more features and functionality, but it has stagnated as a product and may in fact be declining due to M$ removing plug-in and Java support. They have done this because they believe they have a monopoly in browsers, but they're wrong. All Mozilla has to do is add a few new features (expanding upon ideas like disabling pop-up windows) on top of what's already there, and it will be a markedly superior browser. I think this project has enough momentum to continue long past version 1.0, and I believe there's enough creative energy on the development team to come up with a long list of new features to leave IE in the dust once and for all.

AOL seems to be very aggressive in promoting Netscape 6.1. For instance, there's a Netscape banner at the top of cnn.com. If AOL can, at the very least, convince Netscape 4.7 users to upgrade to 6.1, it will have done the Mozilla project a huge favour.

The browser wars are far from over.

Re:I love it!

[Homewrecker]

The browser wars are far from over.

Around Slashdot, maybe, but in the real world the horse is dead. One is hard pressed to find non-nerds using anything but IE. Why? Nothing else offers stability or features that make converting worth the time. Also, IE is by far the most responsive browser on the windows platform and, from my experience, that's one of the biggest things that people like. Everything else feels sluggish in comparison.

aha!

[Requiem]

Excellent. I may now pr0n from work and not have to frantically close windows at the last minute.

Thank you, Mozilla team. My raging codpiece salutes you.

Proxomitron

[BrookHarty]

...the ability to disable the JavaScript window.open() method...
You might not want all pop up windows disabled, only windows with ads.

Check out Proxomitron

It will has rewrite the http stream, so you can rewrite headers, html, cookies, etc. For both incoming and outgoing. Also make Mozilla reply as "User-Agent: Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1)". Its more powerfull than junkbusters, and has more features and filters(all editable)

Re:Proxomitron

[Gerv]

Also make Mozilla reply as "User-Agent: Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1)".

That's a bit silly, isn't it? You'll get served IE DOM content, and it won't render correctly.

Gerv

Re:Proxomitron

[BrookHarty]

It comes in handy on a few sites, and the option is easy to toggle. In fact most sites that do bitch about a non-ie browser work just fine after I fake the user-agent.

Re:Proxomitron

[Phroggy]

Also make Mozilla reply as "User-Agent: Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1)".

Please don't do this. It just makes life harder on web designers. How can we optimize our HTML code to render correctly in your browser, if you lie to us about what browser you're using?

Of course, there may be a few cases where it's necessary to do this temporarily, on a per-site basis, but please don't do this long-term.

Re:Proxomitron

[twoflower]

It just makes life harder on web designers.

No, you're just making life harder on yourself. See below.

How can we optimize our HTML code to render correctly in your browser, if you lie to us about what browser you're using?

Don't do this! Just write to the standard DTD and ignore everything else. Providing "optimized" content just makes the compatibility wars rage on, and if you optimize for IE, you're helping MS subvert the documented standards.

Twoflower

Re:Proxomitron

[noom]

Don't do this! Just write to the standard DTD and ignore everything else. Providing "optimized" content just makes the compatibility wars rage on, and if you optimize for IE, you're helping MS subvert the documented standards.

Would you mind posting the DTD for JavaScript?

Oops! I nearly forgot, scripting languages don't have DTDs!

And sure, some backwards people would prefer no scripts in their HTML, but the rest of us actually prefer the design flexibility... and it's just a fact of life that browsers have different, mutually incompatable bugs in their implementations, as well as "additional features."

-n00m

Re:Proxomitron

[damiam]

Oops! I nearly forgot, scripting languages don't have DTDs

They don't?

Re:Proxomitron

[agdv]

And sure, some backwards people would prefer no scripts in their HTML, but the rest of us actually prefer the design flexibility... and it's just a fact of life that browsers have different, mutually incompatable bugs in their implementations, as well as "additional features."

Sure, being able to have scripts et al is great. But I just hate it when a designer implements a simple link as javascript (there goes my option of opening in a new window), use java for something that could be done with a (smaller) animated gif, reinvent the wheel for doing something that could be done with HTML but they do with javascript/java/whatever, etc. The ones I like are the ones that make Flash banners. I purposefully uninstalled Flash, so I don't have to look at their crap.

Re:Proxomitron

[Gerv]

(there goes my option of opening in a new window)

That's bug 55696.

Gerv

Re:Proxomitron

[Todd Knarr]

How about not optimizing your page code instead? Just write HTML 4.01 or XHTML 1.0 or CSS1/CSS2 or Javascript 1.2 or whatever according to the standards ( see www.w3c.org for all of them ) and make life easy on all of us. I find it annoying to go to a site and see "Sorry, Netscape 6.x isn't supported.", flip the user-agent string to IE5.5 and discover that the site renders perfectly in Mozilla 0.9.recent. To me it says that the site doesn't care what customers it annoys and that the designer doesn't know how to create HTML pages.

Re:Proxomitron

[uchian]

How can we optimize our HTML code to render correctly in your browser, if you lie to us about what browser you're using?

OMG your code must be _horrendous_ if you optimise for every browser.

I mean, surely your not one of these people who writes code which _only_ works on netscape and IE, are you?

Re:Proxomitron

[Nurgled]

Even better (although more annoying in the short term for people who browse sites which sanity-check the string) is to push the point home with a completely stupid-looking user-agent string and a comment in it such as "Browser statistics mean nothing, you fool."

That'll learn 'em!

Re:Proxomitron

[Phroggy]

My code is horrendous, because I try to do interesting things that HTML wasn't designed to do. Each browser has its own peculiar bugs, and I've tried to work around all of them. I mostly use browser detection to determine whether or not your browser will likely support something, and if not, offer an alternative that will definitely work - for example, if you're browsing in Lynx, everything will be fine, but if you set Lynx to pretend it's MSIE, it may not work so well.

Browse around http://phroggy.com/ for awhile. Yes, some of the code is crap. I've tried to test it in Netscape 2-3-4-6, MSIE 3-4-5-6, Mozilla, iCab, OmniWeb, Konqueror, Lynx, Links, and I've just installed Opera so I can try that. Of course I test on Linux and Mac OS and Windows, and have had people send me screen shots from other OSes that I don't have available.

Re:Proxomitron

[Phroggy]

Screw that! HTML should be standard and browser independent!

That's a nice dream. Unfortunately whenever you write code that's browser-independent, it doesn't work in some browsers.

Of course, you can write simple text-based stuff with minimum formatting, but there's a lot more I want to be able to do. Have you seen Slashdot's HTML recently?

Re:Proxomitron

[Phroggy]

How about not optimizing your page code instead? Just write HTML 4.01 or XHTML 1.0 or CSS1/CSS2 or Javascript 1.2 or whatever according to the standards ( see www.w3c.org for all of them ) and make life easy on all of us.

Because then it doesn't render correctly in many browsers. I try to check my code at http://validator.w3.org/ periodically, but the reality of Web design is that the official standards only work on some browsers, and to ignore that fact and not compensate for it means some people may not be able to see the site, or it might look like crap, which reflects badly on the designer.

I find it annoying to go to a site and see "Sorry, Netscape 6.x isn't supported.", flip the user-agent string to IE5.5 and discover that the site renders perfectly in Mozilla 0.9.recent. To me it says that the site doesn't care what customers it annoys and that the designer doesn't know how to create HTML pages.

I have never done this. I do occasionally say "sorry, you must have a frames-compatible browser", and of course Netscape 1.x won't work on any site using name-based virtual hosts, but I'll never turn down a browser I don't recognize, and especially not something as common as Mozilla. "Netscape 6 isn't supported" is acceptible ONLY if the message only remains for a few days while the web designer figures out how to work around some quirk, but any web designer who hasn't already done so should be fired, or given a raise and more time to complete that sort of thing. Don't forget the role PHBs play in this.

Re:Proxomitron

[Phroggy]

Don't do this! Just write to the standard DTD and ignore everything else. Providing "optimized" content just makes the compatibility wars rage on, and if you optimize for IE, you're helping MS subvert the documented standards.

The answer to this is obvious: if I write pure standards-compliant HTML, it will render as a horrendous mess in some browsers, and that reflects very poorly on me as a web designer, when most other sites render just fine. Standards are great for browsers that support them, but limiting myself to the W3C standards is just as bad as limiting myself to only supporting the browsers I happen to like.

I suppose if all the browsers you happen to like all fully support W3C standards, you wouldn't understand how this could be a problem.

Re:Proxomitron

[uchian]

Ok I admit it - I was trying to be funny.

A lot of web designers don't go to the extents that you have - in particular, Here is a particular example of a bad site, that will show NOTHING if your browser doesn't identify itself as either netscape or MSIE. (Take a look at the dsocument source and you can see why...)

If you do, then fair play. Shame more web designers do not have that kind of dedication really.

Re:Proxomitron

[Todd Knarr]

What doesn't render correctly? As of now, IE4/5/6, NS6, Mozilla 0.9.3+ and derived browsers and Opera5 all render HTML4/CSS1/etc. per the standards. Some include proprietary extensions, but all of them handle standards-compliant HTML4 correctly. NS4 doesn't but then NS4 doesn't do HTML4/CSS1, having been designed and built before those standards were finalized. Now if you mean they render the HTML differently, that's true, but then they're allowed by the standard to render it differently. The basic idea is that the browser should render things the way they ought to be rendered on the client system, which can vary from system to system as look-and-feel differs.

Of course, you might be trying to do precise page layout and such. Don't. You don't know anything about the client system, so trying to do pixel-level control of layout or force page widths and such doesn't work. I hate pages that make me scroll horizontally because they were designed for a wider screen than I've got, and I hate pages that leave half my browser window blank because they were designed for an 800x600 screen and my system's 1600x1200 on a 21" monitor.

As for frames, I've seen more than a few pages that, when hit with Mozilla 0.9.3, tell me I need a frames-compatible browser. They do this because they've got a hard-coded list of browsers that support frames, and anything that isn't on that list is assumed not to whether it does or not. Worse yet is the site that requires frames but only uses them for a navigation/menu sidebar. Oddly, if I take that side, scan the source and put in the URL to the nav/menu frame manually, I can navigate the site just fine. I've got to hit the Back button to get the menu back to change areas, but it's entirely navigable and usable without frames even though it was designed to require it. And all it takes is one anchor tag in the NOFRAMES section, but the designer won't do it.

Mod this up!

[iotaborg]

Thanks for the mirror, mozilla seems swamped (can't connect at all).

Speed issues. Moz 9.3/9.4

[helixcode123]

Out of curiosity, I downloaded both Netscape 6.1 and Moz 9.3. I found Netscape to be practically glacial, and while Moz was better it still didn't approach the speed of Netscape 4.7X (which came with my RedHat 6.2 installation).

I guess I'm wondering why the new versions are so much slower than the older versions of Netscape. I know this is the M$ approach to application "upgrades", but is it something we really want to emulate?

Re:Yes, but ...

[quannump]

yeah and i can argue that most of the Ask Slashdot Qs belong on Google not slashdot. if Mozilla annoys you, log in and disable the Mozilla topic from showing up on the front page.

Re:Speed issues. Moz 9.3/9.4

[Gerv]

Lynx is probably faster still. But it doesn't mean you necessarily want to use it. This is not an excuse (Mozilla should be faster than it is), it's just an observation.

Gerv

mailer works a little better, too

[BillyGoatThree]

I've been using Moz daily for almost a year now for both web and mail. I downloaded a daily a couple days ago and it's getting better all the time. The most notable improvement: The mailer isn't a time-sink like it used to be. Even in 0.9.3 it would take me upwards of 1 or even 2 minutes to click "new msg", put in 3 recip addrs, type a subject line and then start writing the body. Luckily I only write about 3 emails a week...

Re:Speed issues. Moz 9.3/9.4

[BZ]

> I guess I'm wondering why the new versions are so
> much slower than the older versions of Netscape

Because it's still beta-quality code? There are many performance issues currently being worked on. Also, some things that NS4 does quickly (eg style resolution) take a lot more time to do _right_.

Oh Great!!

[throx]

Now, in the inevitable war between the annoying ad companies and the poor downtrodden browser users we'll get no more popups, but click-thrus or something even more insidious instead.

I can't wait for "This site cannot be viewed without the EvilPopupsAndPersonalInfoCollector plugin installed".

Don't get me wrong, this is a good interim effort but web advertising is going to continue.

<obtroll>
I also find it interesting that the /. crowd decries the use of Smart Tags (because they change content) but is more than happy to change content they DON'T like (popups and banner ads). Do I smell a note of hypocrisy here?
</obtroll>

Re:Oh Great!!

[wadeb]

I think people are more worried about Smart Tags *adding* content than the ad filters removing it.

Besides, most popup filters give you an option to view the popups, say by Ctrl-clicking or what have you.

-Wade

Re:Oh Great!!

[cmowire]

Actually, I don't think so, at least in this case.

Loving Junkbuster to get rid of ads and hating Smart Tags because they change content is being a hypocrate.

But disabling popups is, at least the way I look at it, different. You are not actually changing the content, just preventing a response to said content.

I like the 0.9.4 version very much simply because I can cut off 99% of the annoying popups while allowing the one or two useful, or at least, impossible to avoid, popups. Like when somebody has a web app that sometimes pops up a tip window in response to clicking on a link instead of a whole new page.

I don't have a problem with people tracking my online usage with cookies, because I figure they are entitled to some information about my browsing habits in return for putting up their sites. I don't even mind banner adverts, even though the only ones I'd have a remote chance of wanting to buy based on a banner advert is Thinkgeek. I just hate popups.

Re:Oh Great!!

[roystgnr]

I also find it interesting that the /. crowd decries the use of Smart Tags (because they change content) but is more than happy to change content they DON'T like (popups and banner ads). Do I smell a note of hypocrisy here?

So I go to foo.com's webpage. I expect foo.com's software to give me the content they want to let me have, and I expect my software to display it based on my personal preferences. Nowhere does a third party need to enter into that transaction. This is not hypocritical, because I extend to Microsoft employees the same priviledge: to view what web sites they visit in the way they choose. If the Apache authors were to insert code that added a "Replace IE now!" button to the top of each webpage requested by Internet Explorer, I would find that just as offensive.

Re:Oh Great!!

[Bostik]

Smart Tags add additional (and unwanted) information. In effect, they change the content from the original and do this in a way that the user has little or no way of finding out. Adbusters have to be configured to remove information we wish not to see. They change the content in a predefined and user-configurable way.

Would you like to see your web pages mutilated with links that weren't there when you made them?

The fact that Smart Tags are made as opt-out, is close to immoral. Anyone who builds webpages has to specifically disable them on each page. Just how do you think casual users are going to know about this? Not everyone with web pages is a professional.

Re:Oh Great!!

[archen]

actually I imagine if Mozilla becomes widespread (?!) that pop up ads would probably find themselves being embeded in the click event via links. Sure some pages you visit might not have any links for you to follow, but most places you go probably have links to other pages or relavent media. It certainly isn't hard to figure out what to do from there...

Re:Oh Great!!

[Eil]

Loving Junkbuster to get rid of ads and hating Smart Tags because they change content is being a hypocrate.

No, not quite. In fact, you're not even close. In both cases, (that is, detesting smart tags while installing junkbuster) it is the user's choice what he wants displayed in his web browser. Isn't it every person's choice to have a desire to see what information they want to see and block out that which they don't?

With Junkbuster, one has to make a conscious effort to install it and block advertisments from appearing on their computer.

If Microsoft (et al) had their way with Smart Tags, you would have no choice. You would see the links and if you clicked on them, you would be sent to a page that Microsoft wishes you to see, not neccessarily what you want to see or what information the content creator wanted to publish. The line is drawn at how you determine what "information" is. I do not consider any form of blatant advertising as information and I presume I'm not alone. Junkbuster gets rid of ads, smart tags modify information. It's really that simple.

Regardless, it all comes down to choice. Junkbuster is a choice for the user, smart tags are not. If you believe that smart tags are alright, then I presume you also wouldn't mind Microsoft adding dynamic animated GIFs to your Start Menu and the occasional pop-up reminding you to upgrade to Office XP everytime you opened a Word 97 document.

Re:Oh Great!!

[tester13]

not at all. I want to see what I want to see and don't want to see what i don't want to see.

I'm not averse to changing content per se. I just want to be the one changing it.

Re:Oh Great!!

[throx]

I want to see what I want to see and don't want to see what i don't want to see

My point exactly (even though it was a troll). In both cases the addition and deletion of content is user controlled, explicitly installed and optional. The only difference is turning off ads will deprive the web site of revenue from your hit, and the webmaster doesn't have the option to force you to view them (he has the option of turning smart tags off). If you don't mind ripping people off for their hard work then junkbusters and popup removers are certainly useful.

Re:Oh Great!!

[throx]

Wow - someone responding to the actual post and not the troll.

That's exactly what I'm afraid of - a "cold war" between the browser developers and the web advertisers. Slowly web ads will become so complex that getting any content will mean clicking on all sorts of advertisers, or running spyware plugins.

Like I said - it's not a bad thing that they've done this, just I can see the stakes being raised and things just getting ugly.

Re:Oh Great!!

[Ms.Taken]

If you don't mind ripping people off for their hard work then junkbusters and popup removers are certainly useful.

So if I turn off ads I'm "ripping people off for their hard work"??????? What if I don't use junkbusters, but just scroll past the ad without looking? What if I look at the ads, but don't read them carefully, or read them but don't click through? What if I buy a newspaper and throw out the advertising inserts? What if I'm watching tv and I leave the room during the commercials?

I'm so ashamed! It turns out I've been ripping of hard-working people on a daily basis for years. I promise that from now on, I'll read every ad in every webpage, magazine and newspaper I see, and if I have to pee, I'll wait until the commercial is over.

Re:Oh Great!!

[cmowire]

Actually, that is true that Junkbuster is also a choice, the choice not to load advertising banners. So I stand corrected on that.

I also think that, in order for Junkbuster to be really useful, it would have to heavily modify a page. Again, I feel for the site owners and figure that they have a right to put banner advertisements on their pages in return for putting it up for free, so I don't actively try to install Junkbuster. But then, I'm a pragmatic moderate who thinks that the web should eventually make money for people and doesn't mind some advertising.

And, no, I don't think SmartTags are allright. The and in my above message is a logical AND, not a conversational AND. If you both hate SmartTags and love Junkbuster, then and only then, are you running the possibility of being a hypocrate.

Re:Oh Great!!

[Homewrecker]

What if I don't use junkbusters, but just scroll past the ad without looking?

Then the site still gets paid for your impression and continues to offer free content as a result. It's not rocket science.

Re:Oh Great!!

[throx]

Don't be so freaking stupid. Sites get paid for the number of hits on the advert. If you use junkbuster then the ad never gets hit and the site doesn't get paid.

This has nothing to do with actually READING the ad, so strawman somewhere else, please.

Re:Oh Great!!

[funky womble]

The Proxomitron is a great adblocker for Windows - lots more modifications possible than with Junkbuster, lots of blocking for certain types of javascript etc., if you want heavy modifications this is definitely the most flexible software. Junkbuster isn't too good at HTTP protocol, you have to fallback to 1.0 which slows things down a lot (see bugzilla).

Re:Oh Great!!

[Ms.Taken]

Please excuse my freaking stupidity, but how do I click on the link if I don't READ it? I think you've actually made my point for me. If the site only gets paid for hits, then it doesn't matter whether I SEE the ad or not. If I don't click through, they don't get paid.

Re:Speed issues. Moz 9.3/9.4

[zachlipton]

In my mind, I don't think of Mozilla or Netscape 6.x to be an upgrade to Netscape 4.x, I think of it as a completely different product. Any time that you rewrite 100% of the product, you can expect the new version to be slower, more infested with bugs, and just "feel" worse than the older version which has been tended for many years.

However, if Netscape decided not to do the 5.0 rewrite, disaster would be the only end. The old code was not mantainible and doesn't allow for the powerful new features and embedding that seamonkey allows for.

Speed is something that is being worked on and is significantly better than before. I won't mention full names here on /. without permission from the people involved, but someone at Netscape (d. hy.) did a lot of work on page loading and a new contributor did a lot of proformence work as well recently (jes.). Mail/news also uses the widget in the folder-paine, which has great speed increases as well.

So we are trying the best we can. As always, patches are welcome.

Zach

Re:Speed issues. Moz 9.3/9.4

[dwlemon]

Mozilla's gui responsiveness and load time is slower, but the rendering engine is fast: it's the only browser that renders big tables like Slashdot comments as they load instead of waiting for the entire page to load (on my 28.8). The only browser I havn't compared it to is Konqueror.

I think it's a fair trade-off. It's always in memory on my machine anyway.

What's new in 0.9.4

[mbrubeck]

The difference in 0.9.4 is that you can disable popups on page-load/page-close only. This gets rid of most popup ads, while preserving less-annoying uses of popup windows (unlike 0.9.3).

See this newsgroup post for details.

Re:What's new in 0.9.4

[fobbman]

Long Live Pornzilla!!!

Re:Speed issues. Moz 9.3/9.4

[BrookHarty]

Links is text based, renders frames, and has HTTP 1.1/keepalive support, color. It supports the mouse in terminals too, so you can just click links.

I perfer links over lynx.

Power

[HanzoSan]

Netscape 4.7 is older, WEAKER, of course its going to be faster.

Think about it.. the most powerful browser cannot be the fastest browser.

Re:Power

[noom]

Netscape 4.7 is older, WEAKER, of course its going to be faster.

Think about it.. the most powerful browser cannot be the fastest browser.

I'm thinking about it.

Ok, I'm scratching my head now, I'm still thinking.

Ok, something is starting to happen, I think I've got it.......!

DOH! Tell your mom not to use her teeth, dammit!!!! I can't stand it when she does that...

-n00m

Re:Speed issues. Moz 9.3/9.4

[JoeBuck]

Konqueror does a much better job than Netscape
4.x of doing styles right, and it's a lot faster
than Mozilla. I wish the Mozilla people well,
but it's simply false to claim that the slowness
is required for a correct implementation.

Re:How Are you gentlemen

[tpackage]

If management isn't already open to at least a discussion of telecommuting then it's likely your request will fall on deaf ears. This however isn't the case if you've been at a company for a long time and have quite a bit of power.

Stop complaining about speed!

[HanzoSan]

Everyone who complains about speed doesnt know anything about how computers work.

New software is bigger, more powerful, and NEEDS a more powerful computer, RAM IS CHEAP, dont tell me Mozilla uses too much ram when you can buy a gig of ram for under $200.

Get a faster harddrive, if mozilla is slow you are most likely using cheap IDE crap.

Now, if you have a modern computer THEN you may use modern browsers and modern software, if you have a computer which was made before Netscape 4.7 was released, then you should be using netscape 4.7, your computer will never be powerful enough to run mozilla.

Re:Stop complaining about speed!

[InsaneGeek]

I shouldn't respond to such a stupid comment, but what the heck...

Firewire rated: 50/mBs (i.e. 400/mbs)
SCSI III rated: 80/mBs (i.e. 640/mbs)
SCSI over fibre channel: 100/mBs (i.e 800/mbs)

Most Firewire drives: 5400 RPM (I've yet to see a faster one, so I'll say most)
Fastest IDE: 7200 RPM
Fastest SCSI & Fibre: 15k RPM

That's a real big nail there.... heck IDE is faster than Firewire

Re:Stop complaining about speed!

[Homewrecker]

You know, that's a really ironic post in light of the usual "Linux isn't bloatware and Windows is" chestbeating that I usually see around here. How many times have Slashbots bitched and complained about being forced into a hardware upgrade by the newest, larger, version of Windows? It's amazing how one's perspective changes when the same criticism is launched against something you like, eh Hanzo?

If I have to upgrade my PC to appease a lumbering beast like Mozilla, it seems to me that the TCO of Linux is actually higher than Win2K and IE, which run fine on my 433/128 without upgrading. Odd, don't you think?

Re:Stop complaining about speed!

[madprof]

IDE is clearly going to be fast enough for everyone who doesn't need the kind of sustained throughput that SCSI allows.

Re:Stop complaining about speed!

[WNight]

I've seen a few firewire drive enclosures that run IDE HDs, one-per controller, so they're all masters, and all seperated from everything else.

They're pretty zippy, and because the drive is 3rd-party, you can put in the fastest drive you can find.

The same drives released as SCSI could be released as Firewire, the same as they could release them as IDE. The reason they don't, more than performance, if that people with SCSI systems will pay a premium.

It's like hardcover/paperback. Some people will pay anything to have 'it' right now.

Re:Stop complaining about speed!

[Trogre]

...and after these words were spoken it came to pass that Bloatware was born. The efficiency beast was banished from the land, never to return.

aww mozilla too slow? Use lynx!

[HanzoSan]

Lynx is lightning fast. its perfect for a nix user who wants weak yet fast software.

Re:Speed issues. Moz 9.3/9.4

[Gerv]

We'd be very interested to hear of pages Konqueror gets right, and Mozilla gets wrong. Please file bugs in Bugzilla here, and then quote the bug numbers. We'll get right on it.

Gerv

"Patches and/or help is welcome!"

[zachlipton]

Rather than sitting around and discussing how Mozilla should load pages 1.5ms faster than it does today, why can't we all get off the ground and help. mozilla.org has made it very easy to find the resources that you need to help out, espically with non-coding work. Yes, you did hear me say it: "You can help with Mozilla without coding!"

If you travel over to one of the following pages on mozilla.org, you can learn all that you can do to get involved. Confirming the unconfirmed (from page number 3 below) is a great way to get involved, doesn't take much time, and is of a big help when all the many bugs come in after a big release like this.

1. http://www.mozilla.org/start
2. http://www.mozilla.org/get-involved.html
3. http://www.mozilla.org/quality/help/

two things to make it my permanent browser

[nilstar]

If mozilla used less memory (I've got a laptop with 192MB and it is still slow!).... and.... (though not critical) if it could read IMAPish email from my @netscape.net account.... then I'd use it all the time....

Re:two things to make it my permanent browser

[ksheff]

If you have 192M of RAM on your laptop and you still feel that it's slow, the problem isn't due to lack of RAM. I've used earlier releases on machines with only 64M and it worked fine.

Re:two things to make it my permanent browser

[Salsaman]

On my system (Linux 2.4.9), mozilla is currently using 23M. Thats hardly a lot if you have 192M available.

The browser is great, but where is the spell check

[(H)elix1]

Come on already.... lets get a spell checker folded into the email client. I need ALL the help I can get. I tried hacking the 6.1 checker in to the last build, but no luck. Is there any word on when - world acording to me - one of the most basic things about an email client will be included?

Snap mouse to default button in Windows

[abischof]

In Windows NT/2000, users can set in the mouse preferences for the mouse to automagically move to the default button in dialog boxes and alerts. However, Mozilla doesn't currently cooperate with this.

The bug has keyword "helpwanted", so if you know how to accomplish this functionality, please speak up :). Or, if you aren't inclined to programming, you can also vote for the bug (of course, you'll need a free Bugzilla account to vote).

Re:Snap mouse to default button in Windows

[Osty]

Having glanced at the bug, I'd guess (without digging into any Mozilla code) that the problem is not that Mozilla dialogs are not Windows dialogs (which, as the bug points out, they are), but that Mozilla buttons and other widgets are not Windows buttons and widgets.

Let me reiterate by stating that I have not looked at the Mozilla source, but it is my understanding that a button, or listbox, or textbox, or label or any other widget is in XUL is defined by XUL. They are not just wrappers for native widgets (they probably should be, with a user-drawn attribute and all the necessary handling code), for if they were, then a default button on a dialog would be a default button on a dialog. As it is, a "button" on a XUL dialog appears to be a bitmap that happens to react to mouseovers, mouseclicks, and possibly keyboard presses, and a "default button" is a fore-mentioned "button" that the dialog knows should accept keyboard input when no other "button" has focus.

If I'm wrong about this, feel free to correct me. Maybe I'll dig into the Mozilla win32 code this evening and find out for myself. But if I'm right, the only real way to fix this that I know of is to actually use wrapped native widgets, at least for a default button.

Re:Snap mouse to default button in Windows

[jchristopher]

Is there a reason for Mozilla to use anything BUT the standard Windows buttons, dialogs, etc? This was one of my big pet peeves with the older releases, i.e., non standard scrollbars, menus, etc.

StarOffice suffers from the same problem. Why confuse the user with a non-standard dialog?

Re:Snap mouse to default button in Windows

[Gerv]

Is there a reason for Mozilla to use anything BUT the standard Windows buttons, dialogs, etc?

Two reasons.
1) If they did, there wouldn't be any ports apart from possibly Macintosh (not enough time)
2) To get the things required by CSS 2 and 3 (animated background images on buttons, anyone?) you need to reinvent your own widget set. IE does the same thing. Having done that, why not use it for a cross-platform UI?

Gerv

Re:The browser is great, but where is the spell ch

[Gerv]

Netscape has a proprietary spell-checker which it ships. No-one has yet found time to write an open-source one. Obviously, no-one at Netscape would spend time doing it, and external contributors are busy on other things. The usual trick is to use a build close to a Netscape release and install their spellchecker.

It would make a good CS project for someone. Fuzzy logic matching isn't all that hard. The UI is open source, it's just the back end that's currently proprietary.

If you are interested, mail me and I'll point you in the right direction.

Gerv

Re:The browser is great, but where is the spell ch

[Salsaman]

Isn't a/ispell open source ?

Re:How Are you gentlemen

[tpackage]

Yeah I know... oops. This was my first post and I got a little anxious.

Interesting

[synclayre]

Im still using an 0.8.x build of Mozilla and for some reason Im failing to see all the instability that the newer releases are striving to thwart.

Maybe its just me? Im not a 'power user' perse... most of the sites I visit aren't multimedia intensive with the exception of some Flash which isnt readily supported anyway.

I guess if something happens in which my browser stops working Ill upgrade, but until then Im happy where I am =/

Kudos to the Mozilla team for all they're hard work and maybe someday Ill upgrade and check out all the new features ;)

Re: Mozilla user agent

[abischof]

Changeable user agents is bug 46029. Feel free to vote for the bug if that issue is important to you.

Re:The browser is great, but where is the spell ch

[BrookHarty]

I used netscapes 6 spellcheck.xpi with an early version of mozilla and it wored fine. It doesnt work with 0.9.4 (just tried) hopefully someone fixes this soon.

First bug post...

[painkillr]

Or maybe it's not the first.

For mozilla 9.4, win32 on winXP using the GrayModernIII skin:

Edit > Preferences > Mail & Newsgroups

The formatting in the window is in a messed up font and the background doesn't show. Blah.

----------------

Re:First bug post...

[zachlipton]

Thank you very much for trying Mozilla and reporting bugs. However, this is not the right place to report a bug as there is no way to track it here with the hundreds of other comments.

Please travel over to http://www.mozilla.org/quality/help/bugzilla-helpe r.html and follow the instructions there to report bugs. Espically, please please please search for duplicates before you file a bug. After a release, the number of bugs filed jumps by huge amounts and many of these are duplicates. Please let us who are working on Mozilla work on code and not weeding out duplicate bugs.

Thanks

Re:First bug post...

[Metrol]

What version of OS was Microsoft pushing when Mozilla got started anyway? I'm thinking it was Windows 98 first edition. In the mean time they've had 3 major upgrades to that, the last of which was a total overhaul.

Sorry, just something that occurred to me as I was reading this. As Mozilla has been working on this browser, it's primary environment has been a moving target. One of the downsides of how long this is taken I'd imagine.

Re:0.9.4 Linux does not work

[Salsaman]

Works fine for me, linux 2.4.9, X 4.1.0.

Re:The browser is great, but where is the spell ch

[abischof]

The bug for getting a spell checker into Mozilla is bug 56301. If you can help out with the effort, that would be fantastic, as the bug is somewhat stalled at the moment.

It used to be that you could install Netscape's spellchecker, but that is no longer supported.

PS Gerv: This message isn't directed at you, but primarily at the parent post to your post.

tag submits forms?!

[smclean]

Anyone else notice that this release's tag now automatically works as a submit button when it is inside a block? What is with that? I am programming a bunch of php software and I love using to attach javascript onClick event handlers to, and this really screws up a bunch of my forms. I know this is not standard, no previous version of mozilla or netscape, and even IE has this functionality.

Anyone know where they got this idea? Anyone know a way to turn it off? Are they planning on keeping this awful behavoir? This sucks, I am going back to 0.9.3 and hoping IE never follows suit.

Sean

Re: tag submits forms?!

[BZ]

> Anyone else notice that this release's tag now
> automatically works as a submit button

What do you mean by that? Are you referring to the <button> element? If so, a <button> with no type specified is supposed to be of type "submit" per the HTML4 spec. If you want it to be type="button", say so in the tag.

Google cache of mirror list

[snol]

Here

This sentence included because the filters are biting. And this one too.

when I was posting stories to /. on Tuesday ...

[timothy]

it was with a daily build (a few weeks old) of Mozilla 0.9.3 -- it never crashed, never even noticeably slowed down, over a more than 14-hour stretch. I was afraid both that the machine would crash and that my ISP connection would break, because phone service was so sketchy (this was not far from DC, but I circuits were busy all over the country). But I *expected* Mozilla to crash (I find that most browsers don't like to be on forever, Mozilla one of the worst offenders), and it never did. Thanks, Mozilla guys :) Good show.

Aside: for years, I've heard people praise IE to the skies. I've tried it out, have sampled a few versions of IE (on Mac and Windows) sometimes, but they've never impressed me as being better for anything *I* do / want to do than either Mozilla or Konqueror, and seem to crash just as often. My iBook came with both Netscape and IE installed, but so far no compelling advantages are apparent.

timothy
(wireless in the backyard, blacksburg va)

Has Fizilla been updated too?

[A_Non_Moose]

Can't get in at the moment, but 0.9.3 was very good on OS X (if a bit of a CPU hog).

On OS X I noticed something odd:
On Internet Explorer 5.X it could not jump to the first 2 anchors on *any* page no matter where they were. As a webmaster I was obiously going "HUH!? WTF?"
Mozilla/Fizilla: Perfect every time.

IIRC I noticed this too:
IE used less cpu but more "kernel space" acccording to my load docklet.
Mozilla/Fizilla used cpu all the time, but "kernel" usage was ~10% to IE's average and spikes of 25 to 50, respectivly.

Stability, well, I found myself using IE heavily.
Why? Well, pegging the cpu like seti@home while using just a browser annoyed me...dunno why besides "that is what seti is supposed to do" and IE made the load docklet look interesting (blue/red graphs).
Sorry, but it is the "blinky light/shiney object syndrome" that has affected me since getting into computers.

Moose

Ooooo, shiney objects!

Re:Has Fizilla been updated too?

[Lysander Luddite]

Yes, there is a OSX build of 0.9.4. It is on the ftp server. I am using it now.

Some observations:
1. I am using the Aqua theme. the scrollbar seems narrower than before because my scroll thumb graphic is munched up appearing non contiguous.
2. According to the Dock, there is a window called "Hidden" that I cannot access nor close. This is in addition to any other windows open.
3. Preferences and settings from the previous build transferred over.
4. Upon launching he first time you get taken to the Mozilla.org home page.
5. The scrollbars have their arrows pointing left and right instead of up and down. Happens on form elements like this text box entry too.

Haven't noticed anything else of import. Overall doesn't appear much changed to me yet.

&lt;button&gt; tag behavoir is whacky!

[smclean]

Anyone else notice that this release's tag now automatically works as a submit button when it is inside a block? What is with that? I am programming a bunch of php software and I love using to attach javascript onClick event handlers to, and this really screws up a bunch of my forms. I know this is not standard, no previous version of mozilla or netscape, and even IE has this functionality.

Anyone know where they got this idea? Anyone know a way to turn it off? Are they planning on keeping this awful behavoir? This sucks, I am going back to 0.9.3 and hoping IE never follows suit.

Sean

Re:&lt;button&gt; tag behavoir is whacky!

[Gerv]

Are they planning on keeping this awful behavoir?

Best way to find out is file a bug and see :-) If you don't like it, and it makes it into Netscape 6.whatever'snext, then if you haven't filed a bug you've only got yourself to blame.

Gerv

Re:&lt;button&gt; tag behavoir is whacky!

[BZ]

The definition of <button>. Of particular interest:

type = submit|button|reset [CI]
This attribute declares the type of the
button. Possible values:
* submit: Creates a submit button. This is
the default value.

Mozilla's just finally gotten around to what is _very_ plain in the spec. use type="button" if you want it to be just a button.

Re:&lt;button&gt; tag behavoir is whacky!

[jesser]

That's bug 97829, "simple in a form causes submit", marked as invalid. You can vote for the bug even though it's currently marked invalid. (I'm currently voting for two 'wontfix' bugs and one 'invalid' bug in addition to a bunch of open bugs.)

Re: links

[seney]

lynx sucks my ballsack. links doesn't.

links for os x doesn't have mouse support, btw.

b ut, it still doesn't like my ballsack.

JavaScript window.open() work a rounds...?

[BLAG-blast]

<I>including the ability to
disable the JavaScript window.open()
method during page load and unload events.</I> <P>
Ah! That's why I'm seeing a large delay before
some pop up ads appear. I.e. the evil JavaScript
program wait until after all loading is done before calling the window.open() method, once the
window has been openned, then it can continue with
loading events.... let's just make then illegal instead. Or just boycott all who use them and hope they go away some day....

Re:Speed issues. Moz 9.3/9.4

[Ungulate]

The reason moz/nscp6 is slower than 4.x is that the entire user interface is built out of XML, CSS, and javascript (something collectively referred to as XPFE). This approach ensures that there's as little platform-specific code as possible, making mozilla available on a wide number of platforms that simply wouldn't have been plausible if everything was being done natively. It's been said many times that if not for XPFE, there wouldn't have been versions for anything but Windows and possibly Macintosh.

The good news is that UI responsiveness improves with each release, and I fully expect it to equal 4.x in time. However, I've read that GTK is a bottleneck of some sort, so that's why Windows has a performance advantage over *nix.

Re:The browser is great, but where is the spell ch

[Gerv]

but that is no longer supported

Well, the 6.1 spellchecker is not supported with current builds. But if you wait for the next Netscape release, the spellchecker will work with those builds and all following builds (until we break it again ;-)

Gerv

Comment removed

[account_deleted]

Comment removed based on user account deletion

come on...

[green pizza]

Mozilla 0.9.4 still feels sluggish on my machine, a 1.1 GHz T-Bird w/ 512 MB of PC133. It feels sluggish under both Linux and Windows. Yet the GIMP and Photoshop (not to menton my games) never felt so fast. If I can shuffle AutoCAD projects with 5+ million elements with ease, I should at least be able to have a complete, yet zippy, web browser.

Granted, newer software is generally larger and more capable, and thus often requires more cpu cycles to do its job. Yet Mosaic and early (pre 2.0) versions of Netscape ran fine on my Sun SPARCstation 10 so many years ago. With advances in coding and cache techniques, not to mention the abilities of modern compilers and the speculative processes doen in the modern cpu -- why must a modern web browser run so slow??

My SPARCstation 10 had a single 50 MHz SPARC processor and 32 MB of rather slow ram. Has Mozilla gotten so far out of hand that even the latest 1+ GHz wizbang PC can't even handle it? Is Mozilla actually more demanding than my Maya rendering daemon??

I say finish up Mozilla. Release 1.0 'when it's done'. Then go back to the drawing board and start over. Bring in some of the old school coders, the folks that didn't have 4+ GFLOPS CPUs. Bring in the old browser folks... Marc Andressen, JWZ, etc.

Sure, Mozilla will be fine by next year when it hits 1.0 and when we all have 2.0 GHz PCs. Browsing will be great at that point. But I pitty the next advance in browsing, because Mozilla 2.0 will certainly bring back the slowness. It is time to start over and do it right.

Re:come on...

[Gerv]

Then go back to the drawing board and start over.

Are you committing to coming to help us if we do what you say? :-)

Gerv

Re:come on...

[dennism]

Have you really used Mozilla for any amount of time?

On a Celeron 700mhz Dell laptop, Mozilla .9.3 felt just as fast as IE, if not faster at times (when going back to a previous page, etc.). With the fast start feature, it's just as easy to start Mozilla as IE. On my iBook (366mhz, 320mb RAM), Mozilla is again as fast as IE on the same box.

I'm using .9.4 right now, and I'm pretty impressed. If 1.0 is optimized and relatively bug-free, then IE is going to be left behind (technical not market). This browser has given me some new hope in what was Netscape.

Re:come on...

[jchristopher]

We can yak all we want about why it's so slow, or whether it's slow on this or that system. Fact is, the benchmark is Internet Explorer. (Opera on some systems is also very fast).

So a better question might be, given that IE is so tightly integrated into Windows (lots of unneeded code for a browser to have) why can't Mozilla be faster than IE? I say lose the useless sidebar and other bloat and just build a BROWSER.

Re:come on...

[Jacek Poplawski]

>Mozilla .9.3 felt just as fast as IE
As far as I remember Microsoft software was always slow . Now I see many comments "something is almost as fast as Word, as Excel or as MSIE - that's great!". What the hell happened? Why this is fast today?
Even, if StarOffice is much slower than MSOffice - that doesn't mean MSOffice is fast! Dillo, links, gnumeric or LyX is fast, well written software. Mozilla (which I am using now), MSIE or MSOffice is not.

Re:come on...

[funky womble]

For all but opening new windows, Mozilla is faster than IE on my windows box. Even with the recently added slow-down-turbo-mode-to-accommodate-the-very-few-p eople-who-use-profiles code it's still faster to start.

Re:come on...

[Gerv]

Yeah, but you've anonymous :-)

So, what's wrong with the codebase? I'm sure that, to make a statement like that, you have at least a passing familiarity with it.

Gerv

Re:come on...

[madprof]

Microsoft software was slow, yes, but consider that processor speed has been continuing to follow Moore's Law and software like Office clearly hasn't become so big and lumbering at the same rate.
IE is actually pretty fast these days, or at least fast enough. It would be nice if Mozilla were able to outdo it in terms of speed and perhaps force Microsoft ro re-evaluate the speed of their browser.
Of course I'm dreaming. :-/

Mozilla vs. Communicator

[under_score]

I've been using Netscape Communicator 4.72 for the last X years. Why? I have over 82000 email messages that I have kept! I do not want the hassle of moving over to Outlook or some other platform for email - lots of filters to set up, _lots_ of folders to set up, and many many thousands of messages to transfer. So I've been waiting for Mozilla to mature. I have tried it a few times over the last two years - and always it has not quite made the cut. In particular, importing the huge number of messages and folders has been a real hassle (often crashing). I'm getting close to switching. This release seems much better. We'll see...


"Blade Runner" the Comic Noir, "Akira" the Film

Re:Mozilla vs. Communicator

[swordgeek]

Mozilla???

Netscape?????

These are web browsers. If your email is so important to you that you can't just archive (or trash) your 8200 messages and pick a new platform, why are you using a web browser for mail services in the first place? Here's a hint: They're not very good at it!

Re:Mozilla vs. Communicator

[J'raxis]

How are the messages stored? mbox format? Then the message files/folders are mostly portable between other email clients (Eudora, `pine`, etc., all use mbox). Most other email clients either use mbox format or can import from it.

mbox files look like this (if opened in your text editor). Each file serves as a single mailbox, containing all the messages as plain text:

From [user] [date]
[headers]
[body]

From [user] [date]
[headers]
[body]

[...]


As for filters, I dont know.

Re:Mozilla vs. Communicator

[under_score]

What would you suggest for an email client? This is really a legacy application problem. 82000 messages (not 8200) might be easy to transfer over if it was just a matter of the headers and content. But unfortunately it is also a matter of the folders and filters. I have tried importing into Outlook (both versions), but the process has been incredibly painful and I have been very happy that I backed up my email before I started the attempts. I'm not comfortable looking at a text-based client. Nor am I (currently) in a position to pay lots of money for an enterprise-quality solution. Please, if you know of a good email client that meets these requirements, let me know! Thanks.

Re:Mozilla vs. Communicator

[Metrol]

Here's a hint: They're not very good at it!

I seriously beg to differ. Netscape 4.7x is still my primary mail client after years of use. I have not run across a single other client that serves my needs nearly as well. I have looked too!

On Windows I tried Outlook out for about a weeks worth of use. Back to Netscape for a far superior handling of address auto fill. This is especially true when your addresses are coming in off an LDAP server. Tried Calypso, Eudora, Pegasus, and all of them missed some crucial feature I've grown to rely upon Netscape's Messenger for.

On the Unix side of the house the story isn't much better. Mostly using KMail instead of Netscape due to the fact the fonts are a lot more readable. Feature wise, it to just doesn't hold up to Messenger. The other *nix mail clients I've tried couldn't hold a candle to most of the cheesey shareware mail apps over on Windows. Every once in a while I really do like the ability to bold some text, or work a little HTML into my E-Mail. Nothing stable outside of Netscape can do that at this point.

I'm starting to really like the looks of Mozilla's Mail, but damn is it memory hungry. Still, I'd LOVE to have an E-Mail only client that offered the basic functions of Netscape's without all the rest of the browser thrown in.

Now I'll just sit back and wait for a troll to suggest that I need to go and code this.

Re:Mozilla vs. Communicator

[Eil]

I wish I could recommend some free ones, but (fortunately) I haven't used Windows in a few years. I used to use Eudora (a *very* nice and very capable email client) back when I was using Windows. The only downfall is that it is commercial software that you must pay for. I'd say if I were still using Windows, I would buy Eudora.

In Linux, I used to use KMail until the latest release of KDE... For some reason, loading any KDE application takes an eternity on every computer I've tried it on. So I switched to Mozilla's email client, which isn't much better but it gets the job done.

I'm going to try out newer versions of some of the popular GTK mail clients soon.

Re:Mozilla vs. Communicator

[swordgeek]

In Windows, I'd say Pegasus or Eudora, although there are other nice ones two. I'm looking at Unix clients now, and am moderately displeased, but Mahogany is looking promising.

My point on the 82000 (!) messages is _why_ do you need access to them all? Dump the whole directory tree to a CD or six, along with a copy of Netscape 4.7 (just so you can read them in the future), and then ignore them. I can't imagine needing ready access to many messages that old, even for a company that needs to keep seven years of records.

Re:Mozilla vs. Communicator

[swordgeek]

Don't worry. As a non-programmer, I have threatened to pummel all of the coders who say, "well I don't want it like that, so go write it yourself."

I haven't played with Mozilla mail since 0.8.1, but it was such an ugly outlookalike (hah! :-) then that I'm not interested in seeing if they've got it functional now. (which it wasn't then)

I'm looking at Mahogany right now. A friend who found that Eudora and Pegasus didn't have all of the features she needed is loving The Bat right now. Check those ones out.

Re:Mozilla vs. Communicator

[uchian]

You mention these features that other email clients don't have. Would you care to elaborate on what they are?

Re:Mozilla vs. Communicator

[maXter]

If you run KDE applications often and you are not using kde, try running a kdeinit process in your startup script for X (xsession or xinitrc depending on how you're login works. It uses about 3.5 megs of memory on my machine, 3 of which is shared. This will speed up startup of all KDE apps considerably. This process is of course always running if you're working in a kde environment. Hope this helps.

Re:Mozilla vs. Communicator

[great throwdini]

I used to use Eudora (a *very* nice and very capable email client) back when I was using Windows. The only downfall is that it is commercial software that you must pay for. I'd say if I were still using Windows, I would buy Eudora.

Since the release of version 5.0 -- it's now at version 5.1 -- Eudora has been "free" in same same sense that Opera is now "free": embedded banner ads.

To be fair, Eudora 5.x has three modes of operation: light, sponsored, and paid. Feel free to read up on them at your leisure... might as well do a little reading on its feature set as well.

Re:Mozilla vs. Communicator

[great throwdini]

feature set :: http://eudora.com/email/features/index.html

Re:Mozilla vs. Communicator

[Eil]

Wow, thanks for the info. I haven't been keeping up Eudora obviously, but if it's still quality software as it was before, it sounds like I might be recommending it to a few family and friends.

Re:Mozilla vs. Communicator

[Metrol]

You mention these features that other email clients don't have. Would you care to elaborate on what they are?

I thought I had, but I'll summarize my top can't live withouts.

1. Addressing. No other E-Mail client handles this half as well as Messenger. A seperate line for each recipient, with each of these being able to toggle to: cc: and bcc: independantly. The full name AND E-Mail address showing up in the address bar. Auto completion of names as you type.

2. LDAP Addressing. All of the above, with a seemless integration to a remote server running LDAP. Other clients have some LDAP abilities, NS handles it darn near perfectly.

3. Multiple address books. This didn't come around until later versions of NS, but when it did it was handled very well.

4. HTML composition and viewing. Pretty much every component of Composer can be utilized when putting together an E-Mail.

5. Threading that works properly.

Yes, I realize that Outlook handles some of these items, but not nearly as well as Netscape. As I personally reviewed these two Outlook came very close feature wise to Messenger, and beat it in certain areas like the handling of multiple accounts. In the final analysis for me, the addressing features won out over multiple account handling. That, and I found over some time that I just preferred the interface that NS provided.

On the *nix side of the house I haven't seen any clients deal with the above features nearly as well. I wish one did, as I really hate the UI on the *nix version of Netscape 4.78. Only thing I'm seeing come close is Mozilla, though I'd personally prefer a stand alone client that called to Mozilla for HTML rendering only. From the looks of things, that ain't happening any time soon.

Re:Mozilla vs. Communicator

[damiam]

Evolution is a free Outlook clone for *nix. Still beta but it's very close to a 1.0 release.

Re:The browser is great, but where is the spell ch

[Gerv]

That won't be fixed. You'll have to wait until the next time Netscape releases a spellchecker.

Gerv

IE specific extensions are the problem...

[Drakula]

and people design their pages to work with IE. Therefore, other browsers who work with the standard can't render those pages correctly.

Although, even IE doesn't get Java pages right all the time.

where is the popup disabling feature?

[jchristopher]

I can't figure out to use the new feature for blocking window.open() attached to OnLoad or onUnLoad in the latest build for Win32. Can anyone point out where it is?

Re:where is the popup disabling feature?

[Gerv]

Have you read the release notes? It details how to do it in there.

Gerv

Re:where is the popup disabling feature?

[jchristopher]

It sure does... it specifies a file that doesn't exist on my machine, in a path that doesn't exist either. Default install.

Re:where is the popup disabling feature?

[Gerv]

OK, so maybe I screwed up the release notes.

Could you tell me what OS you are using, and what the results of searching your hard disk for a file called prefs.js are? Mail me if you like.

Gerv

Re:where is the popup disabling feature?

[jchristopher]

Okay... looks like the file is there, just the readme is wrong. (For windows, anyway.) I'm running Windows 2000, SP2, Mozilla build 20010911303.

The prefs.js file exists in C:\documents and settings\username\application data\mozilla\profiles\profilename

I let it import my 4.7 prefs during the first launch.

I'm no programmer, but it seems to me that most Windows programs store prefs inside Program Files/Program Name/prefs or somesuch place. I know storing all the prefs together elsewhere is a *nix thing, so I can live with it, as long as the readme points you to the correct place.

Can you abuse it more?

[kimihia]

Person who wrote this article: can you be a bit more derogatory with your one line summation?

from the fast-as-a-speeding-snail dept.

Have you even used the thing? What sort of XT system with 64K of RAM do you run it on?

It works. I've used Mozilla solidly for nearly a year now.

(Plus regarding disabling popups, that was around since 0.8, although I hear it now has an interface other than a text editor. I'll install the latest milestone once I post this comment and close Mozilla.)

Re:Can you abuse it more?

[jchristopher]

although I hear it now has an interface other than a text editor. (for disabling popups)

Not that I can find. If it does, do tell.

Re:Can you abuse it more?

[kimihia]

Sorry, false alert. It is just a new user_pref value that can be set from prefs.js.

* user_pref("dom.disable_open_during_load", true);

good work

[Ender Ryan]

I have to say I'm am extremely impressed with the latest releases of mozilla, there has again been a very very nice speed jump.

As a submitter of bugs, it's good to see them getting cleaned up, at this point it's better than many browsers that call themselves 3, 4, 5, 6, 7... whatever.

Stability is getting really good, I haven't been able to crash the latest 0.9.3 nightlies or 0.9.4, even with java, javascript, and flash.

Really excellent work, my thanks goes out to everyone who has helped with Mozilla.

Re:good work

[zmooc]

My little useless addition: @ 0.9.3 I decided to try Mozilla again and I've been using it since (on my debian-unstable x86-box). I've tried a new nightly about every day and have just tried 0.9.4. I do NOT find them very stable. Especially 0.9.4 which has crashed over 10 times the last 6 hours (according to the Quality Feedback Agent). I've not been able to determine what was wrong, but it seems to crash randomly. Anyway. It's starting to become a real neat browser now and I will keep using it even though it crashes more than once every hour.

Similar Option For Omniweb users

[helixblue]

I've been happily using Omniweb under MacOS X for some time, and there is a very nice near-equivalent feature you guys should know about. Hidden in OmniWeb -> Preferences -> Javascript, there is the following dialog:

Scripts are allowed to open new Windows:

* Always
* Only in response to a link being clicked
* never

The second choice of course, is the preferred choice. I'm glad that I'll be able to use Mozilla soon for the sites where Omniweb's javascript doesn't work, as the SMP bug under MacOS X now has an uncommitted patch! Too bad Omniweb will still look better for now ;)

Much props to the Mozilla crew. Keep it coming

Re-querying

[_generica]

I'd like to know if the annoying behviour of requerying a page when "view source" or "save" ing has been changed ?? It makes no sense at all to re-download a page to do these things, rather than use the copy being displayed, and this completely stuffs up pages using forms which you have filled out... Why isn't this behaviour optional ???

Re:Re-querying

[BZ]

This behavior is not optional because there is no API to request a cached load at the moment. This API is being added. It is not yet in 0.9.4. As soon as it's added, I'll change the view source code to use this API and the problem will be gone. There is an existing patch for "save" that with a few small changes will start working then too.

And how about maximizing?

[blamario]

Disabling the popups is great, now is there a way to circumvent those annoying sites that maximize the browser window?

Re:And how about maximizing?

[Gerv]

Yep, you can prevent sites from resizing windows as well. It's possible in roughly the same way as popups, with the same level of control. You'd need to ask in the newsgroups exactly how to do it.

Gerv

Need to Reclaim Real Estate

[idonotexist]

Menu item spacing is larger for Bookmarks Menu. Vote for this bug.

Re:Need to Reclaim Real Estate

[jesser]

Or stop whining and use a different theme!

Two responses:

1. This was a recent regression, and I think it was an unintentional one. The bookmarks menu used to use the same height for menu items as other menus.

2. Configurability is good, but sane default behavior is better. This is especially true here because the setting "height of bookmarks in menu" is part of a theme, which can't be mixed with any other themes. (Also, it would look silly to have "height of bookmarks in menu" listed in preferences.)

Just wait until sites do this...

[Ryu2]

The obvious workaround for advertisers desparate to be as annoying as possible:

var w = window.open(...);
if(w == null)
window.location = "popups_required.html";

redirecting you to a message telling you that you need to enable popups to use the site

Re:Just wait until sites do this...

[Gerv]

In which case we hack window.open() to return a value indicating success :-) It may even do so already.

They can't win, you know...

Gerv

Re:Just wait until sites do this...

[Evangelion]

You do realize that "returning a value indicating success" in this instance is returning a valid javascript window object, don't you?

Fine, hack it to return new Object(), then the code checks to see if it had a document property, etc. Fine, hack it to include a document property. The code checks to see if it's a real document object, or just another new Object(). Add a document object. Then the code is going to check the existence of javascript objects generated by code that's on the popup's page. Fine, then - interpret the page, but don't render it. Then they check for things like window stacking order, and UI events. Fine then, they require interaction with the popup to get into the site...

And they win (assuming your goal is to get into the site without popups).

Re:Just wait until sites do this...

[scrytch]

In which case we hack window.open() to return a value indicating success :-) It may even do so already.
Until the popup itself contains a script that directs the actual site content to load in the original window. Or they just use an interstitial and use unique requests coded on the timestamps of the interstitial (thus defeating your cache while they're at it). Or they can put the ads in flash and use liveconnect to make the last frame of the flash load the page ... now you have to hack on flash (though you could probably have a filtering proxy munge the flash).

They can't win, you know...
They don't have to against the minority population of powerusers. Just Joe Average. But ultimately you can always vote with your feet.

Re:Just wait until sites do this...

[Malcontent]

The chances or this happening are low because the chances of IE implementing the same feature are zero. IE is designed to deliver advertising to windows users not to make your life easier. As long as the sheeple continue to use IE in massive numbers you can continue browsing without those annoying popups. Once again IE (and windows) is a kind of a stupidity tax.

Re:Just wait until sites do this...

[WWWWolf]

var w = window.open(...);

if(w == null)
window.location = "popups_required.html";

user_pref("capability.policy.default.Window.loca ti on", "noAccess");

Problem solved. =)

Re:Just wait until sites do this...

[Gerv]

Create the popup, then - just create it at coordinates 100000, 100000, and destroy it after 10 seconds.

Gerv

Re:Just wait until sites do this...

[lavaforge]

But ultimately you can always vote with your feet.

You mean by giving the advertisers a swift kick in the ass?

.94 & previous releases seem to have memory le

[ILoveMandrake]

This seems to have been a re-occuring problem among all of the releases, but even worse with .94 (.93 seemed to be getting better). Here are my specs : 1.2 ghz athlon w/ 266 fsb, 266MB PC133, gf2 ultra, sblive 5.1...)

I started mozilla having 136 MB with the buffers & cached memory. After about an hour or so of watching realplayer videos in cnn & abc and going to various sites I had only 85 MB free with buffers & cached memory included. I noticed a very aparent slowdown in the way programs loaded. I closed the window with realplayer with no ram gain. I then clozed mozilla. and my free memory went back up to 135 or so.

This may be caused by the fact that I have 4-5 windows open at once. I later tried using the "flush memory" option which only freed up about 5-10MB. The older versions seemed to have much less of this problem, but it was still noticeable. I use opera 5.05TP1 and it not only loads in a second or two, but is much faster at loading web pages and allows you to have up to 10 or more pages open at once in one opera window. This is quite nice for the way I browse. Unfortunately It does not have real player support, but id does have java & flash suport.

Too late for mozilla

[ScrewFinger]

The WORST thing about mozzila is that you can not even build the thing w/o having to be connected to internet to constantly checkout files from thier CVS. If you are on a modem give up. Why can every other open source project have a owrking configure script (no matter how complex, look at gnome & kde) but mozzila has some mostly broken thing that you need to use a web page form to make some shell script for you? Its pointless and counter productive. Once I found I could run Konquer w/o having to have the KDE desktop loaded (eg Konqueror in WindowMaker) and its far more efficent

Re:Too late for mozilla

[unapersson]

That's funny, I managed to download and build Mozilla source (keeping it up to date using CVS) using a 28.8 modem. I just downloaded the huge source tarball first, converted into a CVS tree, then updated it that way from then onwards. This resulted in a very small nightly download, instead of 8-10MB. Building is easy on Linux using gmake if you just follow the instructions.

Speed ...

[konmaskisin]

This release takes 17 seconds to start up on a pII-233 (about 1 second less if you use a reall small bookmark file). Sure, that's a bit slow (given the Netscape 4 takes ~ 12 seonds and Opera takes ~ 7 seconds. But 0.9.3 took 24 seconds (constistantly) so thats about a 30% start up time improvement and there's more to come. I'm not sure what the estimates are but 4 seonds faster on my machine is likey possible and that's without the BRUTAL_SHARING stuff in mozilla yet (which will make it faster) or improvements in gcc, glibc and ld which will likely get startup to ~10 seconds or less on a machine like mine.

Rendering pages is extremely FAST but creating windows is SLOW. The main hitch I have right now is on new window creation (which takes a long time to do). For example on a test page that uses javascript to open and close 75 windows one at a time (see the super simple code at this URL and either copy and make you own test or click on the link on that page):

http://206.191.52.79/MozTester/pagebanger.html

On a P233 running Linux I get the following (you'll likely want to try this on a faster machine - it's the relative comparisons that are interesting).

* Netscape 4.7.* takes about a minute
* Opera takes about 15 seconds
* Mozilla takes about 5 minutes !! (actually I stopped timing it's so slow)
* KFM/Konqueror ?? (old version doesn't work try it with KDE 2.0)
* Galeon ??? (not timed recently - the sort of more "native" GTK GUI might be faster??)
* Embedded Moz etc.
* Other browsers??

On MS Windows the Mozilla GUI is likely faster (haven't tested) and IE of course is very fast ... However IE 5.0 seems SLOWER on rendering pages and only really flies better on creating new windows.

Some of the slowness is due to the server so I engourage you to create your own javascript tests that just openm and close blank windows or something ... but the main slowness in Mozilla comes in drawing its own GUI ... Other than that the performance and speed of Moz is pleasantly peppy even on old machines (though lots RAM is recommended).

Re:Speed ...

[ksheff]

I tried this with Galeon 0.11.4 and it took about 45-50 seconds to loop through the test. At the end of the test, the browser opened the last two windows and just sat there for a minute or so and didn't redraw anything. I had thought it crashed at first.

If one has the "Open Popups in Tabs" option set, the test takes about 7 seconds.

Re:Speed ...

[kevryn]

I tried Netscape 4.78, Mozilla 0.9.3 (haven't installed the new release yet, figured I'd wait a day or two for things to slow down), and Konqueror 2.2 (within KDE 2.2) on a 900MHz Athlon with 128Mb RAM running FreeBSD.

Netscape took about 10 seconds
Mozilla took a bit over a minute
Konqueror took about 5 minutes, then crashed on page 74.

Re:The browser is great, but where is the spell ch

[BZ]

yes. now all you need is a way to call it from the browser and somehow usefully use its output...

Oh, and aspell/ispell is Unix-only,no?

That would be a good warning to me...

[Greyfox]

Not to enter the site.

It's amazing how much less irritating browsing the web is since I disabled popups and animation in Mozilla.

Re:Speed issues. Moz 9.3/9.4

[BZ]

> Konqueror does a much better job than Netscape

> 4.x of doing styles right

True. That's really not saying much. I've tried Konqueror and it's style and DOM support is pretty crappy as of last month...

The perfect feature...

[Kozz]

...which will strengthen Mozilla's place in the browser market as the premiere pr0n-browsing tool.

Geeks, rejoice!

Re:The perfect feature...

[Salsaman]

Yes, and don't forget ctrl-Q, which will instantly close all windows ( in case of emergenecy ;-) ).

Re:Speed issues. Moz 9.3/9.4

[Chakat]

It's because Mozilla is still not running builds with greater optimization (-O2) (reference). Once they finally get all of the bugs worked out so that they can go full steam, I'll bet you'll see a huge difference in the speed of the browser. Give 'em time. Mozilla is still beta, albeit in the very late stages of beta, so they're still getting everything working. I'll bet though that in the next month or two you'll see a huge speedup.

Just can't wait for it to finish downloading so that I can test it out. .9.3 has a fairly significant memory leak that bites you on the ass after time; it gets me somewhat quickly the way I browse - opening most links as new windows. Other than that, it is for the most part rocking.

Re:Mozilla plugin support?

[ksheff]

I have all of those plus the Crossover plugin and they all work fine, even in Galeon. The Java plugin for Mozilla is different than the one for NS 4.x. You may want to check to see what your system is using for the MOZILLA_FIVE_HOME environment variable. If you installed it using the rpms, that is set to /usr/lib/mozilla and the location for system wide plugins is going to be /usr/lib/mozilla/plugins. If you want to put them someplace else, I _think_ you can as long as you set LD_LIBRARY_PATH to include the directories where you actually installed the plugins.

Re:[button] tag behavoir is whacky!

[rjamestaylor]

Yes, turn it off thusly:

onclick="[some javascript];return false;"

That should do the trick... the return false cancels the button click.

why i'm not solely running Moz...

[xuvetyn]

note: i'm predominantly running windows lately on my desktop machine. don't start with me. it's the curse of being a gamer.

last year i ran in to NetCaptor (http://www.netcaptor.com), which, uses IE and, among other things, is "tab-able", and i simply can not go back now. (i'm addicted to tab-able apps. PowerShell rules! having 20+ windows open at any given time doesn't). so, my suggestion to developers is an add-on app that incorporates Moz for this. i'm sure i'm not the only one that would love to see this.

just my .02

Re:why i'm not solely running Moz...

[chavo valdez]

There is a project on mozdev called MultiZilla.It's still alpha and I haven't tried it though.

Keep your head up

chavo

Re:why i'm not solely running Moz...

[bigredlinux]

try multizilla (mozilla.org homepage link) for tabbed windows. Also, if you used galeon in Linux, you would get this same effect. Everyone is giving the option for tabs now because it is becoming excedingly popular, but it will always be an *option* since not everyone likes them.

Re:I must be stupid

[jchristopher]

It found the old prefs automatically and offered to import them on my system. Maybe try starting from scratch with the latest build and try again?

Get MultiZilla

[Dracos]

Go to mozdev.org and install MultiZilla. It puts tabs in Mozilla.

Nice.

[bee-yotch]

"including the ability to disable the JavaScript window.open() method"

Awesome, this'll be great for the free porn sites! ;)

Very little is UNIX-only

[yerricde]

yes. now all you need is a way to call [aspell or ispell] from the browser and somehow usefully use its output...

Pspell is a portable C library providing an interface between apps such as Mozilla and several varieties of spell-check backends (such as Aspell's English algorithm or Ispell's language-independent algorithm), along with command-line apps that call those functions. It's licensed under GNU Lesser GPL.

Oh, and aspell/ispell is Unix-only,no?

No. Pspell is a cross-platform library, and even though Ispell is tuned for POSIX systems, Cygwin provides a good POSIX layer on Win32 systems. With the port of XFree86 4.10 to run on Windows 98/ME and Windows NT/2K, it's very hard to call a piece of source code "designed only for UNIX systems" anymore.

Re:Very little is UNIX-only

[BZ]

OK. All we need now is someone to write something using Pspell to do spell-checking. :) You want to do this? I'm serious. This issue has been bugging me and I don't have the time to learn the spellcheck code while working on other phases of the project.

Also, is pspell easily portable to Mac? I don't know how MacOS Classic's POSIX support is.... and I don't know whether there is something like cygwin for it...

Linux 2.4.x/XF86 4.1.0 users have reported WfM

[yerricde]

Yeah... Mozilla is just useless on Linux (again... this counts for all milestones): X Error of failed request: BadMatch (invalid parameter attributes)

Users on recent Linux 2.4 and XFree86 4.1 have reported WORKSFORME. Do any other apps that link to GTK+ work?

Opera rocks!

[Dog and Pony]

I'm just now playing with the new release, Opera 5. It is so great... all that is missing far as I can tell is some IE-dependant javascripts... which shouldn't be needed, but all you web developers (yeah, you) are using them because they are "cool". :)

Anyways, opera is free if you agree to have an ad in your top-right corner. It is more than worth it - mozilla can go hide... they lost it a long time ago, open source or not.

And boy is it fast and stable... and since it has email and icq clients built in, I'm moving fast into being able to drop windows altogether, which will be a must as soon as 2k isn't keeping up anymore. I am not moving to the gummibear OS no matter what...

Re:Opera rocks!

[El Prebso]

For some of us Opera isn't an option, because we actually care about the license. I would have preferred Mozilla being GPL'ed, but I settle for the MPL.
Opera isn't free, it just doesn't cost anything.

Re:The browser is great, but where is the spell ch

[Swaffs]

"I need ALL the help I can get."

"in to" --> into
"acording" --> according

Hope this helps!

I wonder why advertisers don't realize...

[Dog and Pony]

... that it is because they do this stuff, ad revenues are going down?

What about Real?

[matty]

It used to work, now it crashes Mozilla (since about 0.9.1, I think). When I click on a RV link (like at espn.com) the window opens where Real is supposed to go, then all the Mozilla windows just go away. *poof!*

BTW, I'm using Debian Potato and I install Mozilla into ~/mozilla and drop rpnp.so and raclass.zip into ~/mozilla/plugins.

Any ideas???

User-Agent: that tricks servers but not designers

[yerricde]

It just makes life harder on web designers. How can we optimize our HTML code to render correctly in your browser, if you lie to us about what browser you're using?

If your browser sends

User-Agent: Mozilla/4.0 (compatible; MSIE 6.66; This isn't IE but really Mozilla 0.9.4)

then the server will send IE content, but the web designer will notice the satanic "666" in the User-Agent field of the logs and know something is up. A closer inspection reveals a Mozilla browser.

Never mind...

[matty]

I downloaded everything fresh from scratch and all is well. Maybe something got corrupted somewhere....

Dillo

[Z4rd0Z]

Mozilla has definitely progressed a great deal recently. However, check their page for system requirements: 64 fucking megs of ram! For those of use who like the idea of a small, fast standalone browser, dillo is the browser to watch. It now supports tables, and although it's still quite buggy, as fast as development is moving, I think within a year they may have a usable product, which at this stage is blazingly fast.

Re:Hangs on startup is probably sound "bug"

[rasjani]

This is typical error if you have flash plugin installed or possibly java (I havent really verified this).

Flash plugin opens /dev/dsp in initialization state and depending on your sounddrivers, it wont open if you have, for example xmms (or anything else using sound devices) running. Someone also mentioned that java does the same but i havent noticed this myself.

So, here's a list of what you can really do:

• get rid of offending plugins.
  
• Stop all sounds when starting mozilla
  
• Use esddsp to lauch mozilla
  
• Get Alsa. Oss that comes with stock kernel doesnt support soundmixing and thus, multiple instances cannot use /dev/dsp "simultaneously".
  

Re:What?

[ScrewFinger]

I don't want a binary. I want a single tar.gz like every other damn thing on the planet and I want to configure it w/o the aid of thier website

Omniweb have it

[zauc]

Nothing new here, in the preferences of Omniweb ( http://www.omnigroup.com/applications/omniweb/ ) there's radio buttons to select when a script can open a new window : always, when a link is clicked, or never. It's awesome, I'm glad Mozilla thought of it.

Moz/Netscape latest is a disaster...

[OSgod]

by a classical defintion -- first release of Netscape was technically challenged. No release of Mozilla exists -- it exists on beta only. The time since the last code release of Netscape is growing ever longer.

It is a disaster today. Until they *RELEASE* a product they have no product. Waiting for perfection is not the answer -- MS does at lease one thing right -- they ship.

Re:Moz/Netscape latest is a disaster...

[Gerv]

So what's Netscape 6.1 (released early August, to widespread acclaim) if not a product?

Gerv

Just tried 0.9.3 - Performance is TERRIBLE

[cdn-programmer]

I just installed 0.9.3 There is a LOT there that looks really good. I see some very very excellent ideas in this browser and I personally will switch to it as soon as the performance gets up to par. Nevertheless I'm aghast at the TERRIBLE performance.

I can see where some of this comes from... Part of it is the user interface. There is a generic level of abstraction here that takes a LOT of CPU time and memory. I agree with the abstraction but I can not for the life of me understand why it should literally take SECONDS in some cases for a mouse click to be resolved to a specific function.

At worst - in a given window if even a 1000 separate functions could be selected - then even a linear search through 1000 fields should be instant. These machines run at millions of instructions per second. There is no excuse here other than sloppy coding. An indexed search should take milliseconds. Clean up the GUI first.

Next threading. People typically open GOBS of windows. A separate thread should handle each window IMHO. I can see activities in one window blocking another. I think the underlying reason is that the underlying rendering code is not thread safe. This is my guess - but if so the project will really suffer until this problem is rectified.

Memory management. It looks to me that there is no attention to collecting the memory associated with a given window. It looks to me that 1000's of tiny objects are being malloc()ed with the result that the memory for one page will be interspearsed with the 1000's of tiny objects malloc()ed with other pages.

If 90% of the memory in a given page is malloc()ed for an idle window - and only a single tiny object is malloc()ed for the active window or controlling code - then that page can not be swapped. Effectively inattention to memory management in a browser that opens literally dosens of windows means that the virtual memory subsystem of the computer is neutered.

In addition - malloc()ing 1000's of tiny ojects with the default malloc routines found in systems like Linux is not IMHO a good idea.

The ususal malloc() code searchs for similar sized "holes" which means that (1) it burns a lot of unnecessary CPU and (2) it prohibits the OS from paging the majority of the memory used for idle windows because this memory gets allocated like someone shuffling a deck of cards where each suit represents the memory for a given window. (if you want to use the card deck as an example - then think of 4 windows and 4 suits. If you have 8 windows - then think of 2 decks, etc). A much better idea is to collect the cards in each suit into their own pile and KEEP THE PILES SEPARATE.

There are leaks - but when a window is closed the leaks from the code that malloc()ed ram for the said page is not freed. How can threading be accomplished and how can working with the virtual memory system of the OS be accomodated when memory is allocated amoung all pages like someone shuffling a deck of cards? IMHO this is an area where HUGE performance improvements can be gained.

I think the developers need to address (1) how threading can be layered into the system so that a single thread runs a given window - or perhaps a forked process runs a given window depending on the os... and (2) how all memory for a given window - hense given thread can be collected into a group of virtual memory pages associated with only said window/thread - so that the OS can swap them!

This I think means that people are going to need to look at some fundemental structural issues. Solve these issues first. Once a code base that is multithreaded with thread safe subsystems like the rendering engine and GUI engines are in place - then the rest of the system should fall into line. On the other hand - if the structural issues are not addressed - then I can't see how a multi threaded version will ever come about.

I personally think that a multi threaded code base will solve a HUGE pile of problems. If the developers agree - great. If not - well - its your code base and we each are entitled to our opinions. This is just my 2 cents worth here.

Re:Just tried 0.9.3 - Performance is TERRIBLE

[asa]

It's your codebase too should you decide to jump in and write some code. If you have performance insightss that you think warrant action please post them to Mozilla's performance newsgroup

--Asa

memory leak

[mstich]

This build appears to have fixed that nasty memory leak that was apparent in 0.9.3.

Web browsing speed.

[saintlupus]

why must a modern web browser run so slow??

It's not a matter of "must." Internet Explorer flies on my iMac. So did Konqueror when I was running Yellow Dog Linux. And so do OmniWeb and IE when I reboot into OS X.

It's not modern web browsers that run slowly. It's Mozilla, and in my experience, just Mozilla.

--saint

Am I the only one

[espresso_now]

that noticed definite weirdness with the Classic Theme on the first run?

Re:don't be too excited about popups

[Gerv]

I'm not certain (check the newsgroup post) but I think it disables them during timer events as well.

Gerv

Re:turbo mode in 0.9.4

[Gerv]

It's turned on by default (in installer builds only) because we would like more people to test it. You have an option of turning it off in the installer.

Yes, we are trying to make the code more efficient as well - turbo is not a substitute for performance improvements.

Oh, and the mozillaquest.com article is full of it.

Gerv

Re:What?

[Gerv]

There's a sourceball on ftp.mozilla.org and you don't need the website unless you want weird build options. The command you are looking for is:

gmake -f client.mk build

Hope that helps :-)

Gerv

What is NOT fixed

[Tony Shepps]

There are a whole slew of bugs in textarea handling that make it very difficult and painful to compose posts in textareas. This is not so critical for people who don't, uh, participate in large discussion groups such as this one. But if you DO post to Slashdot etc., you may want to hold off -- until 1.0, which is the target for just too many of these bugs, IMO.

I won't link in the bugs since the Moz folks don't like us to /. their bugzilla, but I'm talking specifically about such bugs as 83650, 82151, 88024, 68331, 75629, and 74383. And, for example, as I compose this in Moz, I know that if I accidentally hit the right arrow button at the end of the post, the cursor will move to the top of the post. It's painful.

Using mutt with mozilla?

[Dwonis]

Does anyone know where I can find out how to set up Mozilla to use mutt as its mailto: handler? Google couldn't tell me anything useful.

Re:Using mutt with mozilla?

[Dwonis]

I have it installed (and I also have mozilla-mailnews installed, because some users on my system use it), and it doesn't do anything yet. Isn't there some per-user setting that needs to be made?

Re:Using mutt with mozilla?

[vs]

For Communicator, there was muttzilla. I don't know if this will work with Mozilla.

And on Konqueror...

[hetz]

Click on Setting, Configure Konqueror - and on the window click on "Konqueror Browser" -> Javascript - now look below.. Yup - JavaScript web popup policy...

GUI - what makes things much easier then you think..

I hope some day that Mozilla will have some "User Agent" menu to bypass some web sites Insisting Explorer or nothing - Don't know why Mozilla people are not adding this. (and no - when a big web side web-admin gets a request to support Mozilla - it's being ignored almost always
)

Re:Speed issues. Moz 9.3/9.4

[rabidcow]

Any time that you rewrite 100% of the product, you can expect the new version to be slower, more infested with bugs, and just "feel" worse than the older version which has been tended for many years.

I disagree, sure there'll be more bugs (more new code), but in fact a complete rewrite should in the end be faster and cleaner. If they somehow made the old Navigator code compliant with the standards, it would be slow nasty spagetti code. A rewrite lets you untange that mess. If it feels worse, maybe you need a different chrome?

The first time I saw the "Deeply Nested Tables" demo, I was shocked! Mozilla is MUCH faster than 4.6 laying out tables alone.

Re:Speed issues. Moz 9.3/9.4

[konmaskisin]

This release takes 17 seconds to start up on a pII-233 (about 1 second less if you use a reall small bookmark file). Sure, that's a bit slow (given the Netscape 4 takes ~ 12 seonds and Opera takes ~ 7 seconds. But 0.9.3 took 24 seconds (constistantly) so thats about a **30% start up time improvement*** and there's more to come. I'm not sure what the estimates are but I heard 4 seonds faster on my machine is a very likely possible and that's without the BRUTAL_SHARING stuff in mozilla yet (which will make it faster) or improvements in gcc, glibc and ld which will likely get startup to ~10 seconds or less on a machine like mine.

Mabye I'm dreaming but it seems to me that rendering pages is *extremely FAST*. It's only creating windows opening dialogas etc that is SLOW. The main hitch I see now is on new window creation (which takes a long time to do). For example on a test page that uses javascript to open and close 75 windows one at a time (see the super simple code at this URL and either copy and make you own test or click on the link on that page):

http://206.191.52.79/MozTester/pagebanger.html

on a P233 running Linux (you likely want to try this on a faster machin it's the relative comparisons that are interesting).

* Netscape 4.7.* takes about a minute
* Opera takes about 15 seconds
* Mozilla takes about 5 minutes !! (actually I stopped timing it's so slow)
* KFM/Konqueror ?? (old version doesn't work try it with KDE 2.0)
* Galeon ??? (not timed recently - the sort of more "native" GTK GUI might be faster??)
* Embedded Moz etc.
* Other browsers??

On MS Windows the Mozilla GUI is likely faster (haven't tested) and IE of course is very fast (does the above in ~ 15-20 seconds) ... However IE 5.0 seems SLOWER on rendering pages and only really flies better on creating new windows.

Some of the slowness on a test page like the one above is due to the server so I encourage you to create your own javascript tests that just open and close blank windows or something ...

If the main slowness in Mozilla comes in drawing its own GUI this seems like a GOOD thing since opitimzing on that part of the beast hasn't even really started in earnest. It makes sense to me that the GUI would be the last thing to be sped up since rendering speed and correctness are the first job.

Galeon w/ Mozilla 0.9.4

[PenguinX]

Just so that everyone knows - it appears that the CVS variant of Galeon works fine with Mozilla 0.9.4.

will it ever function correctly?

[gruntvald]

I was so thrilled at 0.9.3 - it was sooo much better than any build I'd seen before. I got all psyched up to start figuring out what I could now do with this application in my company. Today, I installed 0.9.4. What a fuckin' crock of crap it is. Page rendering takes nearly a minute. This is Windows 98, I haven't checked if it's bad on other Windows, and Linux. However, my temporary excitement has gone. Maybe some day this project will turn out something worth working with. In the meantime, though I refuse to install it on my kids game machine, my dad installed ie6. I think the bar has been raised far beyond the capabilities of the mozilla development team.

Re:.94 & previous releases seem to have memory

[ukyoCE]

It's probably realplayer. I keep Mozilla open all the time, and am always surfing with 6+ windows open at a time. no problems here. but I also use basically no plugins, no flash, no realplayer, etc. If I want to watch streaming video I keep it in a seperate program, so it doesn't crash my browsers when those buggy things go down.

Re:Speed issues. Moz 9.3/9.4

[WhiteKnight07]

As far as speed goes, v0.94 is a major improvment IMHO. Mozilla 0.93 took about 6 seconds to start on my computer. The 0.94 milestone starts in 2 seconds flat! Although v0.94 does use about 29,500k of memory. (according to taskmanager neway) But that doesn't really bother me. Oh, and I'm not using the QuickLaunch option. IE 6.0 starts in about half a second. Pages of any size render instantaneously in both browsers. I would have Opera but I can't find a good cracked version of it. (I hate ads) I don't have Netscape 4.x cause Netscape 4.x is an obsolete piece of crap. <SoapBoxRant>MOZILLA RULES!! GO MOZILLA!!! DEATH TO IE!!! HAHAHA!!!</SoapBoxRant> ahem...

I've been using Mozilla as by default browser since v0.9 and it keepes getting better with each release.

Oh yea, my system specs:

1.4 ghz Athlon (266mhz fsb)

256megs DDR RAM

AsusA7M 266DDR Motherboard

Twin 75gig IBM Deskstars in ATA R.A.I.D. Level 0 (done with a Promise FastTrack TX2 PCI card)

Windows 2000

I can't help it! I'm a speed addict! :)

Re:Speed issues. Moz 9.3/9.4

[konmaskisin]

Hehe ... with such a powerful machine you should do yourself a favour and GET MORE RAM!! (256megs?? My *Mom* has 256 Megs and she uses her computer for "electronic mail" and "recipes").

Also do yourself a favour and install FreeBSD :-P

You're making assumptions...

[lines]

Who says they're using base 10? After 0.9.9, they could simply go to 0.9.A =)

hbz

Re:Speed issues. Moz 9.3/9.4

[WhiteKnight07]

I'm going up to 512megs in a couple of days. And up to a gig in a week or two. hehe....

Re:Speed issues. Moz 9.3/9.4

[WhiteKnight07]

Actually, I would install a few other OS's (linux, FreeBSD, BeOS, WinXP (just to see what its like)) but I can't figure out how to partition my raid array without losing all my data. Partition Magic only goes up to 80gigs and my array is basically a 143gig C drive. A shame really.... Hey, if anyone knows how to partition a 143gig drive without loosing all the data. Let me know.

Re:User-Agent: that tricks servers but not designe

[Phroggy]

Right, which means we have to keep checking the access_log to figure out what interesting tricks people are using these days.

I asked for that a while ago..

[josepha48]

There was an article here on /. and people were talking about popup windows, and I said that it would be nice to be able to disable window.open on a per site basis. I wonder how they implemented this. If it was a global thing for all sites or per site. But this is cook.

A good tool should do this for you

[yerricde]

we have to keep checking the access_log to figure out what interesting tricks people are using these days

Which is easy to do. All you have to do to find new tricks is extract the User-Agent field | sort | uniq, and "MSIE 666, really Mozilla" will out like a sore thumb in any "which IE versions are people using?" query.

-- Damian Yerrick, running nightly Mozilla builds until a milestone is released that includes the trunk fix to bug 30841

hmm...

[Ender Ryan]

What libc version are you running? Sometimes mozilla(among many other programs) seems to have problems with glibc 2.2.

You might want to try building it yourself.

junkbuster is even more "choiceful" than that

[hawk]

As shipped, junkbuster blocks nothing. It has no settings to block all banner ads, or to change content. As near as I can tell, it has *no* ability to change any content in any form.

What it *does* do is allow you to specify sites or pathnames from the root of any site that simply don't load. If it's an image, you get a broken image thingie. If it's a page, junkbuster tells you that the page is blocked.

It can also acti in a similar manner with cookies.
But it never, ever, changes things, and has no universal settings. They explicitly refuse to offer those (though they do tell you how to find sites where you can see what others have done).

hawk