Mozilla 1.2 Beta Released

nberardi writes "Mozilla 1.2 Beta is out. Typeahead now works on Mac and Java now works on Jaguar. On Linux, the classic theme now picks up GTK native theme. See the release notes for more info."

Comments

If only...

[sl3xd]

If only there was a theme that used the OS native widgets, without the ugly 'classic' icons...

Re:If only...

[dalassa]

The classic skin looks alot less ugly than the netscape 6 theme on OS X. Those round gray buttons are really jarring.

Re:If only...

[Anonymous Coward]

Help yourself. Unpack the classic theme (classic.jar is a zip archive) and replace the icons with your own.

Re:If only...

[unixmaster]

It does actually , uses GTK on Linux and native widgets on Mac/Windows when classic theme is selected.

Re:If only...

[sl3xd]

The reason is that the native widgets /should/ use less system resources...

Re:If only...

[evbergen]

Sadly, there isn't even any such thing as OS native widgets on Unix. Every toolkit has its own, and every application gets to choose its own toolkit.

We need an X protocol that works at widget level instead of pixel level. It'd be great if we could design /that/, together with a good widget definition language, and to stop reinventing the OS in huge toolkits that even provide timers and I/O and mistreat X as a dumb framebuffer backend.

Client-side rendering, high-level application frameworks, *yuck*. Provide your high-level GUI stuff through an IPC channel and get out of the way. Let me have my own main loop back. Thank you.

Re:If only...

[GreyWolf3000]

Actually I think that both GTK and QT are good enough development platforms to become "standard"--which is the problem, since neither will back down (neither should back down, since they both have really cool features).

Re:If only...

[ichimunki]

Personally, I'm anxious for the day it uses gtk-2.0 instead of gtk-1.4. I tried it with gtk2 and couldn't do any cutting/pasting (known bug, already in Bugzilla, I believe). Other than that it was great-- they're very close. Even better: once it is stable on gtk2, then Galeon 2 is ready to go. Either way, hats off to all Mozilla coders, Mozilla is a great browser and gets better all the time.

Re:If only...

[IamTheRealMike]

Provide your high-level GUI stuff through an IPC channel and get out of the way. Let me have my own main loop back. Thank you.

Right on. I'm waiting for somebody (it won't be me) to create an SVG based gui_server that accepts binary serialized XML on a socket, transforms it, renders it and synchs events back to the app. That would then in turn render either using X (with spiffy vector extensions) or OpenGL. Evas can do this. That would rock absolutely, and if you wanted to mix and match, you can use XEmbed as well. You could have XUL transformed to SVG :) Why not? Kinda like Quartz and Display PDF except better.

Re:If only...

[RAMMS EIN]

Three words: PicoGUI.

Re:If only...

[ncc74656]

If only there was a theme that used the OS native widgets, without the ugly 'classic' icons...

Phoenix looks like it's going that way. I would be using it right now instead of the new Mozilla beta, but Phoenix doesn't let you disable third-party cookies (you can't check the checkbox that controls third-party cookies, at least not under Win2K). Once they get that fixed, though, I'll more than likely switch over to Phoenix. All I really want is a browser. I use Mutt on my home Linux server for mail, so I don't need a mail client, and I use text editors (such as JOE or Notepad) for editing HTML and CSS.

The thing that bugs me right now about Mozilla 1.2b is that the Pinball theme doesn't work (it didn't work in Phoenix, either, and for the same reason...it hasn't been updated). Classic is ghey (as you noted), and Modern isn't much better. Pinball ought to be the default. :-)

Re:If only...

[oliverthered]

How about a common toolkit interface then, ok so QT is C++ and GTK is in poor mans C++ why can't they get together and sort the interfaces out, 2 toolkits fine, two seperate interfaces cheers mate just what I wanted!

Re:If only...

[uhoreg]

It only uses GTK/etc. to *draw* the widgets. It doesn't use actual GTK/etc. widgets.

Re:If only...

[Misch]

Agreed. Something changed between the 1.1 and 1.2 versions that prevents old skins from being used. I tried to use the Pinball skin when 1.2 alpha came out. It worked fine, except taht I couldn't open up the mail window some of the time on some of the platforms.

I wish I knew what changed between 1.1 and 1.2 on the interface... I'd try and go fix it myself.

Re:If only...

[Rev. Rudolf]

Same on Windows, for what it's worth.

The theme does a reasonable impression of native widgets, and in doing so will help new users feel comfortable with Mozilla, because it looks like what they're used to.

However, it is /just an impression/, a mock-up of the real thing, made from nothing more than XUL, CSS etc. Try /using/ Mozilla, and that illusion will quickly fade, as the widgets don't /behave/ like the real thing.

Re:If only...

[Bedouin X]

Maybe he's counting the Dot Org.

Re:If only...

[Bedouin X]

And guess what?

IE doesn't use native widgets.

pinstripe theme

[Anonymous Cowrad]

Note that if you're using the pinstripe theme, you've got to use the one made for nightlies.

I don't know why.

First thing I noticed.

Re:pinstripe theme

[Anonymous Cowrad]

The funniest part is that I am a homosexual Macintosh user.

I don't know why.

First thing I noticed was that I really liked men. Then I noticed that I really liked Macintoshes.

Re:pinstripe theme

[FaasNat]

Unfortunately the pinstripe theme currently only works on OS X. I would also love to use this them on my Windows and Linux box. Anything out there that's similar?

Re:pinstripe theme

[Dr. Spork]

What the hell is up with that theme? I tried installing it on Linux and it said I have to be running OSX. But why? Isn't the skin API totally cross-platform? I use the same skins on Linux and Windows, why are we making an exception out of OSX?

Re:pinstripe theme

[epsalon]

The API is cross-platform. This specific theme is designed to use OS-native widgets
As stated in the FAQ, the theme uses the OSX Appearance Manager to render the theme, and thus is designed only for OSX.

Link prefetching

[Rhubarb Crumble]

check this out: Link prefetching

seems to mean that if you're reading page 1 of a multi-page article, page 2 will be loaded in the background. nice!

Re:Link prefetching

[Anonymous Coward]

And if your reading a page with links you dont want to click, lets say to a picture of a man stretching his balloon knot open, then they'll be cached for you and swallow up more and more system resources. nice!

Re:Link prefetching

[PEdelman]

This sounds cool, but it looks like the page author has to specify what has to be pre-fetched. Due to the relatively small marker-share of mozilla, there will probably be few sites which implement this feature. Too bad, because it looks like a nice feature to me.

Re:Link prefetching

[Jugalator]

... provided the page is written for link prefetching explicitly. It doesn't mean you can go to a site like Google News and it will start loading the various articles in the background.

Perhaps that's good, although I'd like to see an option where you can choose to apply the feature to all links leading to HTML pages. This combined with a customizable maximum bandwidth restriction for the prefetching would be nice.

Re:Link prefetching

[Rhubarb Crumble]

And if your reading a page with links you dont want to click, lets say to a picture of a man stretching his balloon knot open, then they'll be cached for you and swallow up more and more system resources. nice!

That's a fair point - there is potential for abusem since the web page decides which "hints" to issue. Hopefully it'll eventually have an "enable prefetching for these sites"-type access control, similarly to the way it's done with cookies. Or a limit on the amount of data to prefetch.

Re:Link prefetching

[jlv]

On by default! No UI setting to disable this! ARGH.

(Yes, you can disable it, but you have to edit the preferences. Sheesh)

We are considering adding UI for this preference; however, the overriding theory is that if link prefetching needs to be disabled then there must be something wrong with the implementation.

This should be disabled by default, with a UI preference to enable it.

Re:Link prefetching

[Steveftoth]

What about a button to manually prefetch links on pages that you like? Not all pages should have their pages prefetched. For the advanced users, for the normal people it should be auto -on - or off.

Re:Link prefetching

[pubjames]

Perhaps that's good, although I'd like to see an option where you can choose to apply the feature to all links leading to HTML pages.

What would be great is if it could recognise if a page is of thumbnail images, and then automatically download the linked images. It would make browsing porn much quicker.

Re:Link prefetching

[trollercoaster]

Cool, this could lead to preslashdoting.

Re:Link prefetching

[GeorgePBurdell]

Seems like there's a lot of potential for abuse with this, especially given that right now you have to manually edit the prefs file to turn it off. What's to stop a page from tagging a really huge file, hosted on someone *else's* server as a "prefetch" item. Everyone who goes to page A starts "prefetching" from page B in the background - enough people do this and you've got a DOS going on.

Even if that scenario is not likely, I think it's still an odd choice for Mozilla - the philosophy behind the idea seems to be "the browser knows best and will think for you behind the scenes." On the one hand that sounds great: the browser will anticipate my next move. On the other, that doesn't sound so great... My cable modem starts blinking when I think I'm not grabbing anything and I get suspicious.

Re:Link prefetching

[JanneM]

http://leech.mozdev.org/

Re:Link prefetching

[Eimi Metamorphoumai]

How is this worse than just embedding the image into the webpage, possibly with height=0 width=0? When you go to a webpage you already pretty much give it carte blanche to download what it likes; this doesn't seem very different.

Re:Link prefetching

[Iamthefallen]

I still cherish the days with modem when I could see the goatse.cx pic load before my eyes and go away before it was done, now with DSL it's instant, and you think this is good?!

:)

Re:Link prefetching

[roca]

> What's to stop a page from tagging a really huge
> file, hosted on someone *else's* server as a
> "prefetch" item.

You can already do this by loading someone else's page into a hidden IFRAME.

Nothing new here. Move along.

Re:Link prefetching

[ObligatoryUserName]

How about limiting the prefetching to pages in the same domain as the page doing the prefetching. Perhaps you could explicitly allow addtional domains for prefetching in the head of the document.

Re:Link prefetching

[Ed Avis]

The style of web browsing I use is to load all the links I want to read in new windows by clicking on them with the middle button. Then they can be loading in the background while I read the first part of the article. It forms a kind of queue of pages to read, so when I've finished reading the first page I just close that window and go on to the next (which is ready instantly). The result is up to a hundred browser windows open at once - but I know that I'm not the only person who browses like this. Of course, it helps to have a browser which can open lots of windows without thrashing and slowing the machine to a crawl (like Dillo) or one that has tabbed browsing.

This style of following links can also work well with offline browsing and a proxy server designed for offline use like WWWOFFLE. If you go online briefly and click on all the links you want to load, the proxy remembers to download them. Then a few minutes later you can go online again and all the pages will be loaded ASAP. Once they've loaded you can disconnect again and continue browsing. This makes the most sense for people whose internet access is metered (hmm, I wonder if something like this could work for palmtops).

But what I'd really like to see in a browser is an explicit 'to read' queue. When you click on a link with the middle button, it doesn't immediately open in a new window or tab but instead is added to the queue and starts downloading in the background. On the browser's toolbar there is a 'next page' button which goes to the next URL you have marked for reading.

Automatic prefetching of all links from a page, la wget -r, would be crazy for many heavily-linked sites. But you could have heuristics for it or specify particular sites where the link following should be more aggressive.

Re:Link prefetching

[j7953]

Perhaps that's good, although I'd like to see an option where you can choose to apply the feature to all links leading to HTML pages.

No, that would be a very bad idea. Just right now in the navigation menu of the Slashdot page I'm viewing ("Post Comment"), there are 17 navigation links, plus the category links, etc. You cannot tell me that you'll be following all of those 17 links. Web sites (and probably ISPs as well) would not like such a feature due to the increased bandwidth costs they'd have to account for.

Also note that e.g. this page has a "log out" link that I really do not want to be automatically prefetched for obvious reasons. Granted, it contains a query-string so Mozilla would not prefetch it anyway, but I imagine there will also be web sites that have log out links without query strings in the URL. And there are lots of other actions that might be associated with following a link (think prefetched one-click-shopping).

The HTTP standard (RFC2616) states that "In particular, the convention has been established that the GET and HEAD methods SHOULD NOT have the significance of taking an action other than retrieval. These methods ought to be considered 'safe'", and if there are side effects, "the user did not request the side-effects, so therefore cannot be held accountable for them", but I wouldn't trust on web site administrators knowing this.

Re:Link prefetching

[HiThere]

Right. And so you go to /., open an article, and...

I see a use for a "load in background" click option. That could sometimes be very good. But "load all links"? No. Not even "load all links when selected". There's too much problem with hidden links already.

(Mind you, as long is Mozilla is the browser of a small minority this wouldn't be too bad. But once people start designing web pages to take advantage of this ... unh unh.)

Re:Link prefetching

[Lionel Hutts]

That's fine in theory, but, if it works correctly, why would anyone want to turn it off? Remember, it's only prefetching pages when the server says to. It also automatically interrupts prefetches if the browser is not idle.

Re:Link prefetching

[mpsmps]

Embedding an invisible image has a variety of problems.

1. Relying on obscure side-effects leads to bad code. For example, one could imagine a highly-optimized browser-rendering engine may choose not to read the bits of the image because they won't be visible. It's much better to have an XHTML tag that explicitly expresses the desired semantics and leave it to the presentation tool to properly figure out how to present.

2. Languages, standards, and practices evolve. For example, if my webages are XML interpreted by XSL stylesheets, do I really want to start embedding browser hints in my XML pages (or have my XSL stylesheet assume a browser is the client)?

3. How does the browser know not to start prefetching the image before it has loaded the main page? The prefetching FAQ says that prefetching uses an idle test to avoid doing harm. Embedded images can't readily be optimized by an idle test.

Re:Link prefetching

[SethJohnson]

Ed,

If you're using Mozilla, or the recently-released Phoenix (highly recommended), you can also accomplish your browsing style by right-clicking the links and selecting "open in new tab". The other page will open a new tab within your existing window. When you're done reading the current page, you can click on the tab for the other page without having to juggle windows.

What's nice about Phoenix in this respect is the default behavior is to have the new tab open in the background. I complimented the design team for this on their discussion board and some guy came back and said you can also set this up in Mozilla via the prefreences. It's supposed to be controlled by the checkbox 'Load links in the background'. You can also set middle or right-click to open these tabs in the preferences.

Seth

Re:Link prefetching

[aftk2]

Interesting idea, although it's too bad that, according to the FAQ:

URLs with a query string are not prefetched.

This really limits how useful this feature might be. I can imagine reading a multiple page article, and on page one of that article, the link tag prefetches page two while I'm reading page one, for quick access to the next page. Unfortunately, the URLs to most articles on the web contain query strings (a query string is the question mark (?) that preceeds a bunch of variables in a URL.)

Interesting idea, at least.

Re:Link prefetching

[zmooc]

Perhaps that's good, although I'd like to see an option where you can choose to apply the feature to all links leading to HTML pages. This combined with a customizable maximum bandwidth restriction for the prefetching would be nice.

And that, my friend, would be the end of the Internet. How many of the links on a website do you generally click? On slashdot, I think, it would at most be something like 5%. Let's say 5% of the users would enable this feature. Now their browsers start pre-fetching. Since they normally only click at most 5% of the links, preloading all would multiply their bandwith-usage by 20 times. So. Our 5% of the users uses 20 times as much bandwidth as they would without preloading. So the average bandwidth-usage for web-browsing would about double and that's with only 5% of the users having this feature enabled. Bye bye Internet. There's a reason this really simple to implement feature isn't there yet.

But.... combined with a reasonably large distributed network of caching proxy-servers, pre-fetching might be worth a try.

Re:Link prefetching

[Jorrit]

I would turn this off immediatelly even if it works correctly. I have limited bandwidth every month. I only want to load what I need. Not what the server thinks that I need.

Greetings,

Re:Link prefetching

[thegoldenear]

Mozilla focuses upon being standards compliant, rather than making Netscape and Microsoft's mistake of defining their own techniques outside of the W3C. (I think there's also some degree of themselves informing the W3C when its sensible to). so, is prefetch not a part of a W3C HTML specification?

Re:Link prefetching

[jlv]

If it works correctly, then let there be an option to turn it on.

Basically, they're inventing something new here, and if it isn't turned on by default and in everyone's face, it will never get used. It took 2 point releases of Netscape 2.0 to get them to add the ability to disable JavaScript.

IMHO, I don't care if "Netscape Navigator 7" has this on by default. Mozilla shouldn't.

Re:Link prefetching

[mblase]

How is this worse than just embedding the image into the webpage, possibly with height=0 width=0?

Because (a) height=0 and width=0 are not well-supported -- they'll probably display the image as 1x1 pixel or at full size; (b) presumably prefetching can be turned off on the client side if one so desires.

Re:Link prefetching

[prockcore]

"What's to stop a page from tagging a really huge file, hosted on someone *else's* server as a "prefetch" item"

What's to stop them from doing it now? With iframes?

Re:Link prefetching

[Erik Fish]

Well if there's an enable/disable for prefetching it's sure to be accessable via the best Mozilla plug-in ever!

Re:Link prefetching

[arkanes]

Off topic but relevent: When you open a new window from within an explorer window, with open link in new window or by hitting ctrl-n, then it's a child window of the original, and dies when the parent does. A totally new IE window, opened by invoking iexplore.exe, is a seperate process. As for why it works exactly like this - I don't know.

Re:Link prefetching

[damiam]

It's all in the preferences. Under "Tabbed browsing", select "Load links in background" and "Middle-click of links opens a new tab"

Re:Link prefetching

[LordNimon]

There's a big difference. With embedded images, the browser isn't done loading the page until all embedded images are loaded, and that can affect what you see and how responsive the browser is. Ever try using a web page while it's still downloading? It doesn't always work.

But with this new Mozilla feature, the web browser does the loading in the background, without disturbing you.

Re:Link prefetching

[idontneedanickname]

Naviscope does this much better, it prefetches links with next you can specify. You could for example, have it prefetch all links that say 'page 2' or 'next' and for slashdot 'Read more...' (of couse they would hate that :P )

Re:Link prefetching

[leuk_he]

Nice tool, But it is unsupported now for about 2 years. It does not have autoconfiguration for mozilla or opera above version 4.

Also it has no updates for the anti-ad stuff, what is its main purpose.

It is has had great potential, but th authos seems to have it abandoned,

Re:Link prefetching

[horza]

If you're using Mozilla, or the recently-released Phoenix (highly recommended), you can also accomplish your browsing style by right-clicking the links and selecting "open in new tab".

Even easier is just to click on the link with the middle mouse button, which will open the link in a new tab in the background. You can just scroll down the page middle-clicking on any links you want to read later, and not disturb the flow of what you are reading. Makes for a very addictive style of surfing. You can also middle-click on any tab to instantly make it go away.

Phillip.

Re:Link prefetching

[epsalon]

The W3C recommendation does not mention <LINK TYPE="prefetch"> as a link type, but it is allowed to add new link types.
The W3C recommends adding a profile attribute to the <HEAD> element to sepcify this.

GTK....

[GreyWolf3000]

Mozilla's binaries still depend on gtk 1.x, however when compiling from source you can tell it to use gtk2. I don't know how stable that is, though...

Re:GTK....

[cyco/mico]

As it seems, they are not there yet. But there's a patch from the galeon guys, who seem to be working on that too. You can find the respective hints here (galeon2 installation instructions).

Type ahead find is great

[fault0]

Type ahead find is great. Been using it since Moz 1.2 alpha. The neat thing is that you can type a search phrase, and you can search again with ctrl-G. My only suggestion would be to have type ahead and find searches appear in a history combobox in the find window.

Re:Type ahead find is great

[snake_dad]

In my nightly build the typeahead stuff caused the backspace key to stop behaving as a "backbutton-key" even when I've done no typing at all. I know that alt-left does the same, but one key is easier :) Does it still behave that way in 1.2b?

Mime Types

[nagora]

Have they done anything about adding large numbers of mime types (ie, made it not be a pain in the arse) yet?

TWW

And Blizzard Represents....

[unixmaster]

XFT support on Linux! Now we can get cool anti-aliased fonts on Linux!

You must compile from source with --enable-xft and need fontconfig & xft2 package from www.fontconfig.org and of course freetype2 from www.freetype.org

Great thnx to Chris Blizzard for this!

Oh btw now HTML for controls & scrollbars use your native GTK theme widgets when classic theme is chosen.

Re:And Blizzard Represents....

[Malc]

So how long until I can just download it and have it working straight out of the box. What you describe is too much effort (it's easier for me to stay in Windows).

Re:And Blizzard Represents....

[tempest303]

If you have Red Hat 8, you already have Xft2 and fontconfig installed and working, and Mozilla 1.2-final will ship with Xft2 enabled and (it looks like) GTK2 widgets, too!

So ideally, with a RH8 rig anyhow, there's really no effort at all. Just wait for the Moz 1.2-final RPMS to come out, install them, and voila! Beautiful font rendering, with no hassle. :-)

Re:And Blizzard Represents....

[scotch]

I have Red Hat 8, and among its other problems(*), the xft stuff seems to be broken. Text is written to random parts of the screen (on top of other windows, etc), some text doesn't show up until you move the window or scroll the text, etc. The fonts look nice (when they're drawn in the right place), but the rendering is noticibly slower (800MHZ p3), for example, when text scrolls in a terminal window.

* - Lack of configuration to a fault, for example.

Re:And Blizzard Represents....

[tempest303]

Guess I've never seen the bugs you're referring to, except in Abiword, where I saw something similar.

As for the "lack of configuration", it's not a fault, it's a Good Thing(TM). See here for why: http://www106.pair.com/rhp/free-software-ui.html

Re:And Blizzard Represents....

[npietraniec]

As long as you're installing freetype2, you can make sure to fix the bytecode interpreting

...The difference on RedHat8

Re:And Blizzard Represents....

[MartinG]

are you using a matrox card by any chance?
If so, there are known problems in the render accelleration IIRC. Try searching on bugzilla.redhat.com I'm not sure if thats where i heard about it though. if so add your comments, if not raise a bug!

Re:And Blizzard Represents....

[scotch]

It happens in gnome-terminals and other apps on my system. The text is sometimes written on top of the root window, on my panel, or anywhere else, really. I'm using xinerama, don't know if that has anything to do with it.

As far as lack of configuration, I've read the arguments, and I disagree with them. Linux has over the years attracted users that like having the ability to control their software (all the way from the kernel to the desktop). To piss on these users in favor of the proverbial joe six pack is shortsighted, IMO. If my companies business plan was to piss off existing customers in the hopes of attracting potential customers, I'd find another company to work for

Metacity, in particular (referenced in your url), is an interesting beast. There are umpteen million window managers out there, some guy rejects them all as too buggy or too configurable, so writes yet another window manager. Great. But in metacity, I can't turn off opaque move. This is unacceptable on slower systems. I also can't vertically or horizontally resize windows. This might seem like fluff to you, but it is very powerful and useful on a large (multi-headed) desktop. Keybinds in general suck in metacity. Sawfish, the now-ugly-stepchild of the gnome project, had all of this. Instead of writing YAWM (yet-another-window-manager), why couldn't the author of metacity work with the sawfish (or other existing WM) to work out sensible defaults and cleaner preference system? You can argue that all the configurations of existing window managers make them buggy, bu t this flies in the face of my experience. A new project like metacity is likely more buggy than a project that has been around, has all the features you want, and has been tested by thousands. A classic case of not-invented-here-syndrome, IMO, which plagues linux like a bad STD.

There are lots of nice things about gnome2 and redhat8. Lack of configuration is not one of them. Thank god you can still replace metacity with sawfish (although they've made it *HARDER* than in previous release - sheesh).

Re:And Blizzard Represents....

[scotch]

Thanks, I will - I've been planning to, I looked a bit at the xft and gnome sites, but haven't found anything yet. I do have a matrox card, btw.

Re:And Blizzard Represents....

[dmiller]

- Lack of configuration to a fault, for example.

Have you tried running gnome-font-properties? What more configuration do you want?

Probably a little redundant...

[MacOS_Rules]

Moz 1.2 works like a champ on my iMac under Jaguar. 1.1 was a little sluggish, but 1.2 seems to have corrected that and then some. Startup times are now nearly as fast as IE 5.2.2, and Moz is and hopefully will continue to be much less crash prone than IE. This is in and of itself amazing, considering it is 1.2 BETA.

Great job to all who work on this effort. It is much appreciated by many in the computing field.
Cheers!

Re:Probably a little redundant...

[BlueGecko]

I'm on a 667 MHz G4 PowerBook with 512 MB RAM, and while I dislike IE for the sole reason that it does not have popup protection, there is no way you're going to convince me that Moz 1.2 is loading as fast as IE. Chimera does, but not Moz 1.2. The Mac version of Mozilla proper still has a long way to go before I would consider foisting it on anyone.

Fast releases

[koh]

moz development has been considered sluggish by many a few months ago... now that they have the infrastructure right, they do release early and often. Nice :)

Too bad I'm still stuck to 1.0.1-r1 on my gentoo distro... ;)

Re:Fast releases

[rizzo]

Why are you stuck to 1.0.1-r1? Just unmask 1.1 and you'll get 1.1. You can copy the 1.1 ebuild and make a 1.2a ebuild. As soon as 1.2b source is released I'll be submitting a 1.2b ebuild.

You just need to unmask it by commenting out any mozilla lines in /usr/portage/profiles/package.mask. The gentoo people mask apps to create an aura of stability.

Re:Fast releases

[Mark Bainter]

It's even easier than that. Just emerge /usr/portage/path/to/pkg.ebuild

It won't check for masks if you explicitly give the full path instead of just the package name.

Type-ahead Find

[RPoet]

1.2 is really worth installing just for the Type-Ahead Find feature. It's one of those "how did I ever manage without it" features, and a punch in the stomach of anyone who says free software isn't innovating. This feature almost obsoletes the use of a mouse while surfing (well, almost). You see a link you want to follow, called "Click here". So you type "cl", and that link is marked. Now press enter to follow it. So simple, yet so efficient.

Re:Type-ahead Find

[Anonymous Cowrad]

Typeahead rocks my socks, but the Mozilla team didn't invent it. Internet Explorer for the Mac has had this for quite some time.

A trivial point, maybe, and I certainly agree that Mozilla is innovative, but they weren't first in this case.

Re:Type-ahead Find

[Loligo]

>Typeahead rocks my socks, but the Mozilla team
>didn't invent it. Internet Explorer for the Mac
>has had this for quite some time.

IE has had "fill in the box" type-ahead completion for years, but it sounds like what he's describing is different.

As an example, say you wanted to reply to this article. Instead of clicking on "Reply to This", you'd type enough of "reply" to jump the highlight to the link in the active window.

Not exactly the same thing. Not even remotely the same thing, even.

-l

Re:Type-ahead Find

[Corvaith]

It sounds good, sure. But, I can't be the only person out there for whom it's more of an irritant than a feature. On long pages, if you accidentally type something without focusing on, say, the form box... then it'll scroll you right down to the link it thinks you want.

I'm therefore waiting expectantly for the feature that lets you turn this *off*. I'm sure it's nice for some people, but if you don't want it, being forced to have it is a pain. If there /is/ a place to disable it... it's definitely not anywhere visible.

Re:Type-ahead Find

[Loligo]

>IE does have move-the-focus typeahead

Where? I'm on IE 6, and I don't get anything like the behavior he's describing.

-l

Re:Type-ahead Find

[Kidbro]

The fact that type-ahead works with non-links too (start your search by typing / (slash)), it shouldn't be much trouble. It's usually fairly easy to locate a unique, or at least unusual word near the link in question, and then tab or shift-tab once or twice in order to get to the link you want.

I'm trying this very feature out atm (using Phoenix though), and it works well, even on slashdot (takes a couple of "clicks" to get used to, but that's all).

Re:Type-ahead Find

[BZ]

IE/Mac and IE/Windows have nothing to do with each other. IE/Mac has typeahead find and has for a long time (and is generally a much better browser than IE/Windows).

Re:Type-ahead Find

[tjwhaynes]

Except I am browsing at 3 right now, and there are 29 "Reply to This" links, 4 viewable on my screen. Which means I have to sort through at least 4, or at most 29 links to get to the one I want. Seems easier to just move my hand.

Well - try identifying some unique word just before or after the link you are after.

In getting to and replying to your message, type

/hand

Hit TAB and the "reply to this" link is selected. Hit Enter and you're done.

Cheers,

Toby Haynes

Re:Type-ahead Find

[tibbetts]

It's one of those "how did I ever manage without it" features, and a punch in the stomach of anyone who says free software isn't innovating.

Now, I won't quibble with the belief that this is an example of free software's innovation, but type-ahead (well, vi-style, anyway) has been a feature of at least one even older free Web browser for a while now, if not others.

(For the record, the feature Without Which I Cannot Do has to be mouse gestures. Just click and drag to the left to go back a page!)

Great News

[MercuryWings]

I've been using Mozilla since the 0.6 beta days and count my blessings on a regular basis. It's nice to see they've added the GTK support - now it'll not only be a linux app, but will have the look and feel that is consistent with other GTK-based apps. That part tended to be irritating - didn't feel like GTK, didn't feel like KDE, felt like one of those 'let's design the entire interface to our own personal tastes' programs that one finds far too often on that 'other' OS.

One question I have though - does it support GTK 1.2, or 2.0 (including the anti-aliasing fonts feature)?

Re:Great News

[distributed.karma]

The Galeon page says that the Mozilla GTK2 interface is far from ready. However, there should be ways to enable AA fonts on older GTK, though it's quite unstable.

Beware of GTK themes

[Psiren]

Some of the themes I tried with GTK and Mozilla this morning crashed Mozilla on startup. Others were okay. I guess there are still a few bugs to work out there.

Re:Beware of GTK themes

[l0ungeb0y]

Well of course, that's why I"just say NO" to themes. OS themes, browser themes, any theme at all besides the defaults they come with.

Not because I don't like themes, but they are version specific for each release... and having to drop/change themes with each new release seems like more of a pain in the ass than it's worth.

Maybe someday in the not so distant future, they will build a theme utility that will adjust theme graphics to match the current GUI... but I doubt it.

Re:Beware of GTK themes

[roca]

Please file a bug about any themes you find which crash Mozilla. Or at least mention theme here on Slashdot.

Re:Beware of GTK themes

[Psiren]

Several of the Xenophilia ones. Can't remember which ones though. I ran a talkback build so I sent the crash data that way. Haven't got time to file bugs atm I'm afraid.

Link Pre-fetching is a baaad idea...

[Bonker]

Remember all those offline browsers and 'modem accelerators' that sucked up your modem bandwidth by downloading contantly, spidering every link on every page you visited?

While the Mozilla project is an incredible piece of work, I have to question this feature. It appears that they've designed it so that a page designer or webmaster decides what is appropriate for prefetching or not. Still, if used inappropriately, this feature could lead to more information being transmitted across the internet that is either discarded or unwanted. In a worst-case scenario, an inexperienced web designer might routinely run into his bandwidth cap or unintentionally force users who have bandwidth caps to exhaust their allowance.

If you can only download 3GB per month over your cable modem, do you really want the designer of a page deciding that your browser needs to spend time downloading ads or useless images?

For some people, this could be really useful. For others, it could be a real pain. Team-Moz, if you have any consideration at all, please adjust the default configuration of Mozilla so that this feature is turned OFF.

Re:Link Pre-fetching is a baaad idea...

[Anonymous Coward]

It requires the web-"designer" to hint, what pages to prefetch.
Since, the employer of named designer pays for the bandwith, it surely will be used only at adequate places (Cost/Benefit).

Re:Link Pre-fetching is a baaad idea...

[stratjakt]

Hmm.. How about web designers that:

- add a prefetch tag to the banner ads, making it look as though you'd clicked them.

- prefetch all those pop-up or pop-under ads, so you have a snowballs chance in hell of closing them faster than they can open. (Not all popups are bad - you shouldnt have to disable them completely)

- slashdot trolls embedding prefetch tags into their links to some site that I cant remember right now, something about goats.

There's lots of room for abuse, and it's enabled by default, and can't be easily shut off (easily as in through the UI - not every user is comfortable editting text files). If you think every webmaster can be 'trusted', you're naive.

Re:Link Pre-fetching is a baaad idea...

[Tack]

So far as I can tell from the prefetching FAQ, there's nothing built into this to keep it from, say, prefetching banner ads, which are very typically hosted on a different server than 'real' web content. Thus, the designer of a web page can force your browser to download ads for his benefit and without any cost to him.

This isn't an inherent problem with prefetching. You can do this with regular HTML now. Consider:

• <img src="bigbaduglybanner.gif" width=0 height=0>

Or, you might use CSS and set display:none (although I'm not sure if the browser will fetch the image in that case, but some might). Or, if you want to cause the client to load an html, use an iframe also with 0x0 dimensions. You can see there are ways to do exactly what you're worried about right now, in all browsers, without prefetching.

Jason.

Re:Link Pre-fetching is a baaad idea...

[Draigon]

Not sure if this was mentioned, but wouldn't it be possible to silently download a program to your cache? I skimmed through the FAQ and didn't see any blocks on file types.

Someone could download it to your cache then maybe find a way to execute it.

Re:Link Pre-fetching is a baaad idea...

[srmalloy]

Read the specs, please. It doesn't prefetch ALL links, only those explicitly set as such in the web page. Which, as far as I know, accounts for exactly zero web pages in existence today.

However, it only takes a minimal amount of underhandedness to start screwing people over. Banner ads are everywhere, and a large percentage of them are implemented by having a site drop in a block of code that references a CGI script on a server run by the company managing the distribution of banner ads. If the company running the banner distribution server decides that having their advertising clients' linked pages load faster is a valuable feature, all they would have to do is add the prefetch code to the output of their CGI script -- both Mozilla and IE will happily process a META tag in the body of an HTML document, even though by the specification, a META tag should occur only inside the HEAD tag block. So the user's network connection bandwidth would get usurped to prefetch the advertiser's web page, even if the user has no intention of clicking on the banner ad.

Re:Link Pre-fetching is a baaad idea...

[roca]

This point has been made elsewhere but it needs to be reiterated:

A Web page can already force you to download arbitrary files. For example, it can include a hidden IFRAME linked to some URL. This prefetching feature does not allow Web sites to do anything nefarious that they couldn't do before.

In fact, this prefetching feature is strictly better for users than hidden IFRAMEs or similar, mainly because prefetches are given bottom priority so they never interfere with your other Mozilla network activity.

Re:Link Pre-fetching is a baaad idea...

[roca]

Not if it's something that the browser can display natively, like an image or a Web page.

Re:Link Pre-fetching is a baaad idea...

[roca]

Bad news: they can already do this in any browser. All they have to do is include a zero-height IFRAME in their banner that SRCs the linked page. This is actually a *lot* worse than Mozilla's prefetching, since it will compete for bandwidth with other page loads and it can't easily be disabled.

Mac OS X Users should ignore Mozilla

[toupsie]

Of course, lets heap praise on the Mozilla developers for their hard work. Without it, we would not have the best browser for Mac OS X, Chimera. Mozilla is a full of bloat that Mac OS X users don't need. We already have iChat, Address Book, Mail and iSync built into our beloved UNIX operating system. So a lot of Mozilla's functionality is not needed -- newsgroups are nice but we have better alternatives. Chimera is what Mac OS X users really need. Its blazingly fast, supports standards and gives Microsoft Internet Explorer something to aspire too. Poor Omniweb never knew what hit them.

Get the latest nightly build here!

Re:Mac OS X Users should ignore Mozilla

[analog_line]

The next time I read someone in all seriousness using the term "blazingly fast" the baby seal gets it.

Chimera's great, I use it, but it has zero speed advantage over OmniWeb. Nil. Goosegg. And in my opinion, OmniWeb looks better.

Re:Mac OS X Users should ignore Mozilla

[Macka]

I tried Chimera and didn't like it. Way light on features and just didn't *feel* very nice. So I settled on Netscape 7 + the popup blocker. Now it's the only browser I use and I'm very happy. Mozilla is the proving ground for future versions of Netscape, so I fully support the work they are doing.

Re:Mac OS X Users should ignore Mozilla

[sporty]

What about the XUL support that mozilla provides? I like my neat little google bar, the dom inspector and thigns like that, which mozilla has.

Plus, those apps aren't built in, they are added on. If you don't like iChat, drag it to the trash. We dont' need no steekin' uninstallers.

SO maybe instead of ignoring our options, how about we dismiss them as we please. That's the point of having a choice, nuh?

Re:Mac OS X Users should ignore Mozilla

[bstadil]

What about the XUL support that mozilla provides?

Good point. Try RadialContext ig you haven't already. I love this feature. Wish that this could be incorporated into normal Apps. This is a much better approach than a normal toolbar.

Now if only the spellchecker worked with 1.2b ;-)

Re:Mac OS X Users should ignore Mozilla

[roca]

Be careful with your terminology. Chimera is developed by Mozilla.org and is 90% the same code as the Mozilla application suite.

Re:Mac OS X Users should ignore Mozilla

[toupsie]

Be careful with your terminology. Chimera is developed by Mozilla.org and is 90% the same code as the Mozilla application suite.

That's why I kissed Giant Lizard butt at the beginnning of the post.

Re:Mac OS X Users should ignore Mozilla

[iomud]

My only problem with chimera is it's stability, it crashes constantly. Once the bugs are ironed out and it gets a little more feature complete I think it'll be my browser of choice. I'd really also like to reduce the size of the icons even when they're in small mode it still takes up way to much real estate for my tastes. I'd much prefer something along the lines of ie when configured to use tiny icons.

Re:Mac OS X Users should ignore Mozilla

[sporty]

My resolution is too high. The icons look like smuges. But what doesn't work for me would work for someone else :)

Re:Mac OS X Users should ignore Mozilla

[toupsie]

Sound like you have the same complaints I do with Chimera. However, I find that all browsers on Mac OS X crash more than I like. But since Chimera launches in mere seconds on my Mac, it has become my choice. You are right about the interface, IE is much cleaner. I really need to sit down and hack up a different one.

Re:Mac OS X Users should ignore Mozilla

[Avumede]

I actually use both regularly, and Chimera feels about twice as fast for me. However, I usually still use Omniweb since it looks better, and has better support for international webpages.

Re:Mac OS X Users should ignore Mozilla

[SmittyTheBold]

We already have iChat, Address Book, Mail and iSync built into our beloved UNIX operating system. So a lot of Mozilla's functionality is not needed

Windows users have no need for Mozilla, either. After all, they have Outlook[Express], MSN Messenger, and all the other MS tools built-in. Hell, they have Explorer integrated, even! Therefore, no portion of the Mozilla project is even remotely useful to the Windows crowd.

I'm sure you can admit there are definite reasons for alternatives, even when the already-available software is best-of-breed.

Moz versus IE

[0x0d0a]

With the two rapid releases of Phoenix and Mozilla, with Netscape (the browser) being pushed by AOL, and with Chimera popular on the Mac, IE may have more users, but aside from being more stable and configurable, Moz is now steadily heading for a 1-1 user:browser ratio. Hopefully, this will result in an extremely customized browsing experience.

Re:Moz versus IE

[azzy]

Moz has gone beyond 1-1 user ratio. I have several versions of Mozilla installed.. it's about 1-3 user:browser ratio :)

Xt

[dollargonzo]

XT??? HELLO???

everyone forgot about Xt which works beautifully, and decided to make their own widget sets. this is really annoying when trying to embed Xt stuff into applications that use gtk or qt.

Re:Xt

[dollargonzo]

that is the point: it IS the mother of all toolkits. no one said we didnt need a toolkit, just a standard/default one. Xt fits the bill. it was there and worked, so no need to change

Re:Xt

[roukounas]

XT??? HELLO???

everyone forgot about Xt which works beautifully, and decided to make their own widget sets. this is really annoying when trying to embed Xt stuff into applications that use gtk or qt.

Xt was (is) just a toolkit framework on top of X, it does not change or modify the X protocol. Not only that, but Xt is a mediocre attempt at a toolkit, compared to modern standards: programming with Xt is not easy or intuitive and the on-screen widgets are not up to it.

Xt is not the answer, but a unified toolkit would be nice. I don't think it will happen though, not in this lifetime.

Re:Xt

[dollargonzo]

no one is asking you to use the Xt widgets. you can make them as pretty or as ugly as you want. but embedding other Xt stuff is much easier. there is no reason why gtk widgets can't be built on top of Xt, besides the fact that they use gdk.

Re:Xt

[Redline]

programming with Xt is not easy or intuitive and the on-screen widgets are not up to it.

No joke. To program directly with Xt is to hate your life. But I think you miss the point. Toolkits written *on top of* Xt, like Athena, OLIT, and Motif, are able to interoperate much better than say Qt and Gtk+. You can embed Athena widgets in a Motif app, or vice versa. It is not so easy with non-Xt toolkits. It helps if you think of Xt more like GDK than GTK+, like a sub-toolkit. Nobody writes apps completely with GDK, but *lots* of apps use it indirectly.

Question about typeaheadfind

[Mr_Silver]

Type Ahead Find is currently part of the default install. To turn it off, use:

user_pref ("accessibility.typeaheadfind", false);

Or, to remove it completely, find all files in your installation subdirectories that match *typeaheadfind*, and delete those files.

Whilst it's great that stuff like this is being implemented, is anyone actually working on making a point and click interface to active/deactivate functionality rather than having to get users to resort to deleting or editing files?

If it's already there, for gods sake, why on earth do they insist on giving you these contrived instructions on how to deactivate it?

If the aim of Mozilla is to get a sizeable userbase and encourage developers to avoid writing for IE only then the first thing they should do is make it easy for the common computer user to do this sort of stuff without having to resort to editing text files.

Once they have to do that, then you lose and IE will continue to reign.

Re:Question about typeaheadfind

[aengblom]

is anyone actually working on making a point and click interface to active/deactivate functionality

Of course. Step one is to build the feature. Step two is to make interface. It's called "BETA" be patient.

Anyway, with IE, the feature would just exist. It probably wouldn't even HAVE a way to deactivate it.

[sidenote to editors] It's time to create a Mozilla section. Every release doesn't need front page play.

Re:Question about typeaheadfind

[Kidbro]

If the aim of Mozilla is to get a sizeable userbase and encourage developers to avoid writing for IE only then the first thing they should do is make it easy for the common computer user to do this sort of stuff without having to resort to editing text files.

Good point, but remember that this is the first time we see this feature. I wouldn't expect it to be finished yet (and if you can't live with non finished stuff - don't run betas). I can't speak for the Mozilla team of course, but being a GUI Application developer, I can tell that sometimes you choose between implementing a feature and providing a rough interface to it, or not implementing it at all - as providing a nice "user friendly" (whatever that means) interface would take twice, three or a hundred times longer.
I would expect there to be a nice point and click interface by the time this leaves beta...

Moral of the story: Patience :)

Re:Question about typeaheadfind

[PhxBlue]

If Microsoft has taught us anything, it taught us that added features should be turned on by default and made difficult to turn off through obscurity. Lesson learned, I'd say.

Re:Question about typeaheadfind

[pmz]

...is anyone actually working on making a point and click interface to active/deactivate functionality... ...why on earth do they insist on giving you these contrived instructions on how to deactivate it?

Well, why don't you type "about:config" in your Mozilla location bar. By your argument, there should be pointy-clicky stuff for all 1100+ configurable parameters in Mozilla. Implement that, and Mozilla turns into something like Microsoft Word or the Windows Control Panel (shit everywhere piled under menu upon menu).

Trust me, it is a good thing that Mozilla doesn't put everything in the GUI. Be thankful that the configuration is in a plain text file and not some binary GUI database or, worse, the Registry.

Re:Question about typeaheadfind

[an_mo]

Whilst it's great that stuff like this is being implemented, is anyone actually working on making a point and click interface to active/deactivate functionality rather than having to get users to resort to deleting or editing files?

Get multizilla. Besides a great tabbing interface, it gives you a quick option button that opens a menu containing "open preference editor", where you can edit all user preferences with a mouse click.

Re:Question about typeaheadfind

[an_mo]

It's aready been done in multizilla. Just click on the preference button and on the menu choose "open preferences editor".

Re:Question about typeaheadfind

[Spy Hunter]

Here's an idea. Instead of making a UI for each pref in some monster uber-control panel, give each pref its own url (like about:config/prefgroup/prefname). At this url would be a page with a form to modify the pref and save it in your user.js file. Then next time you read about some new whiz-bang mozilla feature, the article will include the URL, and you can turn it on immediately in the GUI. When someone posts a message to a board somewhere asking how to turn on popup blocking, you can reply with and they would simply click on the link and get a page to turn it on with. Wouldn't that be great?

Re:Question about typeaheadfind

[epsalon]

Don't write it here! Submit it to bugzilla! I suggest using this bug: http://bugzilla.mozilla.org/show_bug.cgi?id=17199

Re:Question about typeaheadfind

[Spy Hunter]

Done! Thanks for the pointer into the maze that is Bugzilla, I never would have found that by myself.

Re:Hooray!

[Lussarn]

I saw the link prefetching feature and thought oh no, there goes our server bandwith. But after reading the FAQ it seems that it's the author of the page that selects what's prefetched and whats not.

Nice feature.

GTK on Phoenix

[distributed.karma]

Since Phoenix (my default browser, as Mozilla is a hog) is built from the Mozilla tree, its latest nightly also has the GTK look. Time to rpm -e galeon.

Link prefetching abuse?

[billybob]

So what happens when the greedy web master decides to add "rel=prefetch" to his tags for banners?

Re:Link prefetching abuse?

[sporty]

It's something mozilla is missing, blocking based off of filters. It'd be nice if I can say, fine, take everything on this server except .swf files.

Re:Link prefetching abuse?

[roca]

The greedy web master doesn't get a cent, because Mozilla doesn't send a referrer for prefetches.

BTW your greedy web master can already just include a hidden IFRAME with SRC pointing to the click-through, which WILL send a referrer, so Mozilla's prefetching adds no new danger here.

Re:Link prefetching abuse?

[Phroggy]

It's something mozilla is missing, blocking based off of filters. It'd be nice if I can say, fine, take everything on this server except .swf files.

Bug 135511.

And Emacs had it forever

[Per Abrahamsen]

In fact, the bugzilla item which typeahead find sprang from was named "implement typeahead find (like Emacs isearch)".

It only prefetches _one_ item...

[Per Abrahamsen]

and only if explicitly specified, and if nothing else is going on (i.e. if you have an active download, prefetch is disbabled).

What about windows?

[jmertic]

When will native widgets find thier way to the Windows versions?. Currently, I moved to Phoenix 0.3 because the Mozilla interface seems to lag on my hardware ( P2-366, 160MB ).

Re:What about windows?

[DrXym]

Answer - NEVER.

If you want native widgets go and run Gecko embedded in something such as K-Meleon.

Of course the concept of 'native widgets' on Windows is little more than a joke since Microsoft routinely inflict their own brand new non-standard monstrosities onto their users with each release of IE or Office. Still, Mozilla uses the Win32 theme engine if its there, so it looks pretty cut on XP.

Re:What about windows?

[roca]

Someone already mentioned this but they don't have enough karma:

We *already* support native look on Windows, but it only works on Windows XP, because only Windows XP gives us the theme engine hooks we need.

Link Prefetching = Waste of Bandwidth

[SuperMario666]

Great feature, but if its use becomes widespread, look for more of your favorite sites to buckle and fold under financial pressure from the increased consumption of bandwidth.

Actually.. No.

[Penguinoflight]

They're releasing slower now, a lot slower. It's just more people care about it. Back in the milestone days you'd get a new release every month or so, now it's about every 2-3 months.

Re:Actually.. No.

[Penguinoflight]

Well, in my personal experience, 1.2a was much more unstable than m18, but then its a lot faster and more feature rich too. I'm not really upset with Mozilla's performance, just pointing out something.

Typeahed adds more - like VI in my browser!

[SuperKendall]

I really like the typeahead find. The great thing is that not only can it apply to links (just by typing) but it can also apply to text across the page just by typing "/" before you type what you are looking for, with Accell-G going to the next match.

UI Not needed

[tweakt]

Whilst it's great that stuff like this is being implemented, is anyone actually working on making a point and click interface to active/deactivate functionality rather than having to get users to resort to deleting or editing files?

1. This is a BETA release. (remember Mozilla is not intended for end users)
2. It's nothing you'd ever need to turn off unless it was causing major problems (ie: crashes).

Re:UI Not needed

[sglane81]

it was causing major problems (ie: crashes)

You should use "eg" instead of "ie" next time when discussing this topic. Everone knows "ie crashes."

Re:UI Not needed

[Phroggy]

2. It's nothing you'd ever need to turn off unless it was causing major problems (ie: crashes).

Or unless you never find it useful and every time you accidentally hit the wrong key, it jumps you to a random place on the web page and you lose your place?

I'll try it for awhile, but I suspect I'll be disabling it.

You could do this before and without too much work

[GoofyBoy]

Here are the instructions

I have it working with Mandrake 9 and Mozilla 1.0.

NO NEED TO RECOMPILE

[Lendrick]

The binary of Mozilla that you have supports antialiasing right now.

Go here and follow the instructions near the top of the page. Provided you have a recent version of FreeType2 on your system and some TrueType fonts for it to find (you have to uncomment a line or two in your unix.js file and tell it where to look), you'll be using antialiased fonts in no time. It looks great, and I wish they'd do it by default. One other thing--you may want to set unhinted to "false", as fonts appear to render better that way. Experiment with your system, though.

I've gotten this to work with the latest Mozilla and an otherwise fresh install of Redhat 8, plus a few .ttf's in the directory "~/.fonts".

Re:NO NEED TO RECOMPILE

[unixmaster]

Normal anti-aliasing & xft2 anti-aliasing are not the same. Xft2 can anti-alias without using XRender extension of X windowing system and Xft2 is far more easy to customize compared to Xft1. :)

Re:NO NEED TO RECOMPILE

[Lendrick]

Ahh, I was wondering about that. All that aside, though, most people are just looking for antialiasing and really aren't concerned with what library is doing it. I wonder why they haven't turned it on by default... it's a frequently requested feature, and not many people are even aware that it's been implemented.

Re:NO NEED TO RECOMPILE

[Dirtside]

I followed those instructions and got it to work -- looks great.

However, my question is, why does this have to be in the unix.js file? Every time I install a new version of Mozilla, it's going to get overwritten. Isn't there some way to set these prefs in my ~/.mozilla dir so that they don't get overwritten when I install new versions? I tried putting a unix.js file in there, but it didn't help.

Re:NO NEED TO RECOMPILE

[Lendrick]

Yeah, I tried the same thing... no luck for me either. You'd think that since it's under "default" there'd be some way to override it, but I have no idea where to start.

At any rate, I used to have a little shell script that would wget the latest mozilla nightly and then copy flash and the java VM into their appropriate places. I'd do something like that, and have it copy unix.js as well. It's kludgy, but it works.

Re:NO NEED TO RECOMPILE

[libre lover]

Isn't there some way to set these prefs in my ~/.mozilla dir so that they don't get overwritten when I install new versions?

Copy your unix.js file to your ~/.mozilla/foo/bar/ directory (the same directory that has your prefs.js file). Then edit your unix.js file. Finally, rename the unix.js file to user.js

Re:NO NEED TO RECOMPILE

[rweir]

Put it in your prefs.js file instead. Moz will leave it alone.

Good defaults are more important

[Per Abrahamsen]

In fact, the Mozilla UI redesign project, Phoenix, is planning to have fewer options (in the UI) than Mozilla, rather than more. Instead they will focus on making Phoenix "do the right thing" out of the box.

Too many options just confuses ordinary users. Of course, that is only true for options in the UI. You can have as many options in obscure text files as you want, only nerds are going to see those.

In my opinion they should have a "Super advanced options" button hidden somewhere, which gave an UI for all the options, for nerds only. Like the customize fascility in Emacs.

Put down the club, step away from the Baby Seal...

[toupsie]

The next time I read someone in all seriousness using the term "blazingly fast" the baby seal gets it.

You toucha da Baby Seal and I'll senda Gianni Jacklone over to breaka your knees. Capiche? Fagetaboutit!!!

Chimera's great, I use it, but it has zero speed advantage over OmniWeb. Nil. Goosegg. And in my opinion, OmniWeb looks better.

OmniWeb wouldn't know a cascading style sheet if it jumped up and bit it on the ass. I constantly find errors in its interpretation of stylesheets. I find OmniWeb's rendering to be substandard compared Chimera. Save the 25 bucks and use Chimera.

Wow - what a bummer

[baptiste]

I've used Mozilla as my primary browser/email for a LONG time. Been happy with it. But I made a clean install of 1.2b (after uninstallin 1.2a) on my Win2K/SP3 laptop, and it won't even go past the splash screen. I guess something in my prefs file is hosing it - sure would like to know what.

Still digging, but it won't even start? Sheesh.

Re:Wow - what a bummer

[baptiste]

I always do this (wipe the directory out completely) Most likely its a prefs.js issue - but I'd rather not blow the profile away if I can help it - but the above skin suggestion may be the ticket. Still digging

Re:Wow - what a bummer

[baptiste]

It may also be a skin problem, the skin format changed (again) recently.

That was it - was using an old SkyPilot. Removed the skin from my profile directory and the lizard lives! Thanks for the tip.

Re:anyone else experiencing this problem?

[sporty]

bugzilla.mozilla.org - you go now!

Prefetch doesn't specify the referer header

[Per Abrahamsen]

I don't know if the banner companies actually uses referer, but if not checking for it might be an easy way for them to combat such abuse.

Re:1.2 beta still has bugs that was meant for 1.0.

[sporty]

I don't think that's very fair to say. Note, that this bug has at least 8 bugs that are dependencies for this to be fixed. It's called a missed goal. It happens from time to time.

-s

Disabling it

[RPoet]

Alright alright, if you really want to disable it, the way to do it is described here. Requires some prefs.js entry though.

Re:Disabling it

[Corvaith]

Thanks! (And to the other person who replied similarly.) Still, the point stands--editing the preferences file shouldn't be required to turn off stuff like this.

Mozilla's feature flood

[Jugalator]

I really have to say that I find the recent development of Mozilla very inspiring as it brings completely new, unique features to the users. First came integrated popup and advertisement blocking. A simple but effective feature. Then came Type Ahead. Then came link prefetching. Now, in what time span?

I don't know about you, but at least my opinion is that the browser software has suffered from some serious stagnation during the past years. Since Internet Explorer 4.0 and its CSS and "DHTML" (mostly Javascript+CSS) support, I haven't seen much development in the browsers at all. Opera was innovative with mouse gestures, but I think the browser that truly turns this stagnation of browser features that's often limited to things like "slightly better CSS support", etc is Mozilla. I'm not even sure how it's possible for the team to bring so many new features in such a short time. Is it a side effect from being open source with browser enthusiasts working on it day and night? Is it "just" because a very flexible and well written code base? An efficient organization of the mozilla developers? A combination?

IMHO, the changes in Mozilla from a late version such as 1.0 are surely larger (at least more useful) than the changes since Internet Explorer 4.0. Each new version is right now bringing lots of new features. Perhaps that will change in the future, but I'll enjoy it while it lasts for sure. :-)

Re:Mozilla's feature flood

[IamTheRealMike]

I'm not even sure how it's possible for the team to bring so many new features in such a short time. Is it a side effect from being open source with browser enthusiasts working on it day and night? Is it "just" because a very flexible and well written code base? An efficient organization of the mozilla developers? A combination?

All of those things. I think type ahead find was written by an open source contributor (of course the module owners helped out, as with most Moz features). Mozilla is very very easy to hack on, as it's very componentized and large parts of it are just text files (xml/js/css). And finally they've been doing stuff like code review, super review, commit for a while so they are pretty slick about it.

Re:Mozilla's feature flood

[goon america]

FYI Type Ahead is not a new, unique feature. IE 5.* for the Mac has had it for the past 3 years. I should know, I've been using it since that time.

Security danger

[Henry V .009]

I noticed pre-caching when I read the release notes last night. In my opinion it is a major security danger.

A lot of police investigations go by the browser cache to see where you have browsed. Now you are giving control over to the cache to someone else.

It would be simple to put a link in the page source to some kiddie porn or other illegal information. You would never see the link on the page and would have no way of knowing what had been inserted in your browser cache until the police inform you of how long you are going to be in jail. Sure, it is possible that the police won't use the browser cache as proof of guilt (don't bet on it), but that requires a lot of trust. And if they want to be technical about it, it is technically illegal to possess that information, no matter how it was acquired.

And the gain isn't at all proportional to the risk. No pre-caching is done except on sites specifically engineered for it. That means next to none.

Re:Security danger

[sisukapalli1]

As someone pointed out earlier...There is nothing preventing a malicious web page designer to load an image into the user's cache (e.g., via javascript preloading, or via img src="xyz.jpg" width="1" height="1" code...

I wonder how the argument would stand in court (Gee, I have 10 Gigs of illegal pr0n on my disk, and I think it was because of mozilla precaching)...

S

Re:Security danger

[AvitarX]

A lot of innocent until proven guilty hinges on the fact that if there is a reasonable doubt that you did not do something then you are innocent, because proof cannot exist with reasonable doubt. So the fact that you are using Mozilla, and Mozilla has the option for totally innocent looking pages to sneak kiddy porn into you cache would be a valid defense. also there would probably be html files cached that employed that tactic in you cache as a very strong defense.

Re:Security danger

[Mnemia]

That's a problem with the police and the law, not with Mozilla. Technology should not be held back because of some irrational and uninformed fear of "bad information" and a mentality that they should try to nail everyone they can. It really seems to me that any decent lawyer would be able to get you off on a charge like that, especially given that they likely could prove that the page was doing it.

Some problems

[macdaddy]

I've got to admit, I am have some problems with Mozilla 1.1 (final release). Recently the problems have led to instability. I was trying to buy a generator on eBay the other day. I had a search going and I opened windows to tabs as I found auctions that interested me. Mozilla kept dumping me. My box is a G4/500MP with 512 RAM running OS X 10.2.1. I've had window focus problems where I click in a Mozilla window text box but my focus is still on another window. I have to switch apps and back to regain control. The javascript driven menus on my PacketShaper 4545 web GUI still don't work. This one is really annoying. And the web GUI to the PacketShaper renders very slowly. Overall I'm pleased with Mozilla. It just needs some honing to fix some of the current problems.

Maybe delaying a release and all new features for a short time to fix existing bugs would be worth it. My $.02.

Re:Some problems

[Anonymous Coward]

eBay is known to cause problems with mozilla.

http://bugzilla.mozilla.org/buglist.cgi?bug_stat us =UNCONFIRMED&bug_status=NEW&bug_status=ASSIGNED&bu g_status=REOPENED&field0-0-0=product&type0-0-0=sub string&value0-0-0=ebay&field0-0-1=component&type0- 0-1=substring&value0-0-1=ebay&field0-0-2=short_des c&type0-0-2=substring&value0-0-2=ebay&field0-0-3=s tatus_whiteboard&type0-0-3=substring&value0-0-3=eb ay

Re:Some problems

[LetterJ]

There's a general Mozilla issue with Ebay as far as I'm concerned. I don't know what it is, but Ebay will bring down my installations of Mozilla 1.2 EVERY time.

Re:1.2 beta still has bugs that was meant for 1.0.

[Jugalator]

> > Sorry, links to Bugzilla from Slashdot are disabled.
> WTF?

The war Bugzilla vs Slashdot sadly had this unfortunate outcome. We will have to live with it. But I'm sure you'll find a way to circumvent the problem. But then again, you're circumventing Bugzilla's access protection and you'll surely be a DMCA case.

Some Tricks To Make Upgrading Easier

[Milican]

You can share bookmarks amoung all your installs of Mozilla, Phoenix, and probably other Gecko browsers (untested). All you do is add the following command to your prefs.js file:

user_pref("browser.bookmarks.file", "C:\\Documents and Settings\\userdude\\Application Data\\Mozilla\\Profiles\\default\\wx4vqyna.slt\\bo okmarks.html");

In addition, you can share plugins by adding the following line to your environment. Her is an example of what I did on my Windows box:

MOZ_PLUGIN_PATH = "C:\Program Files\mozilla.org\Share\Plugin" (in Environment Variables on Win2k)

Really helps so you don't have to redo plugins all the time and you can share one bookmark file for all!

JOhn

Re:Some Tricks To Make Upgrading Easier

[veddermatic]

Or if you are using OS X, just copy the Mozilla.app from the disk image over the one on your HD... boom, you are migrated w/o losing anything.

Re:Some Tricks To Make Upgrading Easier

[an_mo]

I couldn't find the way to add the environment variable on win2k

Re:Some Tricks To Make Upgrading Easier

[aok]

Right-click "My Computer", go to the "Advanced" tab and there should be a button for Environment variables.

Doesn't work with Windows Proxy servers

[oliverthered]

Well, this is almost true,
please vote for this bug (99 votes so-far, lets make it 100)

so that me and anyone else who uses microsoft proxy server 2 or any NTLM authenticating proxy can use mozilla. (this is probably a few million people, and a lot of corporations)

This bug has been there since 2000-01-11, and won't make 1.2, hopefully it'll make 1.3 alpha 1!!!

Re:Doesn't work with Windows Proxy servers

[Monkey]

"Sorry, links to Bugzilla from Slashdot are disabled."

Now why would they bother doing this? I can't see what they have to gain by making it harder to reach bugzilla. Besides, all you have to do is paste the link in another browser window and it works fine.

Re:Doesn't work with Windows Proxy servers

[cant_get_a_good_nick]

Now why would they bother doing this?
Because every time someone posts any Mozilla related story to Slashdot they get, well, slashdotted. Considering it's a DB driven site, it doens't take all that much to drive the server to it's knees and make it unusable. So for a day or so, they can't use a basic productivity tool. Yeah, you can just copy n paste, but this minor increase in effort probably eliminates all the casual clickthroughs.

mail and address book export?

[small_dick]

i tried to find a way to move my mozilla mail and adress book from mozilla to evolution (now the default mail app in redhat 8.0) and could not find a way to do it.

most of the websites about migrating to evolution discuss windows/outlook, or windows/mozilla, not linux/mozilla.

I guess since it's open source I should get off my ass and write a conversion script?

Re:mail and address book export?

[Misch]

For the address book, you can export an LDIF file from inside Mozillas' Address Book. Then, the LDIF file can be imported straight into Evolution.

As for the mail, they're in standard mbox files. You can move/copy these to the default mail directory, and they should appear as local folders within Evolution.

Hope this helps.

Not a good idea

[spitzak]

This is a very very bad idea. You seriously underestimate the enormous complexity of the communication to a "toolkit". Also there is a little thing called "innovation" where people invent new methods of GUI interaction. This would be stopped by such a design, or would force people to write a parallel toolkit anyway (as many (most?) Windows applications are forced to).

If you don't believe GUI innovation happens, imagine if X had an enforced toolkit. It would be Athena, in black and white, with this 1-bit color so written into it that it would be impossible to remove, and everybody would marvel at the fact that you could set it to inverse video and all applications would agree. And defenders would claim that the fact that only the middle mouse button makes the scrollbars move was a *feature*. And any intelligent people would be laughing X off the planet!

Meanwhile, despite it's problems and pretty stupid design even for when it was invented, X is able to replicate interfaces designed 15 or more years after it was invented. This is because of the one intelligent decision they made, which was to keep the GUI widgets out of it!

Now X has problems. There really should be high-level graphics, at least similar to PostScript. Though also complex, it is far less complex than toolkit interfaces, and perhaps more importantly the set of graphics calls needed has been pretty stable for about 20 years. It may even make sense to add calls to "draw a nice raised box" or "clear this to the flat background color" which would do about 99% of what people want "themes" to do.

Also there is a bit of "toolkit" inside X: the "window manager" (even though a seperate process, but the communication protocols are there, and I know for a fact that it takes more code to communicate with the window manager than it would take to draw the window borders and handle moving and raising the windows myself). This also needs to be removed.

But I am serious that putting any kind of "toolkit" interface into the system in a very very bad idea.

Re:Not a good idea

[evbergen]

It's nice that you mention window managers. Although generally not running on the X server side, they are an integral part of the X window system. Even though window managers are specified as an entity separate from the client application, we've seen lots of innovation in window managers.

If the X server would not mandate certain widgets, just as it does not mandate any specific window manager, but would simply allow a 'widget server', then client applications could still be relieved of the burden of managing their own widgets, just as clients are already relieved from the burden of drawing their own window decorations and moving windows around.

It would be great if applications wouldn't have to worry about drawing a check mark in check boxes when clicked, not even at the protocol level, and that a widget server would handle that.

We'd change themes (widget servers) as easy as we change window managers now, and have a well-defined protocol between client application, display server and widget server, just as we have for window managers.

I'm serious. We need a better X protocol. And it's not HTML/XML over HTTP, sorry. The ultimate test there is to get self hosting I guess, i.e. can you implement a web browser as a web application.

But until someone implements gecko in javascript, I'm sceptical ;-).

Re:How much memory will it use with my MAC OSX

[veddermatic]

I'm running most recent OS X (10.2.1) and this version of Mozilla is taking up 8% RAM (of 512 MB)

Another major unfixed bug

[Animats]

The bug that causes crashes and profile corruption if you have both Netscape and Mozilla installed still hasnt't been fixed.

That's been outstanding for most of a year now, which is inexcusable for a major bug that causes data loss and crashes. The Mozilla team still has way too many "don't do that" items in the release notes.

Unless this thing gets cleaned up, it's never going to get market share. Adding additional features of very marginal utility won't help. Could AOL use Mozilla as their standard browser? No way. It's got to just work.

New standard?

[donutello]

Is this a new W3C standard? I wasn't able to find any references from the page you pointed to or from a quick search on google.

So is this an implementation of an existing W3C standard or is it an example of an evil browser deliberately breaking standards compliance?

Re:New standard?

[asa]

http://www.w3.org/TR/REC-html40/struct/links.html# edef-LINK

The link tag has been around for some time. It is used to describe releationships between documents. It was desinged by the w3c with extensibility in mind. The w3c leaves it up to the user agent to determine how to handle link data.

--Asa

Re:New standard?

[donutello]

How is this breaking standards compliance if there's no standard in place? It's not "breaking" the way anything else is interpreted, is it?

It isn't. I was being sarcastic. IE gets accused of breaking standards compliance everytime they introduce a new feature that the w3c hasn't gotten around to standardizing yet.

Yeah, I stopped using that

[Gorimek]

It sounds good in theory, but once you use you realize that pages have far more links than you thought. A typical page can have 20 or 50 links, only 2-4 of which you would be interested in prefetching. Just look around on this page for a good example. It ends up furiously downloading pages, movies etc for as much as your connection can bear, and it's not good for anyone.

The Mozilla approach could actually work. If any designers ever decide to use it.

Prefetching & Standards Complience

[Thenomain]

Maybe I'm missing the standard for it (I'm not on the bleeding edge of things), but I was looking at the HTML 4.01 link rel types and can't find "preload". Fortunately, according to the FAQ, "next" will do just fine.

This is a not nit-pick, but with all the touting of how 100% standards compliant Mozilla is, I'm wondering what the philosophy is on extending the standard, if "preload" isn't in some later HTML standard that I don't yet know about us.

Re:Prefetching & Standards Complience

[FooBarWidget]

Complying to W3C standards doesn't mean not allowed to invent your own standards.
The effect of preload-"tags" is mostly transparent; users of alternate browsers won't be left in the dark just because those browsers don't support that feature.

Re:Prefetching & Standards Complience

[asa]

100% standards compliance means not doing anything forbidden by the standards body specifications and supporting the recommendations of the standards body. If an extension doesn't violate the specification then it is in compliance. The LINK element was specifically designed with room for clients to extend and interpret.

--Asa

Re:Prefetching & Standards Complience

[darinf]

rel=prefetch is something we hope to turn into a standard eventually. it is better than overloading the meaning of rel=next.

Type-ahead find weirdness....

[Fnkmaster]

Is it just me, or does typeahead find have some really strange behaviors? I just d/led 1.2beta and I'm trying it out, and it seems like typeahead find successfully finds about half the occurrences of word patterns, and I can't even figure out if this is detereministic or not. For example, in the article here, I tried to find "moz" and kept pressing F3 to cycle and it was missing quite a few occurrences of the word "Mozilla" which is all over that page. Is there something I'm missing about how this feature is supposed to work?

It'd be nice if they DOCUMNENTED it...

[Lendrick]

Last I heard, the general Mozilla project attitutde about documenting the preferences was that if you don't know what they are, you shouldn't mess with them. As a highly techincal user, I myself would beg to differ. Failing to document all of these options in one place is a cop-out, and their excuse is pure arrogance.

If I'm wrong about this, and there is complete documentation on all the prefs files, I'd love to know about it.

Silly moderators

[0x0d0a]

Score:4 for *that*? I mean, I thought it was kind of cute or I wouldn't have written it, but...

Moderation on Slashdot is a little whacky.

Oh, what really gets my goat is when I post to some post saying "this post deserves to be modded up", and some whimsical moderator mods the "mod up" post up *instead of the parent*. I remember getting a Score:4 saying that the parent post (Score:1) should be modded up.

There's no "right" app for OS X users.

[kitzilla]

I agree that Chimera ("Navigator," officially) is a terrific Mozilla browser for OS X, but we have a lot of choices these days.

Chimera is still pretty sparse on features. I use the nightlies, and run into a fair number of buggy builds. But it's quick, and sure looks like an OS X app. I use it far more than anything else.

KevinG, the guy who did the Pinstripe skin for Mozilla, was nice enough to compile Phoenix 0.3 for OS X. It's just an experiment, not part of the regular project. But damn if it doesn't work, and it has some very cool features. Even *more* OS X choices:

http://www.kmgerich.com/misc.html

This OS X build introduced me to Phoenix, which is now running on my Linux box. Kevin's page says his OS X build requires Jaguar, but I'm using it with 10.1.5 just fine.

Mozilla 1.2b feels very stable on OS X. It's not as fast as Chimera, nor is it as consistent with the Mac human interface standards. But it doesn't suck, and some users like working from within a suite. I know plenty of OS X guys who are more comfortable with Mozilla's mail than Mail.app. It's a matter of preference.

To me, Netscape 7.0 is heavy and gaudy. It has a spellcheck app, however, and isn't a bad choice for those who rely on the Netscape/Mozilla suite for email.

As for Omniweb, it's a great browser. A few more features than Chimera in its current state of development, though don't think it renders as well. Speed is a toss-up.

Every OS X user's needs are different. It's a great time to explore the platform, however. There's a browser for everyone. Run whatever you prefer, and support the community which surrounds it.

Thanks to all the developers who make my online experience more enjoyable. Your work isn't taken for granted.

Re:There's no "right" app for OS X users.

[EricWright]

My biggest complaint with Chimera is the distinct lack of preferences that can be diddled within the app... I'm sure I could find the right prefs file, figure out what keys need to be set with what values, etc., but Chimera REALLY needs to add in a more robust Preferences panel than it currently has.

That and fix the bugs with not saving changes in the fonts panel.

However, I rarely have crashes with Chimera (no more often than mozilla 1.2a/b has)... except for some plugin issues that have been widely reported on bugzilla. Those got squashed sometime last week, and I've only had 1 crash since then, and that was more of a "my keystrokes aren't registering, but the app is otherwise responding" glitch rather than a true error log-producing crash. I typically get one or two of the nightly builds every week just to see if things are better. Speed has definitely improved in the past couple weeks.

Leech

[fialar]

Leech seems to install properly only if you run it as root. (It wanted my to have write access to /usr/local/mozilla/chrome dir.)

It doesn't work as a user. Weird.

Re:Cautionary e-mail tale

[Lendrick]

Sounds like a really old bug. When Moz-Mail crashed, it used to corrupt its mail index files. The trick to getting at your mail again was just deleting the corrupt index. It would reindex them the next time it started. Nowadays, when it sees a corrupt index file, it rebuilds the index automatically.

How long ago did you have this problem? To my knowledge, it's been fine for over a year.

So, does email do more than beep for email yet?

[zrk]

I'd love to know how to get it working for Linux. I tried suggestions found in bugzilla, but no dice in 1.1.

In my office, I can't hear the PC beep. I'd like it to work through the speakers, if possible.

Otherwise, I've not had complaints, and I've been using Moz since .8 or so...

viewing selection source...

[esarjeant]

There also appears to be a View Selection Source option now. So I can highlight a section of a document and view just that HTML source -- very handy for development.

Chimera vs. Mozilla

[EricWright]

Is there any way to import Chimera bookmarks (XML) into Mozilla (HTML)? I did the obvious (Import bookmarks from mozilla and selected my bookmarks.xml file from the Chimera path) but that didn't work...

TIA
Eric

Splash screen

[WowTIP]

You'd think they should have found time to replace that awful (windows) splashscreen by now? :)

CSS is still messed up :-(

[Anonymous Brave Guy]

Unfortunately, according to the release notes, they still stubbornly refuse to use CSS stylesheets that aren't served with a mime type of text/css if you're rendering a "strict" main document. That's lovely for supporting standards compliance, I'm sure, but alas I still can't see the information on a vast number of web sites out there, from major companies to user pages on some popular ISPs who should know better. This is a Very Bad Thing, and whichever overly petty member of the development team chose to do it (and it must have been an active decision) should perhaps consider that this is not the way to gain a user base.

Re:CSS is still messed up :-(

[asa]

they still stubbornly refuse to use CSS stylesheets that aren't served with a mime type of text/css ... I still can't see the information on a vast number of web sites out there

You visit sites that include the "information" in stylesheets? That's completely lame. The whole purpose of CSS is to separate the information from the style. If they're including the content in their style sheets then they're doing a lot more wrong than just serving the incorrect mime type.

--Asa

Re:CSS is still messed up :-(

[Anonymous Brave Guy]

The whole purpose of CSS is to separate the information from the style. If they're including the content in their style sheets then they're doing a lot more wrong than just serving the incorrect mime type.

I was referring to the styling information, not the words on the page. Unfortunately, if a page makes heavy use of CSS for formatting and layout, it can still be very hard to read without it. You could just read the plain text by scrolling around the window lots, but then you could just read the HTML source if you wanted. That's not really the point. IE gets it right: it shows me what I want to see. Moz doesn't. That's an indisputable point to IE, I'm afraid. There is simply no good reason for Moz to be anal about whether it renders using a stylesheet or not, at least not without giving the user an opportunity to override it.

Re:CSS is still messed up :-(

[Anonymous Brave Guy]

so, you think that mozilla should ignore the fact that a web server is saying "this file is text/plain" and rather guess that the file is "text/css" because that's probably what the site author meant should happen.

sounds an awful lot like asking your browser to guess about the author's intent ...

But this is exactly my point: there is no guess involved. CSS files are plain text files. The only thing that says they're not is one string from the web server, which is frequently wrong at their end and of no consequence whatsoever at ours.

If there is a tag at the start of the page that reads

<LINK REL="stylesheet" HREF="styles.css" TYPE="text/css">

and, on requesting a download of the file "styles.css", a properly formatted, CSS-compliant file is received, what other intention did you think the author had?

if the site author screwed up, i don't want my browser to pretend for me that everything is ok.

Please yourself. Me, I'd rather see the page the way it's meant to look. I don't surf to admire how standards-compliant the web pages I visit are, I surf because I'm looking for information and/or entertainment.

Standards are a means to an end, not an end in themselves. I don't expect, and wouldn't want, Moz to imitate IE's mishandling of, say, the standards for margin and padding specification, because that would cause pages written according to the standards to display incorrectly. But there is no sensible alternative here, it's just being anal about standards for no helpful reason. Accepting CSS files sent with an incorrect "text/plain" type wouldn't break any standards-compliant pages, and the only standard it might encourage (or rather, not discourage) breaking serves no useful purpose anyway.

And in case you missed it, the problem here is that it's not usually the site author who screwed up, it's the hosting service. That's typically beyond the site author's control. They frequently get it wrong, but there's nothing the site author can do except ask them to fix their servers. Having been that site author, I know exactly how effective that isn't, and changing hosting service just because of this isn't practical: most of the people visiting that site have IE, and those with Moz will just have to live with it. That's a shame, because I'm pretty sure that both those who visit my site and me would prefer for it to be seen as I designed it.

Re:CSS is still messed up :-(

[Anonymous Brave Guy]

I am of the opinion that web page authors should check their sites with CSS off to make sure they will still be readable.

I certainly agree with that, and with using design techniques that cause the main content to render before navigation bars and such where applicable.

That's not really the point, though, is it? I thought we were trying to encourage the adoption of new browsers and a move towards standards compliance, not discourage it. Petty decisions like this simply mean that if I want my web page to work properly with Moz, and my service provider happens not to send MIME types properly, I can not use CSS at all. I'm back in the dark ages of table-driven layout and in-line formatting. Whose fault is that? Well, frankly, who cares? The damage is done any way you look at it.

Re:CSS is still messed up :-(

[Anonymous Brave Guy]

No, IE gets it wrong. [...] If you don't like it, then complain to the HTTP WG.

I'm aware of the technical details of the relevant standards, thanks. I think you miss, or choose to ignore, my point, however. The standards are just a means to an end. Neither I, nor those who visit my site, care that on some technical detail, my ISP is screwing up. We both observe that my (correct and standard-conforming) web page data displays as I intend it to on their IE browser, but does not display in a particularly pleasant way on Moz (though it's fine if I view it locally using Moz; this is purely a problem with the ISP MIME type settings).

Whether you like it or not, your quest to evangelise W3C standards is coming in a distant second for just about everyone else here. Clearly the de facto standard is superior to the official W3C one in this respect; following it hurts no-one, while following the W3C one hurts anyone with an overly picky browser. Mozilla simply will not take over as a serious, mainstream, alternative browser if they continue to make mistakes like this, which is a shame, given the obvious advantages it has in most respects.

Use / to find non-linked text

[mbrubeck]

If you just start typing "moz...", typeahead will only find text that's part of a link. If you type "/moz..." instead, it will find any text. (Apologies if you already knew this.)

How do I save installed XUL stuff?

[Dirtside]

I installed XUL Planet's Preferences Toolbar on Mozilla, but the next time I installed a new version, it was gone and I had to reinstall it. I know that you can install plugins into your ~/.mozilla directory so that upgrading the browser doesn't require reinstalling the plugins, but is it possible to do this for chrome-like things (like the aforementioned Preferences Toolbar)? I've highly customized the toolbar, as well, and I don't even know where that configuration gets saved. Thanks.

I give up

[0x0d0a]

It took 15 minutes for *that* post to get modded up. I give up.

Re:I give up

[orkysoft]

This post deserves to be modded up :-P

Blocking plugins

[xiox]

There's a bug for blocking plugins from various sites. See bug number 94035 on bugzilla.mozilla.org (no link as bugzilla does not allow ./ linking)

Not quite

[Flammon]

If the iframe has a CSS property of display:none, or visibility:hidden then Mozilla will not load the content of that iframe until the display property changes to something that is visible.

IE loads the content of the iframe not matter what the CSS properties are.

Re:Not quite

[roca]

Then just hide the IFRAME using

Re:Not quite

[roca]

oops.

You could just use an IFRAME with zero height, specified with style="height:0" or whatever. Mozilla will load that.

Of course it would be better if people didn't do this...

Prefetch works with many existing documents

[mbrubeck]

The Mozilla approach could actually work. If any designers ever decide to use it.

The Mozilla approach already works on many documents on the web that have "rel=next" link tags -- especially documents generated from structured markup (DocBook, other SGML). If you're using Mozilla 1.2 beta, you can try it on various W3C recommendations or the GTK programmer's reference, and thousands of other structured documents on the web.

Site Navigation toolbar

[mbrubeck]

I forgot to mention, the Mozilla "Site Navigation bar" (accessible from the View->Show/Hide menu) also adds functionality for documents with link tags like those mentioned above.

Re:Link Pre-fetching is a baaad idea...(indeed)

[*xpenguin*]

Oct 17, 2002:

First goatse.cx link modded informative.

Re:You could do this before and without too much w

[msaavedra]

XFT2 support was only checked in last week, so there's no way you have it working with Mozilla 1.0.

Christopher Blizzard has been releasing xft2-patched binaries of Mozilla for some time. They are available here. You're absolutely right that the new XFT2 font-architecture renders better. I'm using it in a patched Mozilla 1.0 right now, and the font quality rivals Windows. The quality of the old FreeType code was not too great.

Re:You could do this before and without too much w

[gol64738]

if you have redhat 8.0, and don't want to get your hands dirty,
here are Redhat 8.0 RPMS for binaries with Xft2 already compiled in

Best feature...

[dalutong]

the "group" as a homepage. i open slashdot, linuxtoday, and google news... click "use current group" and now when i open mozilla it pops open three tabs each with one of my favorite sites.

Very cool. Phoenix, when will you have this? (Please, please, please put it in!)

Re:You could do this before and without too much w

[gol64738]

Ok, so i just installed those RPMS and all i can say is..

BEAUTIFUL!!

i highly recommend all RH8.0 users to upgrade your mozilla to the above link. you won't regret it!

Insightful?

[Chuck Chunder]

If you have a GTK desktop then Mozilla 1.2 will just match the current GTK gui (a few bugs notwithstanding) in Classic. You don't need to do anything.

i still miss image management...

[ovrclokd]

am i the only mozilla user left who thought that pre-download image management - "ask me before downloading an image" in preferences/images - was one of mozilla's best features? no web bugs, no third-party banner ads, total control... it got axed as the "fix" to bug 110112, because of crashes on multiple confirmation dialogues, and despite repeated pleas (vote for bug 146513!), it's still on the wontfix list.

i'm still running 1.0.1 build 20080808, since that's the last build that still had image confirmation, and so far none of the nifty new features, appealing as they are, have been worth giving it up. anybody want to start a write-in campaign? :)

(i know, "send code" - but i'm a networking geek, not a Real Programmer. i don't have enough coding clue to even understand the code involved, let alone to try to make changes.)

Re:i still miss image management...

[frankie]

anybody want to start a write-in campaign?

We just started something even better -- a pledge drive. Go visit bug 146513 and commit a few bucks to whoever will write the patch for us.

Re:more sites not Mozilla accessible

[SCHecklerX]

if on linux, make sure your jre is a symlink, not a copy. This used to get me when copying my plugins from old moz version to new.

you must have a 1.2b theme

[Khopesh]

no pre 1.2b-branch themes work.
there are only a small number (growing by the day) of compatable themes.
don't get a theme that doesn't specifically list support for 1.2b