Slashdot Mirror
Mozilla 0.9.6 Released
bluephone writes: "Yessireebob. mozilla.org has released the 0.9.6 milestone. Here are Release Notes and a link of files on the FTP server. For milestones 0.9.7 and 0.9.8, the focus is on performace enhancingment, and stability of the Mail/News end of the suite. And boy, is it getting good..."
If the only part of mozilla you like is Gecko, then use only gecko with a simplified interface.
For linux, try Galeon
For windows, try K-Meleon
Well lets look at the system requirements, which as we all know are very conservative
:)
Windows
* Intel Pentium-class 233 MHz (or faster)
* 64 MB RAM
* 26 MB of free hard disk space
Linux
* Intel Pentium-class 233 MHz (or faster)
* 64 MB of RAM
* 26 MB of free hard disk space
Since you probably can't upgrade your processor on your board maybe you should try and bump your RAM to 128MB or so? That would definitely help out. Otherwise I recommend you give Opera a shot. It's right up your alley and it works on Linux and Windows
JOhn
Campaign for Liberty
Idea wishlist:
I'm a very busy person who does some good for the community already in his free time, so don't ask me to implement these features. I just don't have the time.
Perhaps this would be a good time to ask... does anyone know of a proxy that allows you to rewrite packets on the fly? I think the web's got to the point where I want to start overriding some HTML arbitrarily. I know regular expressions, so some sort of regex interpreter would be quite handy.
Er, nice try.
The main reason is that Mozilla makes pretty heavy use of pthreads, and pthreads don't exactly fly on Linux. Windows threading performance is definitely superior to Linux's pthread performance, at least on single processor systems. Why?
Windows is absolutely useless at process creation. Windows threads are about as heavyweight as Linux processes, and Windows processes are hopelessly heavyweight. So, the traditional Unix model of using many processes to complete complicated tasks completely fails on Win32. Microsoft's answer to this failing was to make threading as fast as possible, and to push multithreaded programming as a hack around a fundemental OS problem.
Back in Linux land, the relatively low demand for mutlithreaded apps (because the Unix model really works quite well if you have fast process build up/tear down) finally pushes Linus and friends to implement clone(). The clone() system call was based on Plan9's thread model, and is actually much faster and more advanced than Win32 threads, totally beating out of almost all standard OS kernels at thread performance.
So, now Linux has both faster processes and threads, but thread performance still sucks. Note that I said pthread performance on Linux isn't very fast. Pthreads are POSIX threads, and have very different semantics from clone(), mostly to support implementing multithreading in userspace (ick). So, the standard is a hack to say the least. Unfortunately, it's still a standard, and Linux must map pthread behaviour to clone() in userspace, which is painfully slow. It requires multiple context switches just to created a pthread on Linux. So, pthread-heavy programs like Mozilla just crawl.
So, the moral of the story is that Linux has a much better core, but seeing that the Linux community actually cares about standards, performance isn't quite up to snuff.
Can't comment yet as I'm just downloading now, but in the meantime: where have the themes gone? Whenever I try to download others, I'm met with a 'page not found' error. Is this because Mozilla is moving faster than the theme developers can handle? Is there actually more choice than 'classic' and 'modern'? Not that these aren't good themes anyway. I'm just wondering...
Keep up the good work! (Oh and fix that bloody 'print selection' bug.)
It's appreciably faster than Mozilla as it uses native win mfc components for its buttons, dialogs etc.
Use ISO 8601 dates [YYYY-MM-DD]
for a quick and reasonable-on-mem-usage browser, please see skipstone. it uses the mozilla-embedded lib for page rendoring, and it doesn't come with all the html editor / mail / news crap.
see ports/www/skipstone for FreeBSD users
From my .mozconfig:
ac_add_options --disable-mailnews
ac_add_options --enable-optimize="-O4 -finline -fno-omit-frame-pointer -march=pentiumpro -mcpu=pentiumpro"
I don't know what build options are used for the milestone builds...
I was expecting the worst, but I was quite impressed. Heck, MS even went through the trouble of providing CDE icons. My only complaint, feature-wise, is that you can't browse local directories with IE, but that's minor.
You must not be using a recent version. Get 5.0 and then tell me it's slow and bloated. Fine, maybe it doesn't work for you, but on all of my modern Solaris boxes -- from a humble Blade 100 to an E6k -- IE stomps NS4.x and blows away Mozilla. I'll admit that on my SPARCstation 5 (the 170MHz Fujitsu TurboSPARC model) Netscape 4 provides better performance, but if you're running kit that old, you have more important things to worry about than browser performance. ;-)
I used to be quite the Netscape weenie myself -- hell, even on Windows I used NS4.x up until IE5.0 was released. But not a single person I've ever met (IRL, as opposed to these online flamewars) who's used IE on Solaris and fairly compared it to Netscape and Mozilla (without the anti-MS issues being involved) can deny that it rocks. Seriously, if you're having issues, give me your hardware specs and OS version and I'll try to help you out.
--
I like to watch.
If Opera can run so snappy and fast, why not Mozilla?
Because Opera can't do what Mozilla can do. Opera handles common HTML and CSS just fine, but if you're pushing your web design further, Opera's rendering engine falls apart.
I use translucent PNG images on my web page, and Mozilla does alpha-blending beautifully. Opera, on the other hand, can't even handle transparency, let alone alpha-blending. Opera also can't switch style sheets through DOM or the UI. It cannot install components on the fly. The list goes on...
"JayPee" has made available Navigator-only optimised builds for Linux that you might find useful.
rmathew.com
I mean, it is extremely slow, even on my PII-400 with 256MB of RAM. IE is ways faster in load time and crashes less (trust me, the first time I loaded the latest Mozilla Mail/News, I got a crash within 3 minutes). You may say that IE is preloaded, but what about Opera? I'm using Opera 6 beta from time to time and its browser loads at about the same speed of IE.
Pretty good crash performance; honestly, for me Mozilla crashes once in a few days. Even a year ago, it crashed just once or twice per day day for me. With the relative heaviness I agree though; It works pretty nicely with P3-550 & 256 MB of RAM or even better with my home machine (Thunderbird, more memory), but it's definitely a bit slow with my secondary machine, K6-2 400.
Everyone who makes generalizations should be shot.
Yes, and it worked fine in a clean profile, and if I view the pages anonymously in the new clean profile, it works fine.
But when I log into slashdot, and view the page, it craps out (this is still on a clean profile).
I'm a moderator, so i have one of those 'score' dropdown boxes by every single post. Hundreds of posts = hundreds of dropdown boxes.
I used up that one last moderator point, the dropdown boxes no longer appear; and guess what, the page no longer crashes my browser. Works just dandy.
Off to file a bug...
"Can of worms? The can is open... the worms are everywhere."
I don't know who fucked up the shortcuts, but I must use alt-d over 100 times a day in IE, the shortcut that brings you to the address bar. I had a (not too investigative) look at the Mozilla help, and couldn't find any info on shortcuts
Ctrl-L (for Location) is the keystroke you're looking for.
Use the turbo feature, Mozilla loads faster than IE.
If you use Linux, please help development of Autopac
This is the Unofficial Mozilla Spell-checker FAQ: http://www.mozilla.org.uk/docs/spell-checker-faq.h tml it should address your concerns.
linux's pthreads use clone. they are kernel threads. read the source.
- You can disable Mozilla's JS window.open()
A nice feature, true, but what happens when you go to click on a "help" icon and it can't open a new window?
The mozilla anti-popup feature disables popups on window open, page load, and window close (and timers). So obnoxious auto-pops don't happen, but e.g. The Onion's horoscopes still work.
Sumner
rage, rage against the dying of the light
Ah, that isn't Mozilla, it's Win98 running out of GDI-mem to display all the widgets. Windows 95/ME has the same problem. It's simply running out of window handles (in Windows, EVERY widget is referenced by a window handle).
Windows NT/2K/XP doesn't have this problem.
- Mozilla is Open Source
Zealots aside, why is this better?
It's there if it's needed or wanted, and can't be taken away.
- Mozilla won't accept activeX or other such nonsense
Which limit's its use on heavily scripted, harmless, usefull sites.
I honestly don't know of any sites that are heavily scripted while remaining both harmless and useful.
- Mozilla has tabbled browsing
Which slows down the quick alt+tab everyone uses to switch between browser windows...
Then use windowed browsing. It didn't go anywhere.
- Mozilla doesn't redirect you to MSN (or AOL for that matter) and spill your privacy for all to see
Nor does IE, if you configure it correctly.
I never found an option to prevent getting passed off to MS on a DNS lookup failure, but then I almost never touch IE outside of HTML testing purposes.
- Mozilla has a development team that cares about the end product
More ramblings from a zealot. I'm sure the IE programmers care about IE. They just dont feel the need to sit around and pat each other on the back in public message boards.
I'm sure the IE developers pat each other on the back all the time, but you won't see it since they don't have public mailing lists.
- Mozilla has site-specific image and cookie management
Internet privacy zones. From your top menu in IE6: tools -> internet options -> privacy -> click the edit button. Yep, it works in IE on a site by site basis.
That it does, though personally I prefer the format of Mozilla's privacy tools. Probably a familiarity thing.
- Mozilla is stable (close to 100%) and won't bring down the OS when it crashes
Just like IE6 (which hasnt ever crashed on me, even though I use it roughly 14 hours a day, 7 days a week, for the past few months)
Last time I used IE (5.5) it crashed several times a day (usually on malformed javascript or activex), taking the whole OS with it half the time. I haven't tried IE6 though.
I've been using the Mozilla nightlies, and I haven't had a crash since before the summer. Some really funky regressions like the expando url bar... but no crashes.
I put my browsers through HTML/Java/Flash/script hell though, so YMMV.
"Where shall the word be found, where will the word resound? Not here, there is not enough silence." -T.S. Eliot
Unfortunately this high optimization (>= -O3) will not work while compiling with GCC 3.0 which currently has a strange bug which will cause mozilla to crash on startup. A patch was checked in early today that fixes this problem on the mozilla trunk.
Here are a couple of build options that I frequently use in my .mozconfig when building mozilla to keep it running extreemely well also cutting alot of the cruft out.
These build options are for all the people that are complaining about shoddy mozilla performance under linux and people that would like to have a look at some really new features.
ac_add_options --with-extensions=all
Enables such things as the Chatzilla IRC client and the dom inspector(which I think is extreemely neat for debugging and viewing dynamicly changing html object model) also contains some very experimental things such as xmlterm.
ac_add_options --enable-mathml
Very neat standard for displaying math of all types and sizes in xml.
ac_add_options --enable-crypto
Great option, about a year ago this option wasnt even possible due to netscape not realeasing it's code due to US laws afaik. Now everyone that want to compile the lizard can get ssl support built right into the browser.
ac_add_options --enable-optimize="-O3 -march=i686 -mcpu=i686"
The main optimization part. This option has the biggest leaverage affect on the actual quickness of the browsre itself.
ac_add_options --disable-tests
Get rid of the unneccesary tests.
ac_add_options --disable-debug
We don't need any debuging symbols in th build if where not a developer do we.
ac_add_options --disable-shared
ac_add_options --enable-static
A nice new enhancement of the moz build system which links all of the modules in statically, im experiencing a big speed increase and a decrease of startup times with this option probably because it doesnt need to read each individual shared object from the hard disk.
You forgot to itemize the good XML-CSS-rendering, Mozilla or Galeon are pretty close to w3-standards. I saw some XML documents which were not correctly rendered in IE or Opera. I nearly want to call Mozilla not only browser, but also HTML/CSS/XML-Validator:-) It helps to find bugs in page code!
Although I'm not 100% sure if Konqueror is included, I have tried the KDE for windows package a while back, and it was erm kind of functional. And on their page it says they've successfully ported KDE 2.2.1 to cygwin too. If memory serves me Konqueror is in the kdebase package, which is available (for windows) at.. sourceforge
I'm a leaf on the wind. Watch how I soar.
Opera and Konq also have this nice feature.
I'm hoping Moz steps up to that plate soon....
X will always be slower than GDI, because of the client->server copy and context switch (whereas with GDI it's just fill command buffer, then execute). Not that this is a bad thing - network transparency is great, and the speed difference can theoretically be made arbitrarily small, given enough optimization of the X code =). At this point though the basic problem is that the X protocol is too stateless, and so you waste time sending massive amounts of state between the client and the server. (believe me, I've used LTT to make kernel traces of Mozilla redrawing - it's nasty. It's not just a simple expose event->redraw round-trip; it's several round trips...).
It's icky, but often irrelevant (since most Linux programs just fork additional processes, and this is pretty clean. Under Linux, process creation is much less expensive than in Windows).
That's completely true, and all of my own software uses multiple processes for concurrency rather than threads =). But, we are talking about Mozilla here. Mozilla makes extensive use of pthreads :[. (ideally a web browser should just be one thread using async socket I/O; although I might go along with having a second thread for async disk I/O, and/or doing Java garbage collection asynchronously)
right-click on the tab, select "close tab". This will close the tab without switching to it, albeit with two clicks, not one.
I use Mozilla with KDE, and haven't used Gnome for nearly a year. At the time Galeon required you to have a full Mozilla install - this was because of the licensing of some of the main Gecko components. Supposedly the licensing was changed, and word was that Galeon would no longer require a full Mozilla install because it could just ship with the few Gecko components it needed. However, I notice on the Galeon site it still requires a full Mozilla install - was the component license issue never resolved? What's up? (not a flame, just curious)
Sorry but after installing mozilla and doing some performance tests with dhtml, mozilla is about 70% slower than IE6 (p4 1g ram winXP) even in simple animations making practical dhtml worthless in it and flash seems more attractive by the day as this isnt dependant on a slow javascript and rendering engine, which would be a shame.
:p
Load time isnt even an issue as its so slow even when its loaded.
Javascript to plugin communication still doesnt work out of the box (contrary to what the moz site says) at least ns4 supported it.
standards support is meaningless as no-one supports them , making it more of an "ideal" than a standard.
quote : "standards are great because there are so many to choose from"
i don't think m$ has anything to fear from mozilla in its current state, at least not in this decade
The current daily builds 0.9.6+ do ask for /favicon.ico by default. The released 0.9.6 build does not.
These two options are the defaults for the icon support pulled from all.js.
pref("browser.chrome.site_icons", true);
pref("browser.chrome.favicons", false);
browser.chrome.site_icons refers to<link rel="icon"> tags which will only load an image if explisitly requested by the page. Whereas the browser.chrome.favicons aka favicon.ico is off by default but can be turned on by default if set to true in prefs.js which will request favicon.ico on every site that you visit regardless of weather the image exists or not.
Unfortunately, a lot of the work Mozilla does _is_ tight number-crunching loops of various sorts. What do you think layout is? It's a lot of recursive number-crunching. So yes, the compiler is making a large difference here. Going from -O to -O2 with gcc (the milestones use -O) leads to a 15% speed increase pretty much across the board for all operations (page loading, new window, etc)
Try right-clicking on any link. The context menu that pops up has an option for opening a new tab. Likewise, right click on the newly appeared tabs and you can close them.
It seems to me, from my experience with Mozilla, that after 4 years and starting from scratch, it has nearly reached the quality and usability of Netscape 4.7x. I have yet to see any really major feature enhancements (besides tabbed browsing and support for multiple mixed - IMAP, POP3 - email accounts) that would prompt me to replace IE with it as my or my company's default browser or replace Netscape 4.7x as our default email client. And it is still too darn slow!
I've had IE5, Netscape 4.7x and Mozilla on my desktop for almost two years now. IE is our corporate standard, so I only use it when forced by certain web pages. Until the middle of this year, Netscape was my day-to-day browser. I would load a Mozilla nightly or milestone on a regular basis, but found that it was a bit too buggy and unstable to meet my needs. I was waiting for the day when Mozilla was ready.
That day came in early August with the release of 0.9.3. Ever since, I've been using a version of Mozilla as my default browser, with no problems. Sites that are "optimized" for IE load just fine, and I grit my teeth when I have to go back to Netscape (regression testing, dont' you know). In fact, I've been so damn pleased that I decided to skip 0.9.5, wait for 0.9.6 (downloading right now) and continue running a nightly instead (20010928-09-trunk, to be exact). Sure, there were bug fixes, feature enahancements and performance imporvements that I didn't get, but what I'm running right now worked great. When you skip an application upgrade because you don't see the need, then you have a good appliction on your hands.
Mozilla is a good application that I'm going to push as our new corporate standard.
eric
The welfare of the people has always been the alibi of tyrants. - Albert Camus
Go figure. Who would have thought that Mozilla users would be asking for a feature that Lynx has had for years. :-)
This is one of many reasons why I keep Lynx around: when I'm using a web interface to a bug-tracking system, and I want to, say, paste some code in to the "explanation" textarea before I close the report, I can just pop into my $EDITOR.
I don't know of any other *nix browser which lets me do this (but I haven't looked very hard).
You cannot apply a technological solution to a sociological problem. (Edwards' Law)
Have you tried getting/making patches for 0.9.5->0.9.6, fetching Debian source and making your own 0.9.6? I've found the Debian mozilla packages rather slow to update, so that's probably what I'll do to get to 0.9.6, if it's feasible without pain.