Author Topic: Firefox 3 Released (Read 70817 times)

icekin · « **Reply #75 on:** June 19, 2008, 06:21 PM »

While i appreciate any independent testing on this website, it seems more and more people are posting their "results" with absolutely no details/supporting evidence whatsoever. For example, what sites were the browsers tested on? was the cache empty? were the browsers using pipelining? what addons/widgets were installed? was any 3rd party software present that may have interfered? etc etc.

Ehtyar.
-Ehtyar (June 19, 2008, 04:45 PM)

Apologies for the lack of detailed information on that post. I did what I'd consider to be a fair test. Both FF and Opera were installed, non-portable versions. FF had the following extensions :

Adblock Plus, Better Gmail 2, BugMenot, Customize Google, Firebug, Gmail S/MIME, IE Tab, MR Tech Toolkit, SeoQuake, Stumbleupon, Stylish, Sxipper, Tab Mix Plus (the version posted earlier in this thread), TinyMenu, Web Developer, YSlow

Plugins : IETab, Java, Mozilla Default, Quicktime, Shockwave, Shockwave for Director, Silverlight, Windows Presentation Presentation

Third Party Software that monitors my connection: Admuncher, Spyware Blaster, IE Spy Ads, Comodo Firewall Pro

Opera had about 35 javascript files in the user.js directory, but of course, many of them (about 20) are site specific, so they only come into play when I visit that site. Many of those 20 sites were not part of my test. Although, I am not sure if Opera loads all the js at once, or only when I visit the site.

Sites I tried out (mostly with poor html):

http://football365.com
http://myspace.com (plus a few random user pages)
And a few other from Vincent Flander's site

Test Machine (Windows) :

Acer Travelmate 4070
Intel Centrino 1733 Mhz
2GB 667 Mhz DDR2
60GB 5400 RPM Parallel ATA
Windows XP SP3 (Nlited, many services turned off)

Test Machine (Linux):

1.8Ghz AMD Athlon
512 MB RAM
Xubuntu 8.04 (Hardy Heron)
7200 RPM 250GB Parallel ATA

I realize that I am not presenting as much details as some would like, but frankly I can't recall all the details of my test. But I would like to say (again with no evidence to back this statement) that I am only making my comments based on my own user experience. I do not declare allegiance to either the Mozilla or Opera camp and in fact, I've used both extensively, along with K-Meleon. Hence, when I state that as a user, I've experienced a problem or enjoyed a feature, its because I actually have and I hope that the developers can fix it or implement it elsewhere. I am not trying to make any product appear better than another through my comments.

Ehtyar · « **Reply #76 on:** June 19, 2008, 07:05 PM »

Thank you for the details, though i wouldn't consider that a fair test at all. A fresh windows installation, and clean copies of opera/firefox would be needed for a 'fair' test i should think.

Ehtyar.

icekin · « **Reply #77 on:** June 19, 2008, 08:52 PM »

Thank you for the details, though i wouldn't consider that a fair test at all. A fresh windows installation, and clean copies of opera/firefox would be needed for a 'fair' test i should think.

Ehtyar.
-Ehtyar (June 19, 2008, 07:05 PM)

I should point out that your average end user is not going to be installing opera or FF on top of a fresh install of XP either. Even worse, it could be one of those default Windows XP installs that comes with the machine, pre-filled with auto-starting programs and other crapware. The ideal goal for developers should be to make a piece of software that can run efficiently on the minimum system requirements advertised, under the practical system environment that an end user is going to have. For an example, look at uTorrent or xplorer2. Both programs have always started and responded, even when my system has 30 other applications open.

My own test and its results are not meant to be used as benchmarks (and neither will they be), but they are certainly fair since all three browsers (Opera, FF and K-Meleon) were run on the same machine and started with clear caches. All 3 browsers were given the same set of resouces to work with. Yet, if some of them can perform in some areas and others can't, I think its an issue with the program, not my testing method.

nosh · « **Reply #78 on:** June 19, 2008, 09:10 PM »

I couldn't bear the wait (not while the whole fxing world was talking about nothing else!) so I decided this would be a nice opportunity to start a fresh FF install. I've managed to find most of my extensions or even better replacements.

Got rid of the titlebar & menus using 'Autohide' (you have to run in fullscreen mode & ask it to always show whichever elements you need to see at all times, the rest can be easily toggled on when needed by getting out of FS mode)

MenuEditor is available (yaay!) - though I haven't had the time to put it to use yet.

Toolbars are not as quirky to customize as before but do take a bit of doing.

The Library is awesome, I botched up some stuff while putting it to the test and instinctively hit Ctrl+Z (I'd never have noticed it in the menu) & could undo every step (don't know if this was a feature in FF2 but it's awesome to not lose bookmarks just coz you're in sleep deprived zombie mode! )

The performance: well, Firefox doesn't feel like Firefox anymore. Can't wait to see how much better it'll get after I give it the full Tweakguides workout.

Thank you, Mozilla!

Ehtyar · « **Reply #79 on:** June 19, 2008, 10:09 PM »

I should point out that your average end user is not going to be installing opera or FF on top of a fresh install of XP either.
...
Yet, if some of them can perform in some areas and others can't, I think its an issue with the program, not my testing method.
-icekin (June 19, 2008, 08:52 PM)

I cannot see how you expect to administer a 'fair' test when you've modified each piece of software from its original configuration in different ways across the board, though i will concede that most users won't be using the software without modification on a clean OS install.

Ehtyar.

Grorgy · « **Reply #80 on:** June 19, 2008, 11:09 PM »

Its exactly the sort of test an end user will make, if it isnt faster/slower/better/worse on his/her system then what it does on an unused PC is irrelevant to them, and I for one appreciate hearing users feelings about a program rather than what it should be capable of in an ideal environment.

Curt · « **Reply #81 on:** June 24, 2008, 09:52 AM »

maybe we should post firefox 3 reviews we find:
-mouser (June 18, 2008, 09:28 AM)

http://firefox.s3.si...com/ff3-revealed.zip
-sitepoint (copyrighted PDF)

- thorough, oh yes, but maybe a little too enthusiastic? $:-\$

tomos · « **Reply #82 on:** July 04, 2008, 05:04 PM »

still havnt installed this

Anyone know ...
I'm wondering if I have versions 2 & 3 installed separately could one run both simultaneously ?
I suspect not, normally FF only allows one instance (unfortunately)

Lashiec · « **Reply #83 on:** July 04, 2008, 05:17 PM »

Sandbox it, or use a VM

nosh · « **Reply #84 on:** July 04, 2008, 06:06 PM »

The single instance thing applies across versions. I had both installed before I got rid of FF2.
42 extensions ported now to FF3, it has crashed a couple of times (possibly an extension incompatibility) but when you consider the night & day difference in performance, upgrading is a complete no-brainer. You can create an exclusive profile for FF3 to avoid complications.

Curt · « **Reply #85 on:** July 04, 2008, 07:29 PM »

- yes, that would be necessary, because if installed the normal way, 3 will update 2

I just tried this and ran FF3 for less than half an hour, before I removed it again. I was annoyed that too many addons were deactivated, more than half of them, but I tried it long enough to realize that I am looking very much forward to install it again, when it is ready for me.

cmpm · « **Reply #86 on:** July 04, 2008, 07:52 PM »

The add on's are better in quality for FF3.
Perhaps the name of the add on is not upgraded.
But there will be one similar and better in most cases.
This has been my experience so far.

patteo · « **Reply #87 on:** July 04, 2008, 09:57 PM »

still havnt installed this

Anyone know ...
I'm wondering if I have versions 2 & 3 installed separately could one run both simultaneously ?
I suspect not, normally FF only allows one instance (unfortunately)
-tomos (July 04, 2008, 05:04 PM)

You could run both simultaneously:
You of course need to install FF3 to another directory and also create a separate profile.

I recommend that keep FF2 as your default as otherwise, when another programs calls for a browser, FF3 would load instead and check for compatability of add-ons that you have been using for FF2 which can be a pain in the neck.
http://blog.codefron...ng-at-the-same-time/

And of course backup your default profiles etc before you begin.

If you are still not sure, better use one of the tools like Sandbox instead.

The only tricky part I have not figured out yet is when I'm ready dump FF2, how do I do it properly- possibly reinstall FF3 again, but this time with the default settings.

A word of caution when using tools like Nightly Tester to force compatibility for add-ons.
I had an instance when an incompatible add-on caused instability for FF3.

Also, go easy when walking the table spread of Add-ons - like doing it progressively, one at a time, use it for a while and see if everything is fine before you add the next one.

It's so tempting as you go through the gigantic list of those available and add 5 or 6 at one go.

My FF3 had a case of indigestion and another round of instability even for those which are already compatible and it becomes hard to identify which one is the culprit when you add more than one at once.

So I have learnt my lesson - slow and steady.

Ehtyar · « **Reply #88 on:** July 04, 2008, 09:59 PM »

As mentioned earlier in this thread, the portable version can be run whilst a fully installed version is in use.

Ehtyar.

tomos · « **Reply #89 on:** July 05, 2008, 07:48 AM »

thanks all
esp patteo &
Ehtyar, - I had missed that earlier

As mentioned earlier in this thread, the portable version can be run whilst a fully installed version is in use.
-Ehtyar (July 04, 2008, 09:59 PM)

Ehtyar · « **Reply #90 on:** July 05, 2008, 06:33 PM »

I'm sorry I was probably a bit snippy there, I forget how much of a pain in the a** it is to read through such a long thread. I would recommend the portable version though, it's the least risky solution, and can come in very handy should you ever need a browser on your USB key.

Ehtyar.

tomos · « **Reply #91 on:** July 05, 2008, 07:41 PM »

I'm sorry I was probably a bit snippy there, I forget how much of a pain in the a** it is to read through such a long thread
-Ehtyar (July 05, 2008, 06:33 PM)

no, no!
& I for my part really meant the "thanks" & didnt mean to sound defensive -
I skimmed the thread which practically guarantees I'm going to miss something

lanux128 · « **Reply #92 on:** July 05, 2008, 09:04 PM »

i've made the transition to Firefox 3 with all my old customisations intact except for one addon - SessionSaver. by now, if any addons haven't been updated, i guess they never will be.

cmpm · « **Reply #93 on:** July 05, 2008, 09:14 PM »

Here's one of a few for ff3 lanux.
If it's what you want....

https://addons.mozil...S/firefox/addon/2324

lanux128 · « **Reply #94 on:** July 06, 2008, 12:00 AM »

thanks cmpm, i've already tried Session Manager but as its "undo closed tabs" feature can't be turned off. i had settle for TMP's session-saving feature. for the record, previously i was using this addon.

tomos · « **Reply #95 on:** July 09, 2008, 06:42 AM »

dont know where to post this - re the forum here & how it shows in FF3 in particular

I open unread posts page
open links of interest in the background
when I click on those tabs that were opened in the background, there's a noticeable jump - the page (and text?) reduces in size like something slotting into place -
anyone else experiencing this ??

I've only seen it with dc on FF3 but saying that, I've only tried a couple other forums, otherwise not had any similar trouble with other sites

f0dder · « **Reply #96 on:** July 09, 2008, 07:02 AM »

tomos: don't think I've had that, but perhaps something related? Sometimes when I switch to another tab, it's as if another tab than the one I'm selecting is rendered for a very short time, and then I get to the correct tab. Haven't looked into what it takes to reproduce, as it doesn't happen very often.

dmtrs · « **Reply #97 on:** July 09, 2008, 02:11 PM »

when I click on those tabs that were opened in the background, there's a noticeable jump - the page (and text?) reduces in size like something slotting into place -
-tomos (July 09, 2008, 06:42 AM)

This looks like the zoom memory, where, by default, the zoom setting is remembered per site (and across sections). It is a new feature in Firefox 3.

To check, you can reset the zoom for donationcoder.com (Ctrl+0, or View > Zoom > Reset).

Zoom memory can be disabled via about:config:

http://kb.mozillazin...er.zoom.siteSpecific

tomos · « **Reply #98 on:** July 09, 2008, 02:45 PM »

This looks like the zoom memory, where, by default, the zoom setting is remembered per site (and across sections). It is a new feature in Firefox 3.

To check, you can reset the zoom for donationcoder.com (Ctrl+0, or View > Zoom > Reset)
-dmtrs (July 09, 2008, 02:11 PM)

that's it all right dmtrs, thanks

strange that it decided to resize one site and not most others - or not in f0dder's Firefox for example !!
Just checking: so if I set zoom to 120% for dc, yes, it does remember ...

I would reckon it was previously at about 110% or less (-I work with scales so I'm a good guage

) and when I clicked on the tab it ^jumped back to 100% (or 10% less

)
anyways, sorted now, thanks

f0dder · « **Reply #99 on:** July 09, 2008, 02:53 PM »

tomos: it's actually a feature of FF3, that it remembers zoom per domain, not browser-wide or per-tab... I quite like the feature, since some sites use ludicrously small fonts

. Guess it might explain my behavior as well... yep, just tested, indeed it does - if I change zoom for one open DC tab and switch to another, first it renders (or shows cached render?) with the original zoom, then quickly re-renders to the new zoom level.