Firewhybother

[Tannin]

Prompted, I suppose, by the Firefox update mechanism that triggered just now, I cruised over to their extensions page and scrolled through the first five pages of stuff - i.e., the most popular extensions. And there was not one that looked to me as if it might be worth the trouble of downloading it. Essentially, they seem to fall into four, and only four, categories. In order of frequency:

1: Extensions that woud be totally useless to me, but might be nice for some other person.
2: Extensions that would be totally useless to me, and totaly useless to any person with intelligence equal or better than a yeast mould.
3: Extensions that claim to add a feature Opera has already had for ages
4: Extension that claim to add something Mozilla has already had for ages.

WTF? Why do people even bother?

 

[Handruin]

Web developer toolbar
Adblock
Filterset.G
Dictionary Search

All seem pretty useful to me and was worth the time spent to download.

 

[Mercutio]

Because some of the extensions do great and worthwhile things. Some of them fall into the category of "Why doesn't everyone use that?"

Perfect example: Linky

Linky lets you select a whole bunch of links at once, right click and open them all in tabs.

So simple it blows my mind, but amazingly useful. Yet this is something I can't do in other browsers, and something that, since I have it, I can't do without it.

Want another one?

DictionarySearch

Select a word, anywhere in your browser, right click on it, and Dictionary- search opens a tab with a dictionary definition for it. It's a quick spell check, if nothing else.

How about another one?
IEView and IEtab

These two guys, taken together, let me tag a site as "IE Only", such that if I visit that site again, it opens up a new tab with an Internet Explorer instance in it. All my banking can be done through Firefox these days, but the jerks to whom I pay my electric bill make me use IE... which I do, through Firefox.

Want another one?

SyncMarks

I store my bookmarks on an FTP site, and every so often, changes to my bookmarks are synchronized to all the other computers I use. No wildly different bookmarks on different computers. Huzzah.

One more:

CustomizeGoogle

This removes all of Google's text advertising, sanitizes the information I send to Google, adds links to different search engines to my Google results, adds predictive choices to my searches and has additional functionality for almost every google subsite.

How are those tools not useful?

 

[Buck]

How are those tools not useful?

Remember, you're talking to an owl.

 

[timwhit]

I guess you can't teach an old owl new tricks.

 

[mubs]

Tannin, I guess the design philosophy of FF was to write just the core and let users add whatever the features they want. A savvy user could end up with a highly customized browser. You keep harping that it's made for idiots (thanks for the compliments, BTW), but this aspect of it, allowing extensive customization, makes it a product for geeks. See it the way you will.

I use only one extension: Tabmix Plus. It allows great control of tab behaviour, and I've been using it (or its predecessor) since day one. Can't live without it.

 

[LunarMist]

It would be nice if they sold a version with all of the extenders included and provided a custom installation option where the desired ones could be selected.

 

[CougTek]

I'm no expert on FF extensions (I haven't used any), but I think there are way too many to make an installation package of a reasonnable size.

 

[sechs]

Why does Tannin bother?

He's a closet Firefox user. He can't get enough of it. But has made a public stance against it, and cannot admit that he was wrong. So, he must keep up this polemic to save face.

How can we blame him? He lives in Australia, a country with cities like "Fairy Meadow." The seasons are backwards there. Shipping from Taiwan is cheap. They call people from "the other place" Mexicans.

It's enough to make a man decended from convicts go crazy!

 

[Mercutio]

It would be nice if they sold a version with all of the extenders included and provided a custom installation option where the desired ones could be selected.

That's completely opposite to the philosophy of the developers.
There are premade "Thumb Drive Friendly" versions of Firefox which can include extensions at the whim of the user. The one on my thumb drive has everything *I* need. You could do it too, if you wanted.

 

[Adcadet]

anyone else having problems with FF 1.5 with Adblock and Gmail? Seems either the entire page is blocked or nothing is.

 

[LunarMist]

That's completely opposite to the philosophy of the developers.
There are premade "Thumb Drive Friendly" versions of Firefox which can include extensions at the whim of the user. The one on my thumb drive has everything *I* need. You could do it too, if you wanted.

The philosophy of the developers seems to be "if it ain't broke, keep fixing it." I have no idea where the "premade" versions are, and am too old to have interest in building one. Of course expending effort to continually maintain the Firefox is understandable if it your profession, but for consumers it does not seem to be an economical use of time.

 

[Mercutio]

I've linked it here several times before.

Ahem.

 

[Tannin]

Geez you've got stroppy ever since you made that howler re air safety Sechs. Mate, get over it.

 

[sechs]

I guess that it's good to know that Tannin's humour removal was successful.

Did the doctors say anything about when this crankiness would wear off?

 

[LiamC]

http://www.internetweek.cmp.com/179101486

IE 7.0b v The wily fox v Tannin v Maxtor--meh Maxthon or some such.

Not that I'm wanting to fan the flames or anything... :mrgrn:

 

[Mercutio]

Wow, three ass-sucking browsers and Firefox.

Firefox memory usage never bothers me. I put gobs of memory in my computers so I can HAVE the occasional process use 1.2GB RAM. But some people bitch about Firefox using more memory than they think it should.

Ben Goodger, Firefox Project Manager and author of the ultimate porn-downloading extension, Magpie, had ]this to say on the subject.

 

[timwhit]

Ben Goodger, Firefox Project Manager and author of the ultimate porn-downloading extension, Magpie, had this to say on the subject.

I've been wondering for months why Firefox uses so much memory. Read the comments on that page (they are not positive). Why would I need to be able to go back 5 or 8 times PER tab? I use other applications on my system as well, I don't like it when Firefox uses 300MB+ RAM and even if I close tabs it still continues to use this much RAM. If I minimize the application most of the memory usage is pushed to VM, then when I maximize Firefox it takes an eternity to push all information back into system memory.

This is craptacular application design in my opinion.

I will be changing that setting tonight to see if it makes a difference.

 

[Gilbo]

Personally, I consider it a good use of my RAM. I like it when my computer does what I tell it to pronto, and I frequently go back through several pages when browsing. Your problem likely has more to do with Windows' determination to page out anything that doesn't absolutely have to be in memory this second --a very annoying predilection in my opinion. Everything is beautifully snappy when I use Firefox on Linux even on boxes with 512MB of RAM, KDE, and several other applications running.

Don't blame the program when it's actually the OS .

 

[timwhit]

The programmer should know about the OS' weaknesses when he is designing the software.

Is there any way to change the way Windows pages?

 

[Mercutio]

More Firefox hackage

The programmer should know about the OS' weaknesses when he is designing the software.

Is there any way to change the way Windows pages?

When I took a look on this machine, the relevant value was set to -1. I don't know what that means, but I didn't change it. FasterFox probably did.

As far as changing Paging behavior... there isn't much you can do about it. Internally, Windows uses 4k pages for everything, and flushes them out to disk if they haven't been used in a while, even if you have free RAM available. You can control pagefile size and location from System Properties, and that's about as much control as you're meant to have. I dimly recall some registry settings that helped paging performance on NT4, but it's been ages since I've felt the need to do tuning on that level.

 

[LiamC]

I've been wondering for months why Firefox uses so much memory. Read the comments on that page (they are not positive). Why would I need to be able to go back 5 or 8 times PER tab? I use other applications on my system as well, I don't like it when Firefox uses 300MB+ RAM and even if I close tabs it still continues to use this much RAM. If I minimize the application most of the memory usage is pushed to VM, then when I maximize Firefox it takes an eternity to push all information back into system memory.

This is craptacular application design in my opinion.

I will be changing that setting tonight to see if it makes a difference.

Eh? So it doesn't suit you? How does that make it craptacular? Or are you saying that your choices are the best for everyone?

It is a design decision, which I will point out, you can change. That is a great design decision IMO (allowing the user to decide). In other browsers you might not be so lucky. Personally I use the back button a lot. It may not be a great default setting choice (for you), but it is changeable.

You will never make everyone happy when designing/building everyone. People have different preferences. That's what makes the world an interesting place. It gives us kindergarten painting and Michaelangelo. My prose, Dan Brown, and John Steinbeck. To rail against something because it does not suit you (in the context of the larger world) leads us to Fundamentalism (note I did not say Islamic—I consider Catholic, Christian and other forms, just as "bad"), Islamic cartoon riots and juntas.

 

[Tannin]

Firefox has a worse problem than the memory usage (which, as I see it, is maybe not perfect but no big deal). This is its bad habit of grabbing a massive slice of the CPU and never giving it back.

I meet this problem regularly. It appears to be, however, not a Firefox issue per se, but a Gecko problem, or possibly related to some other feature shared between the Fox and Mozilla, because Moz does it too - possibly even more often than Firefox. (Though this is probably just because I use Moz a heap more than the Fox.) I haven't seen it on Seamonkey yet, but then I'm not using Seamonkey all that much at present. It is probably still there, but time and testing will tell.

I don't know what triggers it, but suspect that it is some particular sort of page, presumably pages with the wrong sort of, or just too much Javascript. Or maybe some other scripting crap. It only comes up every now and again. You don't really notice it unless you do something else that needs a reasonable slice of CPU - for me, this is usually opening a folder with a lot of pictures, and instead of populating the folder with thumbs right away, they pop into place one at a time. If you are trying to find a particular image in a folder that has 500MB or 1GB of image data in it, it is unbearabe.

So you look at what's going on, and sure enough, one or other of the Gecko browsers is hogging 50% or more of the CPU time - even though it's doing nothing! There is no way to get this CPU usage down short of closing every single tab and exiting Firefox. (Or Mozilla, as the case may be.)

I usually meet this problem when I have a gazillion tabs open at once, but I don't think it has to do with that. It appears to simply be something more-or-less random, but the more pages you have open, the more likely you are to get the problem. Something triggers it, dunno what. The other day I got it in Firefox with only about 6 or 8 tabs open, but that's rare. (And, damn it, I should have noted which pages they were, so that I could start trying to nail it down to a particular type of script or something.)

I've met it with several different Mozilla versions and at least two Firefox versions. Never seen it with any of the non-Gecko browsers, and although I don't use IE much at all, you'd think I'd have hit the issue with Opera by now because I use Opera maybe 10 times as much as I use the Fox, or about equally as much as I use Moz/Seamonkey.

 

[mubs]

Tannin, I have the same exact problem, except in my case I cannot even scroll the browser window, that's how slow my system becomes. I've gone to task manager, and it will say FF is not responding, but if I wait a minute FF will begin to respond. This problem usually goes away for me when I close the tab that has the offending web page open.

Because I have a slow CPU, this is the most frustrating aspect of surfing for me. It's something in those web pages that FF doesn't handle well. I haven't noted which pages these are, but in my daily surfing of tech news, I encounter the problem a lot.

 

[ddrueding]

An interesting observation Tannin, this occured to me for the first time yesterday. Only one window open, and here is the history of the browser window:

http://www.google.com
http://msdn.microsoft.com
MS Knowledgebase article

opened into 3 tabs, (one each), and left open for about 6 minutes.

 

[CityK]

Are you guys running your mail app in the background?

For a good while running now, I had noticed that on occasion something was pegging my CPU at ~25%. At first I thought it was simply a case of mal-behaved web pages -- those encumbered with flash crap came to mind. Then came the day that I took the time to close one tab after another to see which was the offender. None of them. "Hmmmm", I says to myself, "What else do I have running?". "Well, Ize gotz thiz Thunderbird running, and itz axing mez to inszrt my password"...."It can't be that. Can it?"...Yep, it was that. Entering password or clicking cancel results in CPU use immediately dropping to 0.

Moral of the story -- "A bird in the hand is worth more then a fox in a henhouse"...No, no, that wasn't it....Um, "where there's fire there's thunder"...No, no, that wasn't it either...Okay, how about "an early password gets the browser"...No, that's not right either... hmm...maybe there wasn't a moral after all.

 

[Tannin]

But that would result in the CPU utilisation being due to Thunderbird, wouldn't it, which is not what I see in the task manager. Or are there wheels within wheels here?

 

[mubs]

Nope. I don;t run Tbird. This happens even when FF is the only program running.

 

[CityK]

But that would result in the CPU utilisation being due to Thunderbird, wouldn't it, which is not what I see in the task manager.

I would persume so..but I can't say for sure...I wasn't looking at what was displayed in the process table, as I was pretty darn sure it was related to FF and not anything else ---> I think I checked originally, as I seem to recall viewing FF being the cpu hog...But when I finally nailed it down, I was just watching the cpu utilization in KSysGuard (which is essentially KDE's graphical equivalent to Task Mangler). I'll check it tommorow, as it is easy to simulate...but right now I'm a little sleepy.

Or are there wheels within wheels here?

It may be the case. I typically launch "read mail" from FF .... sometimes I forget that I opened it up, or don't get around to checking my mail till much later ... and instead of just sitting idle in the background, the password authentication box seems to be doing something which is demanding a fair number of cpu cycles.

Interestingly, up until very recently, I was experiencing a fair number of password authentication requests. For example, I'd enter TB, provide my passwords for the various accounts I would check, begin reading or writing emails, and would be periodically interrupted by the password authentication when the client tried to check for new mail on the accounts. Hard to say what this annoyingness was related too, given:
- since around the time of updating the browser to 1.5.01, this behaviour seems to have stopped...but couldn't pinpoint precisely when it stopped - prior or post, your guess is as good as mine
- email client got updated to 1.5 too, can't remember when or what affect that had on the authentication annoyingness
- email host might have been fiddling with their servers...wouldn't surprise me in the least if it was all due to them.

 

[ddrueding]

The system I was using had no mail cliens installed, it was a server running 2003.

 

[Bozo]

http://www.theinquirer.net/?article=29707


Bozo :mrgrn:

 

[Mercutio]

Thunderbird and Firefox are discrete applications. All the files they need live in their program directories.

Tbird 1.5 seems to have issues with asking for passwords when it shouldn't - I KNOW I'm giving it the right password for my IMAP server but it'll tell me the password has been rejected. Probably a bug somewhere. I reverted to a 1.0 version and that problem went away.

I just watched Tbird (1.0something) and FF (1.51) start on my server 2003 Windows machine. Process Explorer says FF - one open tab (this one, oddly enough) - is using ~2% of my CPU and about 25MB RAM. Thunderbird is using 0% of my CPU and 23MB RAM... and it's polling my IMAP server at home every 2 minutes and connected to a ~2GB mail spool. 25MB RAM seems to be Thunderbird's consistent memory utilization. I checked it on Linux and a couple Windows machines, and it's pretty close to that on all of them.

If I remember, I'll check how much RAM Firefox is using a couple times today. I know it grows quite a bit, but I'm going to try to pin down WHEN it grows.

 

[time]

Eh? So it doesn't suit you? How does that make it craptacular? Or are you saying that your choices are the best for everyone?

This is craptacular application design in my opinion.

It is a design decision, which I will point out, you can change. That is a great design decision IMO (allowing the user to decide). In other browsers you might not be so lucky. Personally I use the back button a lot. It may not be a great default setting choice (for you), but it is changeable.

Firefox uses 300MB+ RAM and even if I close tabs it still continues to use this much RAM. ... when I maximize Firefox it takes an eternity to push all information back into system memory.

A lot of people, including those in the referenced thread, think this is "craptacular application design", regardless of whether or not you can limit the effects by changing a parameter.

And apart from IE 6, in which browsers would you cease to be lucky? Opera Preferences lets you easily control memory and disk usage; finer cache parameters can be set as arguments, but who needs any of this when you can instantaneously access at least 20 prior pages without committing your entire memory?

To rail against something because it does not suit you (in the context of the larger world) leads us to Fundamentalism (note I did not say Islamic—I consider Catholic, Christian and other forms, just as "bad"), Islamic cartoon riots and juntas.

I know who's railing here, and it isn't timwhit.

 

[CougTek]

I surf a lot, most of the time with Firefox, but I've never seen it take more than 76MB of RAM. I wonder what you people do to make it climb up to 300MB. Also, I don't understand people regularly opening 10+ tabs and windows. You got to be lazy as Hell not to use back/forward or your bookmarks instead. Takes two seconds, one mouse click, a simple wrist gesture. It's barely longer than switching tab/page and probably less confusing.

 

[Mercutio]

I usually have at least 20 or 30 tabs open, and at times I exceed 100. I usually have two open Firefox windows, one for porn and one for work. I use the Linky extension a lot (and not just for porn) which contributes greatly to having lots of open tabs. Right now I have 16 tabs open on this computer and memory use of 97MB, which doesn't strike me as unusual, but I actually restarted Firefox today, which is something that only happens maybe bi-weekly.

The highest I've seen from FF is around 2.4GB, but that came from a link JoJo sent me that I think was made to do that. I see FF using 1.5 or 1.6GB pretty regularly from my "normal" use (which is to leave FF and dozens of tabs open for weeks at a time).

 

[mubs]

You lose, Merc. Tannin always has 65,536 browser windows open, each with 65,536 tabs.

 

[CougTek]

And the owl doesn't unserstand why his browser is slow...

 

[LiamC]

A lot of people, including those in the referenced thread, think this is "craptacular application design", regardless of whether or not you can limit the effects by changing a parameter.

And apart from IE 6, in which browsers would you cease to be lucky? Opera Preferences lets you easily control memory and disk usage; finer cache parameters can be set as arguments, but who needs any of this when you can instantaneously access at least 20 prior pages without committing your entire memory?

So? If you all think it is a "craptacular application design", that is your choice. I say that you are targetting the wrong thing, as it is changeable, it is a default choice, not a design decision, and, not everybody experiences this issue, so it probably isn't even a design issue at all. It may relate more to individual user setups, and given the number of possible combinations of of user setups, allowing the user to choose may be the only sensible choice.

As for the browser bit. You named IE, validating my statement. Your point about Opera? This is a FF discussion. What has Opera got to do with it? Opera works? Great. But if you are going to introduce other browsers, what happens with K-Meleon, Safari, Avanti, Maxthon, IE 7, and a host of others. I don't know, so that is why I did not mention specific browser functionality.

To rail against something because it does not suit you (in the context of the larger world) leads us to Fundamentalism (note I did not say Islamic—I consider Catholic, Christian and other forms, just as "bad"), Islamic cartoon riots and juntas.

I know who's railing here, and it isn't timwhit.[/quote]

Now this just sounds petty. This is a user discussion forum, where people are allowed dissenting opinions. Nice bit of selective editing too. Why did you leave out the important part—the part putting the above into context? Are you miffed (still) because I disagreed with your comments on CVT?

Note also, when I post, I do try to explain why I agree, or disagree. This being a discussion forum, if I post why, it gives others a chance to either change their viewpoint, or respond and tell me why they think I'm mistaken, i.e. attempt to make me change my viewpoint. The above is just an ad-hominem.

 

[time]

You named IE, validating my statement.

I excluded IE 6, partly because it's about to be replaced.

Your point about Opera? This is a FF discussion. What has Opera got to do with it? Opera works? Great. But if you are going to introduce other browsers, what happens with K-Meleon, Safari, Avanti, Maxthon, IE 7, and a host of others. I don't know, so that is why I did not mention specific browser functionality.

I was clearly responding to your statement: "In other browsers you might not be so lucky."

Are you miffed (still) because I disagreed with your comments on CVT?

Your rapid raising of the stakes in these discussions is explicitly why I did not respond to you about CVT. Some of your claims in that thread were in fact wrong, but it became impossible to point this out without you taking personal offence.

The above is just an ad-hominem.

Whether you meant to or not, I feel you made a thinly veiled attack on Timwhit without substance. I understand and empathize all too well with your current frustration, but "rabid" is a word that springs to mind. Chill, my friend, you're looking in the wrong place for dragons to slay.

 

[CityK]

Okay, I checked that password authentication issue. I opened up two instances of KSysGuard, having one display system load and the other the process table.

Sure enough, as soon as the password authenticate box comes up, cpu usage rises ~20%. Looking in the process table shows two firefox processes (firefox and firefox-bin) and similarily two for thunderbird (Thunderbird and Thunderbird-bin). The apps themselves aren't using any of the processor, but their "app-bin" process are each using ~1% approximately every other second. Still, nothing to account for the big spike. "So then", you ask, "What exactly is eating all the remaining cycles?" Well, apart from a few other small processes, the newly identified culprit is none other then X, which is bouncing around ~15%.
Somehow I doubt the client_app-server interaction is supposed to cause such an effect.

As soon as I enter the password or, alternatively, click cancel, cpu usage returns to its expected level (given X drops to ~2-4% ... which, this remaining amount, I would hazzard is attribute to interaction with the KSysGuard clients)