The Secrets of Firefox about:config

jcatcw writes "While Firefox is very customizable, many of its settings aren't in the Options. Each setting is named and stored as a string, integer, or Boolean in a file called prefs.js and accessed via about:config from the nav bar. Computerworld provides instructions on 20 tweaks for speeding up page loads, making tabs behave, reducing memory drain, and generally making the interface act the way you want it to. Customization also comes through the must-have FF extensions (but be sure to skip these)."

While it's nice..

[microbee (682094)]

Do not tune stuff that is hidden unless you know what you are doing.

Re:While it's nice..

[Zoxed (676559)]

> Do not tune stuff that is hidden unless you know what you are doing.

s/Insightful/Redundant/

This is Slashdot: we all think we know we are doing :-)

Re:

[Yvan256 (722131)]

That's why I prefer Opera. If you think Firefox is bad, try it on OS X.

Re:While it's nice..

[hahafaha (844574)]

So if these options really make pages load faster, offer less memory drain, and even feed the dog, why aren't they a part of the settings to begin with?

Basically, because, although they may give more speed, they have drawbacks as well. Your question is like asking, ``If people can overclock their processors to so much faster, why isn't it overclocked by default?''

Re:While it's nice..

[bunratty (545641)]

Firefox, while better than IE7, is a fucking hog and getting worse by the release. Why should users, who are already iffy about switching to Firefox, have to go through archaic setup commands in order to have the browser work well?

Firefox works well with the default settings. If you have to go into about:config and twiddle parameters, there is something very wrong with your Firefox setup. Try creating a new profile and see if that fixes the problems. In reality, the settings suggested in the article can slow Firefox down (by setting the initial paint delay too low), cause sites not to display properly (by using pipelining with servers that don't understand it), and get you blocked from servers (by setting the maximum connections way up). You can avoid all that trouble by sticking with the defaults.

Re:

[Yoozer (1055188)]

If you have to go into about:config and twiddle parameters, there is something very wrong with your Firefox setup.

• The "oomph" noise when you can't find something. Nowhere in the Windows sound scheme. Nowhere in the regular settings. Why?
• Old style tabs - I don't want closing buttons on every tab, I prefer the 1.5 way of handling things.
• 30+ tabs on a screen without having it absorb 'm in a list, I prefer the 1.5 way of handling things.
• (something I haven't killed yet) - the entirely superfluous usage of

Re:While it's nice..

[by Anonymous Coward]

I particularly love the "pipelining" part. Send requests before getting valid acknowledgments from previous requests. ...

It's rude, annoying and breaks the rules/protocol.

From RFC 2616 (HTTP/1.1) [ietf.org] section 8.1.1:

HTTP requests and responses can be pipelined on a connection. Pipelining allows a client to make multiple requests without waiting for each response, allowing a single TCP connection to be used much more efficiently, with much lower elapsed time.

Re:

[MrNaz (730548)]

If it's valid behavior according to the protocol, and it's faster, and it's not bad nettiquette, then why, pray tell, isn't it on by default?

Re:While it's nice..

[maxume (22995)]

It isn't faster for everybody, it doesn't work with all servers...

Re:While it's nice..

[jesser (77961)]

If it's valid behavior according to the protocol, and it's faster, and it's not bad nettiquette, then why, pray tell, isn't it on by default?

Because some servers violate the protocol by responding incorrectly to pipelined requests. At least, that was the reason 2 years ago.

Re:While it's nice..

[SailorFrag (231277)]

This is a tactic spammers use with mail servers. It's rude, annoying and breaks the rules/protocol.

RFC 2920 [rfc-editor.org] is the SMTP extension for pipelining. Pipelining is a perfectly valid strategy to reduce the time it takes to send mail by reducing the number of round-trips.

What's rude is violating the RFC that says that certain round-trips are required and the spammers tend to violate those rules (such as asking if a message body can be sent before actually sending it, and waiting for the server's introduction message before the client introduces itself). Pipelining itself is actually quite good.

I won't comment on HTTP pipelining because someone else did already.

Re:While it's nice..

[smitty_one_each (243267)]

http://subversion.tigris.org/faq.html#in-place-imp ort [tigris.org]
Install Subversion, and use it on your config files.
Subversion: it's not just for projects anymore.

Re:While it's nice..

[RobertM1968 (951074)]

Why? He's 100% right! Just follow the instructions and you are all set with no chance of there being problems. You see, the instructions on that web page clearly state in bold letters: "Keep a log of everything you change, or make backups."

So, either:

• Firefox acts weird or doesnt run at all, and you restore prefs.js from backup and have no problems
• or it worsens performance, and you restore from backup and have no problems
• or it improves performance and you happily surf away and have no problems

So, because he is correct, he's a fanboy? With IE, you run the possibility of having to do much more than restore a preferences file if you hose something. With Firefox, if you follow the instructions (and something goes wrong), it takes you a few extra seconds to restore the file to original state and "nothing major" happens (other than a wasted few minutes in total trying the tweaks).

So, if he's a fanboy, what does that make you? Just curious.

Extensions to Avoid?

[CharAznable (702598)]

I thought we agreed that ComputerWorld article was mostly crap...

I just want

[by Anonymous Coward]

I just want to make it stop going to Google when it's looking up a URL. If it can't find it in DNS, I want it to return a 404, not ask some fsking company where they think I should go.

I tried changing every entry that mentions google.com, and sometimes it still queries. WTF!

Re:I just want

[Fry-kun (619632)]

try this setting:
browser.xul.error_pages.enabled

set it to "true"

Re:

[the_cowgod (133070)]

Set "keyword.enabled" to false.

Re:I just want

[jonasj (538692)]

If it can't find it in DNS, I want it to return a 404

404 is an HTTP status code. If firefox cannot find the server you want to connect to, where do you want that 404 to come from?

Tabs

[rustalot42684 (1055008)]

In Soviet Russia, Firefox keeps tabs on YOU!

Camino?

[commodoresloat (172735)]

I still use Camino [caminobrowser.org], a Mozilla-based browser for OS X. Is there a similar guide to configuring Camino options or do most of these work as is?

Re:Camino?

[644bd346996 (1012333)]

Did you even bother to try it out? Camino's about:config page is almost identical to FF's page. Any options that are named the same in Camino as in FF will do the same thing. (Camino is just a different front end on Gecko, and about:config options are almost all Gecko options, not browser specific.)

Re:

[Hooded One (684008)]

about:config options related to the Firefox UI (e.g. tabs, but also other features that necessarily have a different implementation) will generally have no effect in Camino. The reason they show up at all is that purging them would be a lot of work. However, most options dealing with page rendering, javascript, etc. work the same across all Mozilla browsers. As for the missing browser.tabs.closeButtons (and undoubtedly others), the latest Camino release is from a branch made long before the introduction

Re:

[maxume (22995)]

I think this is what you want:

http://www.firefoxtutor.com/39/loc-bar-search/ [firefoxtutor.com]

And really, they should have called Iceweasel IreOx, at least until mozilla.org asked them to stop.

link to one page article

[maj1k (33968)]

here [computerworld.com]

Re:link to one page article

[thc69 (98798)]

Who modded that offtopic? maj1k posted the "print this article" link so we didn't have to wait seven times for ad-ridden pages to load.

Foons!

[SimonTheSoundMan (1012395)]

Well, a lot of these "tweaks" will have negative effects.

Example: nglayout.initialpaint.delay as 0. This will slow rendering of the page as it causes reflows. Fools.

Re:Foons!

[MedicinalMan (1061338)]

Damn right. Here's what mozilla says about nglayout.initialpaint.delay [mozillazine.org]

Lower values will make a page initially display more quickly, but will make the page take longer to finish rendering. Higher values will have the opposite effect.

Re:

[AcidPenguin9873 (911493)]

Example: nglayout.initialpaint.delay as 0. This will slow rendering of the page as it causes reflows. Fools.

From the article, just below the section on nglayout.initialpaint.delay:

Reduce the number of reflows
When Firefox is actively loading a page, it periodically reformats or "reflows" the page as it loads, based on what data has been received. Create a content.notify.interval integer preference to control the minimum number of microseconds (millionths of a second) that elapse between reflows. If it's not explicitly set, it defaults to 120000 (.12 of a second).

Too many reflows may make the browser feel sluggish, so you can increase the interval between reflows by raising this to 500000 (500,000, or 1/2 second) or even to 1000000 (1 million, or 1 second). If you set this value, be sure to also create a Boolean value called content.notify.ontimer and set it to true.

Seems like setting nglayout.initialpaint.delay to 0 and bumping up the reflow interval can get you the page quicker and avoid too many reflows.

Re:Foons!

[daeg (828071)]

Opera has sensible pipelining defaults. Most "Firefox tip" articles have you set them to values that when combined with other network settings makes your browser appear like a misbehaving robot, proxy, or hacking attempt. Firefox with sensible values doesn't get blocked.

A bigger question

[DoofusOfDeath (636671)]

Is why useful tweaks are hidden behind and obscure and risky-to-use interface like about:config. If the tweaks are worth doing, shouldn't they have first-class support in the main configuration GUI?

Re:A bigger question

[jj110888 (791178)]

Perhaps these tweaks are hidden because they are *not* worth doing?

Re:A bigger question

[leathered (780018)]

Listen sonny, as a network admin I perform 'miracles' every day with a CLI and hidden options in config files. It impresses the PHBs, earns respect and keeps my salary up. And now you want to further trivialise my job with more GUI options. Oh for the good old days when all we had were ones and zeroes.

Re: A bigger question

[Black Parrot (19622)]

Is why useful tweaks are hidden behind and obscure and risky-to-use interface like about:config. If the tweaks are worth doing, shouldn't they have first-class support in the main configuration GUI?

One philosophy is to nanny the unwashed masses away from "advanced" options. A second is that there's not a lot of reasons to support every possible option in a UI, especially if some of them are rarely used.

FWIW, I used to change some stuff and it would be back to the default next time I started the broweser. Ditto if I changed it in the config file. It finally took when I changed it in the GNOME configuration manager; I guess it was masking the application-specific configs.

official mozilla reference

[wizardforce (1005805)]

http://kb.mozillazine.org/About:config_entries [mozillazine.org]

Why aren't these real options?

[Sigma 7 (266129)]

Some of these tweaks cut down on memory usage. Given that there are still plenty of computers with 512MB of ram (e.g. notebook computers), you don't want applications pinning 100% CPU or memory as it slows down the rest of the system. This is more important with notebook computers, since a second lost through CPU usage or hard drive thrashing is a second lost from battery charge.

The notebook I'm using right now has this amount of memory, and was easily available in stores 1 year ago. Last time I checked, a web browser should never require the absolute latest system for day-to-day operations (which include having another application in the background, such as a word processor or even MSVC 2005.)

Re:

[telso (924323)]

If you don't want to lose CPU cycles (and therefore battery power) from using your browser, why are you on Slashdot?

a little OT kinda, but the video downloader ext...

[3seas (184403)]

...works really well if you first watch the video you want to download, putting it in your cache. Then going to video downloader, and regardless of the file size, takes just a few seconds and you are done. Apparently it can grab it from your cache and make it a file on you system (very little for it to really do - very low bandwidth to convert).

In fact, it seems to me that when it doesn't work, "service not available" only happens when I don't watch it first, not in my cache.

The Art of Performance Tuning -- a Fable

[sillivalley (411349)]

A long time ago, when computers remembered using little donuts made of rust, I worked on on a mainframe computer system (CP/V) that supported batch, timesharing, realtime, the works. It had performance monitoring tools, and a large basketload of parameters for sys admins to twiddle.

One of our favorite parameters was SL:BB, documented as batch bias, an input to the process scheduler. When someone called or wrote to us saying they were having problems with performance tuning, we usually suggested they redo their tests varying the setting of SL:BB and let us know what happened. Try different values, 0, 1, 5, 20, 50, 100, things like that. Try it and get back to us.

And lo, they would go off and redo performance runs, and report back.

And we would collect their results and go and muse over them, usually over beer.

SL:BB told us a lot about the user, because SL:BB was a knob that wasn't connected to anything. Oh, the value was range-checked by the parameter setting tool, and dutifully stored in memory, and displayed on performance displays, but it didn't change system performance in any way at all.

That's not what the documentation said, but who believes documentation? We had plans for SL:BB, we just hadn't gotten around to writing the code yet.

So if the user reported that setting SL:BB to 25, but not 24 or 26 gave them incredibly better (or worse) results, we definitely factored that into our analysis.

Those that reported back that the setting of SL:BB didn't make a damn bit of difference, and there were some, we honored as brothers, took into our confidences, and shared beer with at the soonest opportunity. Their bug reports and feature requests received far more attention, for they had passed an important test.

And how many of these Firefox parameters are like SL:BB?

Re:The Art of Performance Tuning -- a Fable

[MulluskO (305219)]

"The general.config.obscure_value preference specifies how the configuration file is obscured. Firefox expects that each byte in the file will be rotated by the specified value. The default value is 13. If this value is left unchanged, then the configuration file must be encoded as ROT13. Autoconfig will fail if the cfg file is not encoded as specified by this preference. A value of 0 indicates that the file is unencoded-- i.e. it is unobscured plain text. It is recommended that you set this value to 0. (This will allow you to skip the encoding step in part 3.)"

Hee.

Re:

[zeroduck (691015)]

I guess someone didn't get invited out for beers.

Re:

[syousef (465911)]

YOU are the IT expert, not the end user. If they're not following instructions you're suppose to help them, not hang them out to dry or make them an object of ridicule. You have the professionalism and people skills of a grizzly bear.

Re:

[syousef (465911)]

So that means when I have a problem with an aircraft requiring a service, I should go find an auto mechanic?

The guy's job was obviously to look at the performance of the system and help users out when they had problems. They might be IT users, but system performance isn't their area. In this context they are end users even if they are also programmers. They just aren't specialists in what this guy does, and it's not their job to do his specialised job on top of theirs.

Trust /. to mod me as troll for saying

Hacking Firefox

[Dominare (856385)]

Gah. Why is it that these people insist on calling anything not found on the main options page "hacking"? As for the above questions - usually the reason things like that are 'hidden' is to stop people fiddling with them. A good example is the old 'coolbits' entry in the registry for nVidia cards - the overclocking functionality was there, but you had to do something non-standard to enable it. That way, the company's ass is covered if you melt your card; you can't pretend you enabled the options accidentally. Since Firefox is free and nobody is paying tech-support, I'm not sure why these things aren't available - but the fact of the matter is, anyone messing around with fundamental parameters should _not_ be the kind of person who lets random articles on the internet tell them what to change.

Thunderbird also...

[thejuggler (610249)]

You can configure many settings in Thunderbird using a similar interface. However, in Thunderbird you can get to the config section from the Options menu Advanced tab. I have reduced the size of the attachment icons this way. set mailnews.attachments.display.largeView to False.

Opera Version

[Deliveranc3 (629997)]

opera:config

Hidden slashdot tweaks THEY don't want you to know

[noidentity (188756)]

I've found some secret hidden tweaks you can make to Slashdot articles, just by editing this thing called a URL. For instance, you can set the comment threshold HIGHER than 5 by editing the number after "threshold=", ensuring only the absolute best comments (apparently none in this story meet that standard yet). Why do they hide useful options like this from us?

Re:Hidden slashdot tweaks THEY don't want you to k

[Dogtanian (588974)]

For instance, you can set the comment threshold HIGHER than 5 by editing the number after "threshold="

It works the other way too; you can set the threshold to -2 or lower via the URL. Most people never get to see such posts, which is the whole point- you'd be shocked if you knew what was there. Things modded down to -2 include:

• Secrets of the Illuminati
• The truth behind the JFK assassination
• Clear evidence that Steve Jobs is not God, Bill Gates is not the devil and Steve Ballmer is not *actually* a chair-throwing ape
• Anything linking to Zonk's blackmail pics of Taco and Cowboy Neal

Re:Which option to make the Firehose work again?

[larry bagina (561269)]

set org.slashdot.dont_make_changes_on_the_live_server_ yes_im_talking_to_you_cmdrtaco=1

Re:Still can't turn off favicons in the bookmark m

[Dracos (107777)]

In Firefox 2.0.3, I opened up the DOM inspector, chose the main window, and started drilling down in to the element tree: I found the icons which you loathe.

Open up userChrome.css (in your profile: [profile dir]/chrome/).

In it, the following CSS rule should work to hide the icons:

.bookmark-item > .menu-iconic-left > image { display: none; }

(This selector appears in chrome://browser/skin/browser.css, if you know where that is).

Re:Documentation

[bunratty (545641)]

My question is simple: Is there good, concise documentation of the about:config page and its options?

Yes.

If yes, where is it and is there an an easy why to find it?

MozillaZine Knowledge Base Article on about:config entries [mozillazine.org]

It is the first page Google finds when you search for "about:config". I'll let you decide whether that's easy to find.

If no, why not? If this is all about choice, should people be able to learn about their choices?

If there wasn't, you'd be able to put it in the Knowledge Base yourself.