I unsubscribe.
I think it’s just a perception that there are so many things for firefox because it comes with some of these things per default.
Download IE7Pro www.ie7pro.com and you have all this for Internet Explorer too, full scripting api for creating addons and superb ad blocking and all those things that supposedly make Firefox better.
It’s great and free, the only difference is that IE doesn’t come with it from start.
Maybe comparing default Safari to default Firefox is apples to oranges. I tend to think so. If you’ve gone out of your way to install Firefox, and you’re using it on a normal basis, than that “installation” part is probably not that much work. Safari and IE tend to market well to people who just don’t care, and want to just browse the internet.
I install Firefox because of the myriad of tools available that just aren’t present (or that just plain SUCK) in other browsers. Ever tried debugging javascript in IE? How does that compare to Firebug? I’d much rather start with a small installation, and build on it as I would like. I definitely don’t want the kitchen sink, in which I don’t use half the features.
I do think, however, that Firefox performance with TONS of Firefox extensions installed needs to be looked into.
@Paul
I honestly don’t think you looked very hard, there are myriads of tools and scripts for IE7 too. Perhaps not as collected to one single website as it is for Firefox.
This is from the same guy who wrote several pieces about trading features for simplicity, feature bloat, etc.
Just make up your mind already. It’s disingenuous to complain about software becoming too complex to use and learn, too bloated to download, and then criticize a piece of software for literally not having enough features that “everybody” wants.
Draggable images
already in firefox 3, actually, it can do more than safari, you can drag elements across tabs, which safari can’t; you can also drag combination of elements, while safari can only let you do one at a time
Resizable text areas
this is first an extension for firefox 4 years ago, since when it becomes a “steal from safari” topic?
perhaps a good middle ground should be an “uber extension” package system whereby users can download packs containing several popular extensions based upon their personalities. Alternately, Mozilla could mine usage data on extensions to see which ones should make it into core, much the same way the Drupal CMS team does with its highly extensable product (there are thousands of useful Drupal “modules”)
How many of those nifty items in your home are specialized functionality-- and how many are crude spackle over missing items that really should have come with the rent? I mean, surely everyone needs “a bed”? Why not include “a bed” in the rent?
The Firefox development team does look to extensions for guidance on features to include – that’s where things like the tab management improvements and session store and spellchecking in Firefox 2 came from, for example. In FF3 we have the download statusbar and some search-keyword management elements at least, and we spent a lot of effort on infrastructure around download management and bookmarks/history to make them more fertile ground for experimentation, to say nothing of improving the security, memory efficiency, and ease-of-installation around add-ons themselves.
But that an extension is popular among the set of people who install extensions and blog about it, or that some of the extension’s features are candidates for uplift, doesn’t mean that that extension itself is ready or appropriate for 165M users to have by default. The bar is much much higher there in terms of robustness in the presence of unusual configurations (how many of those extensions work correctly in right-to-left locales? with a screen reader? in the presence of ligatures in the text?), efficiency, user interface polish. Where we have incorporated extension features, it has always meant a large amount of work, often in concert with the author of the inspiring extension. Successful extensions are usually exactly that: great sources of inspiration, typically not very effective sources of ready-to-serve code.
I think it would be a horrible idea to add those extensions to the browser core. I have about 20 extensions installed, but I have only 3 of the top 10. Adding these extensions to the browser core takes away the customizability that makes Firefox better than any other browser out there. I don’t want to have either unnecessary features, like I do with Opera, or absolutely nothing at all, like I would with IE. Firefox is a perfect middle ground.
couldn’t disagree more about incorporating the top 5 extensions into firefox. your top 5 extensions aren’t the same as my top 5 extensions and either won’t be the same as my wife’s top 5 extenstions. forcing bloatware on users is an too common occurrence. let’s keep firefox on a diet, shall we?
This is a good discussion, and represents a lot of the tradeoffs we (as the Firefox development community) do in fact face every time we sit down and think about how we want to improve Firefox.
- What is core to a web browser?
- What will be useful to the majority of our users?
- How can we ensure that the improvement is worthy of the risk it involves, either to the functionality or simplicity of the rest of the product?
In short, a lot of the questions that come up on a href="http://www.37signals.com/svn/posts/913-question-your-work"this slide/a come up whenever we think about adding a feature.
As for the list of 9 issues, I actually think that Firefox 3 Beta 4 has achieved 5 of them (improved current field highlighting on OSX and Vista, which have OS-native support for that; better font rendering; cleaner download dialog; faster HTML rendering; image dragging). The other 4 are things we weren’t able to address either because the costs were too significant, or the solutions weren’t immediately clear.
We’ll continue to listen and uplift, with a focus on making the web a better, faster, and more straightforward place for everyone, and enabling people with specialized use cases and tasks to customize their browser, and indeed their web (see: greasemonkey) to be the way they want.
Oops, “this slide” – http://www.37signals.com/svn/posts/913-question-your-work didn’t notice the HTML restriction.
It’s a great slide! 
Firefox uses the OS to do text rendering, it doesn’t contain its own renderer. On OSX, Firefox renders identical to Safari 3. On Windows, Firefox renders identical to IE7.
http://codeplusplus.blogspot.com/2008/01/web-font-shorts.html
I discovered that Linux’s text renderer was actually more accurate than either windows or osx.
Obviously, it’s not enough to look at which extensions are popular and base decisions on that. The important thing to do is to look at why particular extensions are popular and examine whether there’s a more suitable core solution or not.
Let’s look at Adblock - why is Adblock popular? Some adverts are really distracting/annoying. Most static/text-based adverts have never really been that much of a problem on the web, but moving and obtrusive adverts are. I see no reason why Firefox couldn’t pause all GIF animations by default, thus eliminating a good chunk of annoying advertising outright. Flash would probably need a more robust solution; Firefox would need to know whether the media is paused or not, and pause it directly via an API if it’s going to auto-play.
NoScript - I think it’s mostly popular due to advertising being annoying (see above).
FlashGot - Many features, not all of which would really fit. Being able to pass files off to an external download manager directly might be a nice Firefox in-built option though.
Download Statusbar - Probably popular because the single-window download manager is a separate window and because it’s harder to track individual downloads without having it fully visible.
“but when extensions reach critical download mass and the same extensions appear repeatedly on best Firefox extension lists, they’re filling a universal need.”
Which extensions are those, then? Those most downloaded extensions are not filling a universal need, they’re filling a need amongst geeks who download a lot of files and like to flip javascript on and off for different sites. Everyone has a different view of which extensions are “must have” - I’m not sure I’ve seen one that appears in every “must have” list.
Put up a blog post with the 5 universal extensions that you think should be in by default, and see if 90% of people agree with your 5…
Why extension should be added to the core browser:
A browser is a browser. I think if an extension adds considerable value to a browser, such as spell checking or highlighting current field, it deserves to be integrated into the core browser installation so everyone can benefit.
If a handful of extensions are used by a LOT of people, then someone needs to test those extensions properly, verify they are compatible with firefox and other extensions, etc.
Why an extension shouldn’t be added to the core browser:
Features such as “Gmail Notifier” doesn’t belong in a browser. It doesn’t benefit the large population of users. Yes, it’s nice since you don’t have to constantly reload pages or keep a separate window open for Gmail.
Or GreaseMonkey. I don’t want to write Javascript to browse the web. You do, but I don’t, and I would guess a large population of people don’t either.
I agree, but the only problem that I think could come from constantly loading each new release with the best extensions, is that it would slowly get more and more bloated.
This, to a certain extent, is what has been happening to Windows since it’s creation.
Hey Now Jeff,
Great post! Cool Iris is kinda nice add in too.
Coding Horror Fan,
Catto
It seems like sort of middle ground could be adding a tab in the installer (of course this does not help those that don’t use an installer) which displays a small selection of best-rated mass-market extensions that you can simply click a checkbox to include in the installation (installs as normal extension so you can disable, upgrade etc.). It would be a good opportunity to teach about extensions in general, like where to find them.
Of course, this would increase the testing burden somewhat, because Mozilla would need to test every combination of the suggested plugins in many, many different settings. Some of that testing is actually already happening with top ranked extensions.