Loading ...So what’s bloatware anyway, Part II
So what’s bloatware anyway? For a web browser, my definition is whatever a browser includes that does nothing to improve the web browsing experience of 90% of its users. Feel free to lower or rise the percentage to suite your preference but not too much.
With this definition in mind let’s look for the bloat. From posts I’ve read around, it seems Firefox 1.0 is regarded as the lightest version of Firefox so far, so let’s review what features have been added since:
Firefox 1.5 brought, from the release notes:
- Automatic Updates, as recently reported, this feature alone sets Firefox apart and ahead in the whole software industry not only web browsers. 95% of Firefox 2 users are up to date so I think it’s safe to say 90%+ is using it.
- Faster browser navigation, I frequently press the back and forward buttons and I guess 90%+ people does it.
- Tabs drag and drop. I don’t usually do this and I know how. I guess less than 90% has mental powers strong enough to guess they can drag it, so I’ll mark it as bloat until it becomes obvious that you can move the tabs around.
- Clear private data. I use this, not very often but better to have it there than not. And
- Improvements to popup blocking. I’m pretty sure I’m seeing less popups every day, and so should be 90+% of users.
- Broken web sites reporting. Mainly for developers or knowledgeable users, conscious developers or knowledgeable users. 90-% for sure. Bloatware.
- Support for SVG, CSS2, CSS3 and JavaScript 1.6. I don’t know the web other way and so do 90%+ of users.
Firefox 2.0 brought, again, from the release notes:
- Visual refresh. Whether I like it or not it’s not a feature but a change in data just like changing default bookmarks or default search engines.
- Anti-phishing. I use this one. Very unlikely I will fall for a scam but when I get one of those emails that link to a phishing site I use it to report the site. Who may not want to be warned of a phishing attack? Even if it isn’t 100% accurate but the best available according to a past study.
- Enhanced search capabilities: open search, search engines manager, search engine detection, all welcome and in use. My only reference is a past Mozilla Links poll which showed 67% of voters have added one or more search engines. A high number, but sticking to my definition I’ll label this one as bloat.
- Improved tab browsing, including mainly tab overflow. I use it everyday. Not the List All Tabs button but I scroll with the mouse wheel all day. Not sure who would prefer two letter descriptions for their tabs.
- Session restore. I like the convenience of being able to close Firefox knowing all will be there when I restart it and can’t imagine it other way. 90% of users appreciate having all their tabs restored and emails recovered where they were after a crash. I’m just guessing people likes free insurance though.
- Web feed preview and subscription options. I am all for Live Bookmarks so don’t really need more preview and subscription options. I’m pretty sure less that 90% of people have a clue of what web feeds are. Don’t want to think about RSS or Atom specifics. Bloat.
- Spell checking. In use every day. My guess is everybody writes long text some day and we all need and appreciate good spelling.
- Live titles. I haven’t added one yet. Bloat.
- Improved add-ons. Not a new feature.
So, according to my definition and assumptions, these features have, feature-wise, contributed bloat to Firefox:
- Tabs drag and drop.
- Broken web sites reporting
- Enhanced search capabilities
- Live titles
- Web feed preview and subscriptions
Note that I don’t consider them bloat personally but it’s likely that 90% of Firefox users would consider them as it as most probably they don’t use it.
Live titles is a case apart because it involves a new technology which success mostly depends on content providers using the technology. Just as web feeds become more and more prevalent making Live Bookmarks more and more relevant, I expect the same to happen to live titles in the next few years.
I would like to know precise numbers of how much system resources consume each of these features: I imagine tab events listeners and drag and drop code, user interface for web sites reporting, additional web pages checks for search plugins, the search engine manager user interface, a clock listener that keeps checking for live titles to update and user interface for web feed preview and online web feed readers. All of this remaining untouched. In summary: a couple or trio of listeners and some user interface, aside from the code that is just not executed.
This will probably remain a subjective matter as no specific metrics are known, but it doesn’t sound like a big deal to me.
Previous: Part I: The forgotten ware
Coming next: Part III: Memory issues
Subscribe RSS
Subscribe email
May 24th, 2007 at 11:29 am
“Tabs drag and drop” is a very useful feature!
[Reply]
May 24th, 2007 at 11:34 am
Saul, indeed it is if you know it’s there. How useful can a DVD player be if it’s hidden below the driver’s seat. It depends on whether you get to learn it’s there or not.
[Reply]
May 24th, 2007 at 1:12 pm
Even though I found the article very interesting, I think your definition of bloat is simplistic and wrong.
First of all, there some features which are rarely used, because of their very nature. This includes things like Page Info dialog, custom font and color settings, cookies manager, or even printing. All of those fall under the 90% mark, yet no one would reasonably want to cut them.
Second of all, your approach leaves too little room for innovation. It takes time for certain kinds of innovative features to catch on as users get accustomed to the new concept. Live Titles (microsummaries) is an example of that. I don’t if they’re going to catch on, but even if they’re going become the hugely popular the initial user base has to be small.
Number three: you are overestimating average users’ ability and knowledge. I can assure you that History is used by less than 90% all users. You’re not proposing removing History, are you? The “bloat threshold” should be much lower.
Finally, the most important point. If you evaluated the usage of all the features, you may see that there are twenty features which are used only by 7% of users. The point is that IT IS A DIFFERENT 7% OF THEM EACH TIME, so removing these things would affect half of the users, not 7%. I reckon that the amount of people that need a feature is only part of the criteria.
Let me describe my own philosophy by taking tab drag and drop as an example. It’s true that probably most people don’t use it. However, this feature is totally unintrusive. It’s right there if I ever need it; if I don’t then it’s invisible. It doesn’t slow anything down, it doesn’t increase memory consumption, and it doesn’t add any additional UI whatsoever. How can it be called “bloat”? To me, it’s a perfect example of a not-bloating functionality.
[Reply]
May 24th, 2007 at 2:37 pm
Ancestor, the initial assumption of the article is that in the beginning, Firefox 1.0 had no bloat, so I didn’t analyze Firefox 1.0 features.
Regarding the specific case of drag and drop, as I said I find it useful and while not discoverable at all it may be called a keeper because of its unobtrusiveness. However not because a feature is unobtrusive it can’t be considered bloat: think about Microsoft Office or Windows itself. There’s so much code, apps and services that remain hidden somewhere and doesn’t bother my interaction but are there stealig previous CPU cycles. The thing is it’s hard to measure how much it is taking to be able to balance cost/benefit.
Having said that, the purpose of the series is to dissect the whole bloat issue from an end user point of view. So the definition is over simplistic on purpose. But my point is that even from an oversimplistic point of view, a brief analysis reveals that claims of bloat creeping out are at the very leats, not accurate.
[Reply]
May 24th, 2007 at 5:07 pm
I’m sorry but you must be kidding me… it’s definitevely bloat - a feature, that should be included in extension.
And you’re saying that tab reordering is bloat because users don’t know how to do it. I can name 50 things in Firefox that work but you have no idea unless you try. Interface isn’t perfect, how could you possibly mark that tab can be moved around? ;)
Mozilla should allow choosing officialy recommended extensions while instaling Firefox, not move closer to Opera.
[Reply]
May 24th, 2007 at 5:33 pm
Riddle, so you don’t care about correct spelling?
For some time, Firefox 2 nightlies featured a move icon when hovering over tabs, it was obviously phased out later, but at least it has been tried.
It would help if you mention some of those 50 things. If it is true, Firefox would be a discoverability nightmare, which obviously isn’t. Note that drag and drop is the only way to reorder tabs which makes it a discoverability issue. The dozens of mouse and keyboard shortcuts are precisely that: shortcuts for two or three step discoverable tasks.
[Reply]
May 24th, 2007 at 7:41 pm
I don’t see how drag & drop tabs are bloat, seeing as I use it almost everyday.
And I find it odd that you think that the Web feed preview and subscriptions are bloat, seeing as it gives the clueless web user the ability make use of those features, as otherwise they would just be confused as to what exactly is these lines of XML code.
The only thing I think counts as bloat is the arrow that displays all the tabs in one click, the thingo on the side of the browser bar. I haven’t used that once in all my time using firefox.
[Reply]
May 24th, 2007 at 8:21 pm
I guess I am part of the less than 90% that use ‘Tabs drag and drop’
But for the rest of these:
* Broken web sites reporting (never used)
* Enhanced search capabilities (added engines but rarley use)
* Live titles (what the heck is this?)
* Web feed preview and subscriptions (that is what I use Flock for)
I would agree with you on!
[Reply]
May 25th, 2007 at 8:34 am
The biggest bloat that I find is when I see things like “add-on X is going to be, or has been added to the next version of Firefox”. If an add-on is really successful, with lots of people using it, why should it be added to the core code?
I have two add-ons. Since day one I’ve had one add-on (IEView) and only recently added my second add-on (Canadian dictionary for the spelling engine). When they changed the tab behaviour to include some add-on that I don’t remember (puts the X on each tab and a few other things), I was one of the people that switched my UI back to the old style.
In effect, I am undoing added features because while they were popular add-ons, I never wanted them. I use the keyboard for closing/opening tabs … making the UI “better” to help mouse users doesn’t help me.
These add-ons that become core components are all bloatware to me.
Put another way … if it doesn’t help with the display and interaction with a web page, does it really belong as part of the core browser?
I too started to use Firefox because of the slim application. Now I’m finding it harder to argue that Firefox is better when IE loads faster on a friend’s or family member’s computer that is a few year old.
[Reply]
May 25th, 2007 at 10:33 am
My logic sees it the other way exactly: if an extension is used by a large number of users it would make sense to include it by default and save them from installing the same extension. Now the thing is what is “large” enough. An extension with 1,000,000 downloads would have just proved to appeal to 1% of the estimates Firefox userbase.
Considering there are 1000s of extensions out there now it’s very very hard for Firefox to include a feature that is not already available as an extension. Then there’s the main problem with extensions: they don’t receive the full QA treatment Firefox or any other Mozilla product does.
However, this is not stopping Mozilla to further ease extension development with the inclusion of FUEL, a set of libraries for extension developers. FUEL 0.2 should be ready for Gran Paradiso Alpha 5. An added benefit of FUEL is that developers won’t have to implement some “low level” functions but instead will use the available API which has all the necessary QA.
In the end it’s all about the interaction with the web page. Please provide some examples that don’t. Bookmarks, history, page save, open, options and similar which seem to be UI oriented, are all for a better, more personal interaction in the end.
I may be wrong but I don’t remember Firefox loading faster than IE, ever. Windows preloads a lot of IE components as they’re part of OS now, so it’s very hard to beat it.
Firefox has plenty of advantages though: much smaller security vulnerability exposure windows, extensibility, themes, spell checking (seriously, how can people don’t care about spelling? I find it insulting when I get a messy message), web standards, and of course a much cooler provider and community. ;)
[Reply]
May 25th, 2007 at 6:17 pm
Tab drag-and-drop: wasn’t that something lots of tabbed browsing extensions added? Being able to keep your tabs in order is a fairly fundamental part of the tabbed browsing paradigm.
Broken web sites reporting: maybe not many use it, but it’s an essential part of a browser that does things so differently to IE.
Support for SVG, CSS2, CSS3 and JavaScript 1.6: if support for web standards is bloatware, go use IE5.02.
Enhanced search capabilities: search is also a fundamental part of web use. The ‘Search Bar’ makes it easier to use and has pretty high user adoption. The Open Search system and manager just makes it easier to use a capability that was already in Fx1.0
Web feed preview and subscription options: you may only use Live Bookmarks, but there are lots of feed readers out there and lots of people who use them. The ability to integrate a feed reader into the feed subscription process is fairly essential.
While Fx 1.5 & 2.0 have added features, most are simply implementing best or most popular practice. Even if some people choose not to use them, they don’t get in the way of what Fx was made for: browsing the web as simply and effectively as possible.
[Reply]
May 27th, 2007 at 3:30 pm
Interesting.
It got me to thinking too, although I worry more about extraneous features that result in “unpredictable behavior” (from the eyes of a non-power user). For example: where did my tab go? (When it gets accidentally dragged.)
I can see how the live bookmark indicator may seem non-essential, but I like that it gives users an opportunity to question/learn about rss feeds and how they can use them. Also, I think it’s necessary for users to discover them without having to search to install an extension.
I think I just started using the spell check suggested spellings yesterday….wow, it shows up first in the context menu…just like in word processors so it’s completely intuitive…bloat or not I think it should stay.
I’m still having trouble with users uncomfortable with tabs…especially more than one. I’ve started using stylish and a script to make the active tab stand out more.
I think my users also really like speed dial (new extension) because it gives them the ability to visually choose a bookmark. I’m considering making it the default homepage in my installations from now on.
Also I think the proxy button should be standard. But it doesn’t need to show up unless you use a proxy. (Which is the minority of users). It’s useful for company laptops when they travel away from the work place.
I think ietab should be standard for the windows binary.
[Reply]
September 18th, 2007 at 10:52 pm
The fact remains that the dmg for Firefox 1.0.4 was a light 8.7 MB. Firefox 2.0.0.6 installed is pushing 50 MB. Huge difference.
[Reply]
September 18th, 2007 at 10:58 pm
It is my personal opinion that Firefox has turned into bloatware.
Firefox is becoming sluggish and unpredictable. Perhaps someone needs to rewrite it from scratch? Java and Flash are a mess. Seems Firefox has been slowly allowed to become bloatware …
[Reply]
April 22nd, 2008 at 1:34 pm
Yeah — I’ve been reading these with interest, but somehow I can’t equate an evolving browser reaching maturity to ‘bloat’. If I want lean’n'mean, I can have it. If I DO want bloat, I can add it (and more memory too!). That is only ONE of the sweet things about FF.
I don’t know . . . even in FF3v5b, I wouldn’t say there is any bloat — yet. Just my 2¢ opinion — not worth a holy war.
[Reply]