Firefox Lite And Old PCs Could Crush IE

Eatfrank writes "A recent CNet article suggests that Mozilla should pipe a lite version of Firefox into older PCs to further attack IE's dominance: 'Firefox supporters, take note. A bare-bones Firefox will get the browser into more houses, increasing the Fox's market share and keeps it in novice users' eyes for when they get a new PC ... a truly great super-lightweight browser would have the security of Firefox, without the add-ons, without the tabs, yes, even without favourites, history lists and customisability. The Firefox name is synonymous with security and Web-browsing vigilance. Why not give this to the processing lightweights of the PC world?'"

They've had this idea before...

Produce a stripped-down Mozilla light, that will be faster and have a much smaller memory footprint, and will run well on old hardware.

If my memory serves me well, it was going to be called "Firefox".

Re:They've had this idea before...

Actually, it was going to be called Phoenix. Then Firebird. THEN Firefox ;)

Re:They've had this idea before...

I think this comment is just silly. What evidence is there to support this? I mean, Firefox isn't that big as it is. Plus, does hard drive space even matter anymore? The smallest hard drive you can find if you stolled into a best buy is like a 160 GB so what's the point.

It's not about the size of the installation, it's about memory and CPU usage. Firefox is ridiculous in both of these categories. Good luck running it on old (heck, even old-ish) hardware.

Re:They've had this idea before...

When the local Wal-Mart is selling new PCs for $280, who cares about old machines?

That's a horrible attitude to have about software development, and unfortunately I see it all the time from programmers.

Just because machines are faster and have more RAM shouldn't give programmers a blank check to write programs that hog memory and CPU cycles. People should write software to take advantage of that extra performance, not penalize those who don't have it.

If we write inefficient and, honestly, dumb software, on the assumption that hardware will compensate for our bad choices, how is the new hardware an improvement at all? It's like you're purchasing upgrades every year to keep up with the increasing laziness of bad programmers.

Re:They've had this idea before...

If C|net really thinks that removing trivial user-interface functionality like bookmarks and history is going to significantly reduce Firefox's memory footprint or CPU usage, then I would suggest that C|net is not qualified to be giving advice to the Mozilla team on this subject. This guy is C|net's "expert on digital music and portable media"; has he ever even written a line of code?

Let's look at his other suggestions. Removing tabs would probably result in people opening fewer pages at a time, but people are already free to ignore tabs if they don't want to use them. There is no point in removing the functionality. (In fact, I would be willing to bet that one window with three tabs uses less memory than three windows). The same goes for extensions; people are free to not install any and removing the functionality would likely not further reduce the memory footprint.

Yup, basically, this guy has no idea what causes memory usage in Firefox. I'm glad that the Mozilla team will undoubtedly ignore his misguided advice. Here's a hint: the main driver of Firefox memory and CPU use is web pages. Parsing, rendering, and running scripts. Web pages are huge nowadays, with tons of scripting, huge images, and even videos, and all that stuff has to be kept in memory while you have a page open. If you want to make Firefox more efficient, don't look at the UI. Look at Gecko. Unfortunately, this means you have to be a programmer to make informed comments about Firefox's memory use.

Re:They've had this idea before...

"160go hard drive don't work under some old computers (PII, 1st or 2nd generation of celeron with old bios)."

Stick the drive in a newer linux box, and make a small boot partition (/dev/hda1) that is small enough for the current bios (I've stuck 160 gig hard drives in machines that won't see beyond 8 gigs with this trick), so it works. Make /dev/hda2, /dev/hda3, etc in your preferred layout. Format each as ext3, reiserfs, fat32, whatevr your poison. This works for win9x, winnt, as well as linux.

Move the drive to the old box, boot off the cd-rom, install on the only visible partition (the small one).

Reboot, and at the end of the boot process, your other partitions are visible. On wn9x systems, your other partitions are limited to 32 gig each, so make sure you do under that. On winnt systems, you're limited to 128gig (yes, I know, its supposed to be 256 gig, but don't do it ... you might even want to stay below the 32 gig limit and stick with vfat, just to be safe), on linux systems, you're limited to ... whatever your heart desires.

Note: if you're using fat32 and are foolish enough to go over the 32 gig limit, you WILL be sorry. It will appear to format, and it will even appear to hold, say, 128 gig. However, once you try to write past the first 32 gig, it will over-write from the beginning of the partition, erasing data ... so if you're doing this with a windows box, your absolute safest bet is:

1. boot partition the maximum size your bios can see (say, 8 gigs, assuming really old hardware - a p150, for example)
2. 4 partitions of 32 gigs each
3. one partition with the rest.

Opera?

Several comments in that blog already point out the obvious: Opera already does this. It runs perfectly even on old hardware, and you won't even have to sacrifice any functionality. Hell, there's even K-Meleon, which uses Gecko.

"Nate is CNET.co.uk's expert on digital music and portable media"

Expert? He hasn't even figured out that the Opera browser even runs on mobile phones, and using the same engine as the desktop version...

Re:Opera?

You've missed the point entirely.

The whole idea is to create a new FF version that does the things that Opera or K-Meleon do but still carries the branding of Firefox.

I think it's the Firefox team that's completely missed the point. At my firm, we have several hundred PCs with 256MB RAM running Win2K and XP. We wanted to get rid of the buggy IE7 which go tinstalled as critical update on XP.

Firefox and Opera were evaluated - and the latter won. It appeared Firefox was not only 'compatible' with IE and rendered all IE-only pages, it was bloated and clumsy like IE as well. The development team seems to have gotten hijacked by a few misguided elements, probably under influence from Microsoft. Firefox on Windows behaves differently to Firefox on Linux - but Opera stays the same.

The only plus for Firefox is the numerous plug-ins, but what we like to see is pluck-outs that would ensure no memory leaks and lesser footprint. Until those things happen, Firefox will be a product that never reached it's potential.

Re:Opera?

I have found the magic number for ram to be something just short of 512 megs for a bare XP install.

When I sat magic number, I mean the point where more memory doesn't make a difference on a default install. I have always encouraged people to raise their ram to at least this number and on more then one occasion have heard reports of "it acts like a new computer".

If you are using just 256 of ram, you probably won't know what your computer could do so you don't see it as a slow down. However, When Someone who does have enough memory, even if it is the same processor, uses the same computer, they will think it is slow. If your ever using XP with less then 512 memory, try begging, barrowing or whatever you have to do to jump it up to at least 512 and you will see the difference. You will probably wonder why you ever had less. And yes, this does effect just using it to write letters, surf the interweb, and check email. You don't need to be doing anything fancy to see the benefits.

Re:Opera?

Lite version of Firefox is not going to be ever possible unless they drop XUL completely and leave Gecko + 100% binary UI.

You mean something like K-Meleon [sourceforge.net]? Please try it and see if you find it any faster -- I didn't.

Re:Opera?

Chasing after a declining marketshare is a poor business strategy. Windows 98 and ME boxes will be replaced as the years go on.

Current security bugs often require completely different patches to fix the security flaw. The code base that was used to develop Firefox 2, Gecko 1.8, became largely static in August of 2005. This means that security patches for Firefox 2 start taking significantly more developer time as code bases diverge. The Gecko 1.8 and 1.9 have already have significant differences in the code base different graphics rendering platform, text layout and html processing just to name a few.

Firefox 3 and Gecko 1.9 will not run on any version of Windows earlier than 2000. This means that the project he suggests would need to be build off the Gecko 1.8 code base. This code base is too old for new projects to be developed on it. The last security patch on the Gecko 1.8 code base will be about a year from now. This leaves any code using this open to any security issues discovered.

Re:Opera?

"Nate is CNET.co.uk's expert on digital music and portable media" Er, I think that means he's the one who's got an iPod.

Re:Opera?

He hasn't even figured out that the Opera browser even runs on mobile phones, and using the same engine as the desktop version

If you're referring to Opera mini, which is pretty much the only genericly available "runs on mobile phones" browser (as opposed to versions that run on some smartphones), then no, it doesn't use the same engine. Opera mini is a very lightweight client to a proxy that runs at Opera's HQ, which reformats web pages to create much simplier pages containing more or less the same content.

It's very useful, better than nothing, and arguably better than trying to fit a "real" rendering of a webpage onto such a small screen (I also have the Nintendo DS version of Opera, which has a mode based upon this simplified page rendering as well as a full HTML mode, and the former is, usually, more practical than the latter. Something for Apple to think about, perhaps), but it's certainly not using the same engine or anything close to it as the desktop Opera.

Re:Opera?

No, I was referring to Opera Mobile. It has been preinstalled on phones with as little as 4 MB of memory or so. It might not run on all phones, but it certainly runs on phones that Gecko could never even dream of touching. Oh, and it runs on the Nintendo DS Lite, which is underpowered even compared to today's mobile phones.

How much extra work?

This doesn't sound like too bad an idea. One issue would be maintenence - if the full version and the lite version had to be maintained seperately, it probably wouldn't be worth it. To keep relevent bugfixes and such applicable to both branches, the code would need to be well designed and presumably fairly modular. Any Mozilla developers (or people familiar with the code) around and willing to comment on whether this would be feasible?

Just how old are these machines

This post is being written on a machine with a 633 MHz processor. It's fairly ancient. It runs the full version of Firefox just fine. Mind you, it isn't running Windows, it's running DamnSmallLinux.

If I were to want a stripped down Firefox, it would probably be for embedded devices where resources are often quite limited.

Opposite effect?

"Firefox...without favourites, history lists and customisability"

Firefox without favourites? Without history? Let's just get this straight - you want people to switch to a browser which has less functionality than the one they are currently using? Again - a browser without favourites? How is this going to give people a positive experience of Firefox and make them want to do anything but work out how to uninstall it...?

Most braindead idea I have heard all week.

And, as someone else has already pointed out, originally, Firefox was supposed to be the lite version of the oh-so-slow-and-bloated Mozilla Suite. Would that they had stayed true to their original intentions...

iqu :|

Re:Opposite effect?

Firefox without favourites? Without history?

Yes, it's called a default homepage. You build a local index.html that includes the links you want (you call them Favorites, but we in the non-Windows world call them Bookmarks) and load that as your default homepage in FF-Slim.

This is not an issue at all. History might be a problem, but you can always use 'about:cache' or 'about:history' to derive that.

webkit

Well before going to extreme removing everything useful (heck Netscape 3 had a history and I remember running it on really slow computers) why not first change the rendering engine to use webkit which uses a lot less memory? Why do you think phone companies are investing in it over mozilla?

Re:PC-Lite? Hell, I want that on MY desktop!

I suspect he means the history that shows when you start typing in the address bar. While I find that immensely useful, I could see how it would be annoying on a system not fast enough to do it well.

But yeah, he's trolling... The Flash comment proved that. There's nothing wrong with Flash itself, only how people use it. He then goes on to prove that at least 1 person did it well by his own standards, but he refuses to look at any other Flash. If we were talking about humans, this would commonly be called 'prejudice' and people would be up in arms. It's still prejudice (but without the human connotation), but this is Slashdot and tech-prejudice is expected here. (Call that flamebait if you want, but the lines here are clearly drawn and accepted.)

As for 'Firefox Lite'... My immediate thought was 'who the fsck would want a browser without bookmarks/favorites?' But then I remembered Del.icio.us and how much better it does the bookmarks, and that I never actually use them on my browser now, except for a single bar below my address bar. And that could be done away with using a good homepage. (Maybe modeled after Opera's Speed Dial.)

Why not Lynx?

Excuse me: "without the tabs" and "Firefox ... is synonymous with security"? For me Firefox is also - and actually formost - synonym with tabbed browsing.

My own windows box has IE 7 for the sake of those few sites that really need IE (Windows Update, mainly). Of course I use Mozilla (albeit Seamonkey, not Firefox) for all other browsing on Linux as well as Windows. But recently I had the misfortune of having to intensively use IE 6 for two months "at work". The one thing that I hated most was the absense of tabs, not the lesser security.

Don't get me wrong, the security argument is very valid. But the target audience is going to be much more convinced by the tabs. If not, I suggest putting Lynx on the machines. It's even more leightweight, and it even has more security advantages, since no hacker targets it (anymore) and since features that aren't there can't be abused. Now really...

Re:Why not Lynx?

For even greater security, I use telnet and mentally parse the source.

Bookmarks

I don't understand why bookmarks get the flick, since they've been around since the year dot, and are both 'light' and important. Other than that it sounds like a good idea. Having said that, until two years ago I was using an AMD 400 as my internet PC, it seemed just fine on dialup, using FF 1.5 or so, on W98SE. I don't get the impression that browser speed or footprint is a big deal in itself, most of the processing seems to be content.

Ok I have some old 486's and up....

....To get them web-able would mean I could find a usable home for them....

But I get the impression that what is referred to as old here is system produced 5 -7 years ago.

Hell I'm running off an overclocked to just over 500Mhz box right now using Ubuntu. Its my main internet system. It does just fine.

Having been screwed badly by the computer industry during the commodore fall and its thieving aftermath I haven't found a good enough reason to upgrade to the latest and greatest but rather wait for perfectly good hardware to be tossed out. I'll make smaller purchases in fixing or upgrading some tossed out systems but that's not very often. Getting to be just DVD R/W drives anymore. And that is so I can run live Linux CDs such as Dynebolic.

But this doesn't work for the older systems.

So to me old system fall in the category of 486's to Pentium I, and I have quite a few of those that will either make it into next years Decatur High free electronics recycling mine (yes, electronic based hardware has more mineral value in it than its weight in raw dirt based ore and such... And to think some places want to charge you to recycle) or I'll find an easy way to make them useful again which is the preferred method even with recyclers.

So if the software industry got back to lean and mean OSs and small but very usable internet applications and put together a package that could be test run via CD (or floppy/cd combo for those old system that just can't boot from CD) there could possible be an extension to the usable life of systems that otherwise make it to the landfill or recycling mine.

I'd been hoping that AROS would fit here but unless someone take on dev for old 486 systems, its not going to happen.

Anyone know of any such a package easy to test on old systems (live cd or floppy/cd bootable)?

Lite version not necessary

I am writing this from a 500 Mhz, 384 MB RAM, Windows 2000 PC. It is 7 years old. I run the latest and greatest Opera, IE, Firefox, and Eclipse (w/ many plugins) all simultaneously for web development. I don't experience any problems in doing so. Eclipse takes a while to start up, but hell, it does so on my modern PC at the office as well. Face it, web browsing doesn't require much hardware at all - even with the newest browsers.

Why on old PC's?

Why not on new PC's. Are there people who enjoy having to run bloated software? Make it available and if it ALSO runs on older hardware, that is even better.

When did we stop to program with bounderies in our minds? Memory is not an issue. Drive space is not an issue. So please start programming again with these things in mind.

There already is one , for windows at least

Kmeleon [sourceforge.net]

K-Meleon is an extremely fast, customizable, lightweight web browser for the Win32 (Windows) platform based on the Gecko layout engine (the rendering engine of Mozilla). K-Meleon is free, open source software released under the GNU General Public License.

Tabs heavy? I'd suggest the opposite...

Some years ago I made a set of patches [geocities.com] to the Dillo [dillo.org] browser to support tabs and frames and other such things. I kept logs of the increase in memory use, binary size and other metrics. While the binary size and memory use went up a tiny little bit (several kilobytes) this should be offset against running several open windows or instances of a program. Compared to that using tabs actually saves memory, not to mention hassle when not using a tabbed window manager.

My P-3 550 with W2K and NS7.2 ran fine

Until last year I had a P-3 slot 1 550 (upgraded from a 400-500-550-750 then the 750 died went back to the 550) with 224MB RAM running W2KSP5 and Netscape 7.2 and it ran fine. If I were doing the same thing today I'd just replace NS7.2 with Seamonkey 1.1.3 which is what I have running today on all the other machines. I am typing this on a homemade slot 370 P-3 1200Mhz machine with 512MB RAM and Seamonkey. A little slower to start than new machines but entirely usable.

Seamonkey is a bit lighter and quicker than FF, it handles multiuser profiles a bit better than FF and most of the useful xpi extensions run on it.

Re:mozilla firefox ???

That's what I was thinking. There's a reason an Apple team, lead by a former Gecko developer, went for KHTML for Safari. It's the same reason that Nokia later picked WebKit for their mobile devices. WebKit is much lighter than Gecko, and runs quite happily on devices with 200MHz ARM9 chips and 32MB of RAM (no swap space). If you want a light-weight browser for old PCs, I'd look at WebKit, not Gecko.

Re:GNOME

I know this is flamebait; but for the record the Gnome project DO have a web browser (Epiphany) which uses the Gecko engine, but native Gtk/Gnome widgets. Last time I used it, it was smaller and faster than the full blown Firefox (at least in terms of UI-response); had all the main features you would expect from a web browser and integrated with the Gnome desktop quite nicely. -- Sam