Firefox port is '95 percent complete'

By Chris Williams. Published: 15th Apr 2005, 13:20:11 | Permalink | Printable

Opinions of port's public demonstration

On Monday night, Peter Naulls revealed the progress of his Firefox web browser RISC OS port to a Cambridge user group. Punters from as far away as Northampton and Surrey attended the talk, in which Peter assured his audience that a ported Firefox would be publicly released sometime in 2005. It was understood that Peter is about 95% of the way to producing an initial beta version release.

The evening began with a run down of web browsers released for RISC OS, from WebsterXL to Browse to the Oreganos, to illustrate the web technologies currently available to RISC OS users. After demonstrating his Dillo port, a simple but capable web browser for X Windows, Peter moved onto unveiling his work on Firefox. According to those present, the browser took around 20 seconds to start up and appear on the iconbar, on a StrongARM RiscPC. It was understood that if Firefox can run at an acceptable pace on a 200MHz RiscPC, it should certainly be usable on a 600MHz XScale powered Iyonix. According to Peter, users will need 64M of free RAM to use Firefox.

Due to the lack of an Internet connection at the meeting, Peter used a WebJames web server to serve local copies of example websites to the Firefox web browser.

"Rendering speed of pages was quite reasonable, considering that it was a Risc PC, but the user interface seemed a little sluggish," said James Byrne, who attended the meeting (now back from Japan).

"Peter explained that there was still a lot of work to be done improving the libraries used for the RISC OS port, and that he expected performance to improve quite a bit before the browser was released."

Another attendee, Dr. Dave Lane felt the page rendering speed was "slow". He commented, "Much of the user interface of the Firefox window was 'Microsoft style' with drop down menus along the bar at the top of the window." According to Dave, Peter will initially concentrate on fixing the bugs in the port and increasing the application's performance, before making the user interface more RISC OS consistent.

Peter also faced questions about web browser plugins, such as Flash and Real media players. The simple answer was that it's not the web browser's job to provide such functionality, although a Flash port (from the open source client) already exists and a Real media one could in theory be produced once an NDA with Real is signed.

Dave continued: "I asked about Javascript and Peter replied that an up to date version of Javascript was not a problem as it came with the porting of Firefox. I thought this was a real plus point over our current RISC OS browsers. Peter also mentioned Kaffe, his JVM port for RISC OS, which he would continue to work on after producing a usable Firefox. There were several technical questions about APIs and threading which Peter answered.

"As far as I could tell, the web pages [rendered by Firefox] were presented correctly. Judging by some of the questions asked, there were several technical experts present, presumably programmers, and they seemed satisfied with Peter's answers. My only further thought is that I really hope Peter succeeds, as Firefox would be a major step forwards for our platform."

James added: "There is clearly plenty of work still to be done before Firefox is ready for general use, but getting it to its current state is a major achievement. This is a significant project which, when released, will fill an important gap in RISC OS applications. It will give the platform an up-to-date browser, and one that can be kept up-to-date so we don't fall behind again."

Of course, we can't forget the 80:20 rule of projects.

Links

Firefox screenshots
Firefox port announced Peter Naulls is a drobe.co.uk editor

Previous: Writing internationalised software
Next: Desktop coding book republished online

Discussion

Viewing threaded comments | View comments unthreaded, listed by date | Skip to the end

Aah, how wonderful to hear about this, especially since it already runs at an acceptable speed on a SARPC and Peter aims to improve the libraries used by this port. I'm happy - I'll really be smiling when I see it running ...

Peter is really doing a superb job, I can't wait to thank him in person.

hEgelia on 15/4/05 1:37PM
[ Reply | Permalink | Report ]

Stretching the pointless questions here a little, any idea if skins can work :@P. Nice to hear your still going Peter.

I have been using Firefox under windows XP for some time, and have found it to be the best.

Working with the ability to actually render pages on and SA in an achievement, as firefox is a power hungry little bugger.

Not supprised about the memory consumption....whould would like to browse with anythign less than 64 Megs? :@) I wouldn't!

em2ac on 15/4/05 2:25PM
[ Reply | Permalink | Report ]

Looks like this computer will need another memory upgrade to run Firefox :| I've only got 64Mb free at the moment, and that is normal for this computer.

AndrewDuffell on 15/4/05 3:20PM
[ Reply | Permalink | Report ]

It sounds like Dave Lane was confused between java and javascript. Having the latest javascript support will make little or no noticable difference over, say, Oregano2. Hence it can hardly be a plus point. The benefit only comes hand in hand with better DOM support, which of course Firefox has.

senduran on 15/4/05 5:45PM
[ Reply | Permalink | Report ]

senduran> so are you saying in conclusion that Firefox will have a noticable difference due to the fact it has better DOM support?

jonix on 15/4/05 6:29PM
[ Reply | Permalink | Report ]

"the browser took around 20 seconds to start up and appear on the iconbar"

How long does it take on similar hardware running Linux?

(I'd try it myself, but my NET100 card doesn't seem to be supported by Linux.)

Stoppers on 15/4/05 7:14PM
[ Reply | Permalink | Report ]

Stoppers:

even on a modern x86 machine firefox can take a little while to load first time - no idea why

senduran:

firefox's javascript support is *far* better than any of the oreganos so far...

ROHC on 15/4/05 7:19PM
[ Reply | Permalink | Report ]

ROHC: As firefox's Javascript library and Oregano's Javascript library are one and the same (and have been since Oregano 1) the Javascript support is the same. Senduran is quite correct in the fact that it is the DOM support that is better in Firefox than it is in Oregano 1/2.

_Jez_ on 15/4/05 8:07PM
[ Reply | Permalink | Report ]

Dave was slightly confused about Javascript, but I'm sure he understands the difference from Java.

The point about Javascript in Mozilla/Firefox is that it's the reference implementation - as Jez has pointed out, the same engine is used in a number of browsers. But that's only part of the JS story, and why it's not enough to just stick it in NetSurf. The other part is tying it to browser behaviour which is very involved (and includes DOM). Again, Mozilla should (bugs not withstanding) implement the full JS requirements.

And just to confuse the issue, there are differences in MS's implementation in IE.

This is just one of a huge list of reasons which makes writing a browser so hard.

mrchocky on 15/4/05 8:33PM
[ Reply | Permalink | Report ]

jonix: Yes, the DOM support is what will make the noticable difference. The difference, ultimately, being that with DOM level 1/2 more sites will work correctly, or at all. As a user that's what you're really interested in I think, and so it's perhaps worth noting that even having the latest javascript and DOM support (Oregano3?) doesn't guarantee you get into every site - there are many badly written ones. For that reason it's easy to imagine that Firefox could have an advantage because more users have submitted more bug reports and more broken sites have been made to work. That may or may not be true however.

_Jez_: Surely the library got updated between the release of Oregano 1/ 2? So the code underlying the Oregano's is indeed out of date? Just making an assumption however.

senduran on 15/4/05 8:37PM
[ Reply | Permalink | Report ]

My RO4 RAM is currently only 41, a bit more to go to get to 64mb.

Once I have at least 64Mb of RAM, would it make much difference to Firefox if there was a bit more than that?

How much RAM can be fitted to RO4 RiscPC600 machine?

Sawadee on 15/4/05 9:05PM
[ Reply | Permalink | Report ]

sawadee:

RPCs can take a maximum of 256MB, in two SIMMs (maximum 128MB each); 128MB SIMMs are relatively hard to come by - and also tend to be bulky. 64s are much more common and (usually) smaller, so one of each is often the best bet - and will give you plenty of memory to use firefox and every other application you can think of simultaneously

chrisj on 15/4/05 9:59PM
[ Reply | Permalink | Report ]

I just hope Firefox will not work too well on RiscPCs so people can be persuaded to upgrade to newer computers either Iyonix or Omega

Gulli on 16/4/05 12:47AM
[ Reply | Permalink | Report ]

I hope firefox will work about 1% faster as for example Oregano-2

EasyKees on 16/4/05 1:05AM
[ Reply | Permalink | Report ]

In reply to chrisj:

Thanks for that info.

I recall they go up to 128, but didn't know that you could get two 128Mb's for the two slots.

Maybe a couple of 64s may do the trick to run Firefox?

I'm quite happy just to have and see Firefox running when it is released.

The only features I miss in Oregano2 that I use, is the lack of Javascript on certain links in my business sites that I cannot activate, and the ability to print "selected" text from web pages.

Sawadee on 16/4/05 1:28AM
[ Reply | Permalink | Report ]

chrisj:

128Mb hard to come by ?

No they're not. You obviously haven't visited the APDL web site.

However if people don't actually need more than 128Mb I would normally advise 2 x 64Mb rather than 1 x 128Mb as it's cheaper.

apdl on 16/4/05 7:09AM
[ Reply | Permalink | Report ]

If you are living in the netherlands you can buy new 64Mb's for 18 euro's

look at [link]

(you must have a PC for that website rage

none of the 6 browsers for RISC-OS will display it correctly --- perhaps firefox can do it----

)

EasyKees on 16/4/05 9:22AM
[ Reply | Permalink | Report ]

Gulli:

"I just hope Firefox will not work too well on RiscPCs so people can be persuaded to upgrade to newer computers either Iyonix or Omega"

Heh, buying a new computer to run Firefox? Hmmm. There are a lot of options there. -- Spriteman

Spriteman on 17/4/05 2:14PM
[ Reply | Permalink | Report ]

Spriteman:

Very true, but people that have already decided to stick with RISC OS might be persuaded to upgrade their RISC OS computers.

Gulli on 17/4/05 4:51PM
[ Reply | Permalink | Report ]

Is that 64M *free* an absolute minimum, or would a 64M machine running nothing else be viable?

jess on 17/4/05 7:55PM
[ Reply | Permalink | Report ]

The biggest weakness for running Firefox on a RPC would probably be the CPU more than quantity of RAM. Firefox is horribly CPU intensive. Even the Iyonix's CPU is less than ideal for Firefox, sadly. But 256mb of RAM would be nicer than 40mb, though how much RAM you'll need to run it comfortably probably depends on whether you'd typically browse one site at once or have a heap of tabs open.

Clades on 18/4/05 4:18AM
[ Reply | Permalink | Report ]

Clades: It's already been said that Firefox runs at an "acceptable speed" on a RiscPC, so although I doubt it's going to fly on any RISC OS machine there's no reason to raise the speed issue.

SimonC on 18/4/05 10:09AM
[ Reply | Permalink | Report ]

It has been said, but it wasn't mentioned in the article, and perhaps said in a moment of wishful thinking by people replying to the article in these comments.

Certainly the version I demonstrated at the talk wasn't usable in any practical sense on a RiscPC - or indeed an Iyonix. Part of this had to do with attempting to render several thousand 2x2 pixel sprites, and other minor things that are currently being improved.

I have performed miracles on speeding up programs in the past, and there are quite a number of things I still plan to do which will certainly help. On the other hand, we are dealing with what really is quite old hardware in terms of RiscPCs. There's only so much than can be done, and Firefox does a great deal more than any other RISC OS browser.

What will be noticeable is that FF's method of rendering is very different to that of most RO browsers - sometimes this means pages can display much faster - especially those with complex tables.

I recommend you see it for yourself - whether that's at Wakefield or elsewhere before making any predictions on speed. Guessing isn't in anyone's interest.

mrchocky on 18/4/05 10:37AM
[ Reply | Permalink | Report ]

"It was understood that if (emphasis added by me) Firefox can run at an acceptable pace on a 200MHz RiscPC, it should certainly be useable on a 600MHz XScale powered Iyonix."

So... let's keep our fingers crossed (but not too tightly)

"Rendering speed of pages was quite reasonable, considering that it was a Risc PC, but the user interface seemed a little sluggish." was the opinion of James Byrne. Remember it was demonstrated from local copies of websites, not from online servers. It was a bit over-enthusiastic and foolish of me to state at the first comment that "...it already runs at an acceptable speed on a SARPC..." and I hope it didn't create any misunderstandings. However, there seems to be still a substantial way to go at optimising Firefox for RISC OS machines and we should be grateful for that.

hEgelia on 18/4/05 12:13PM
[ Reply | Permalink | Report ]

Ah, I'd missed the "if".

SimonC on 18/4/05 1:07PM
[ Reply | Permalink | Report ]

mrchocky: "sometimes this means pages can display much faster - especially those with complex tables" Hmmm but Firefox (or rather Mozilla) doesn't handle large tables very well. (slow and sometimes even wrong) I have a old case open in bugzilla about that (233532) but it isn't fixed yet

. If you get near that code please fix it

Jaco on 18/4/05 7:04PM
[ Reply | Permalink | Report ]

Please login before posting a comment. Use the form on the right to do so or create a free account.

Search the archives

Today's featured article

South West 2005 show report
Coverage of Castle, Qercus, UPP and ROL theatres

42 comments, latest by jc on 03/03/05 11:12AM. Published: 28 Feb 2005

Random article

Castle directors patch up 'disagreement'
Row over future resolved with open source plans

12 comments, latest by jc on 28/11/06 2:05PM. Published: 24 Nov 2006

Useful links

News and media:
Iconbar • MyRISCOS • ArcSite • RISCOScode • ANS • C.S.A.Announce • Archive • Qercus • RiscWorld • Drag'n'Drop • GAG-News

Top developers:
RISCOS Ltd • RISC OS Open • MW Software • R-Comp • Advantage Six • VirtualAcorn

Dealers:
CJE Micros • APDL • Castle • a4 • X-Ample • Liquid Silicon • Webmonster

Usergroups:
WROCC • RONE • NKACC • IRUG • SASAUG • ROUGOL • RONWUG • MUG • WAUG • GAG • RISCOS.be

Useful:
RISCOS.org.uk • RISCOS.org • RISCOS.info • Filebase • Chris Why's Acorn/RISC OS collection • NetSurf

Non-RISC OS:
The Register • The Inquirer • Apple Insider • BBC News • Sky News • Google News • xkcd • diodesign

© 1999-2009 The Drobe Team. Some rights reserved, click here for more information
Powered by MiniDrobeCMS, based on J4U | Statistics
"Oh, and making up stories and quoting private emails out of context isn't damaging then?"
Page generated in 0.2115 seconds.