Iyonix-only Firefox 2 port released online | RISC OS news and information from Drobe


Beta! \| About us \| Contact \| Submit news \| RSS \| Twitter	Webspace \| Tech docs \| Downloads \| BBC Micro \| Gallery \| Wallpaper

Iyonix-only Firefox 2 port released online

Published: 2nd Dec 2006, 10:15:09.

Christmas day present hinted

The Iyonix-only version 2 of the RISC OS Firefox port was released this weekend. Developer Peter Naulls made the open source web browser available for free download after generous punters donated a grand in cash to the project in little over a fortnight. A second release could be posted online in time for Christmas day if Peter secures more funding.

Version 2 has no RISC OS-style menus, the iconbar icon has been removed, and the while browser does support secure 'https' websites, users will have to manually verify the identity of certain sites. Peter also hopes to use the NetSurf Unicode engine in a future release to display non-Latin characters, and he said version 2 is faster than the previous version 1.5 port although more can be done to speed up the software.

On the Firefox 2 port website, Peter wrote: "This version was released after reaching £1000 in donations. Thanks to all who contributed.

"In order to be able to justify further development - in particular, full RISC OS integration and other issues mentioned in the documentation - I am after a further £4000. I am still some way from this, but I expect to make the next release on Christmas day if it is reached."

Speculation
Last month when Peter announced his intention to release an Iyonix-only RISC OS Firefox 2, it was speculated that the software will be Iyonix-only because the executable would be too large to fit inside the RISC OS 4 28MB wimpslot. The binary released this weekend is 19.1MB in size, dispelling that theory.

However, following discussions on the GCCSDK list, we believe Peter has configured GCC to produce code that uses ARMv5 instructions when building the Firefox 2 port. This machine code can be executed by the Iyonix's Intel XScale IOP321 processor, but not the RiscPC's StrongARM and the A9home's Samsung ARM9 chips. The ARMv5 instructions could squeeze extra performance out of the mammoth web browser.

Peter has so far refused to comment on the mandatory requirement of an Iyonix.

An A9home release is on the cards if there is sufficient "financial interest", although it's understood no version for RiscPC-class machines will be released. AdvantageSix and CJE Micros, who produce and sell the A9home, APDL, who sells RiscPC upgrades, and VirtualAcorn, who produce and sell VirtualRiscPC, sponsored the Firefox 1.5 project.

Update at 17:35 2/12/2006
The Firefox 2 port binary does include CLZ instructions which are ARMv5; the XScale in the Castle Iyonix can execute these, unlike RiscPC StrongARMs and Samsung ARM920Ts in A9homes.

Links
Firefox 2 website

Discussion

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

Great!

hzn on 2/12/06 12:33PM [ Reply | Permalink | Report ]

Any other 4.39 users who'd bei interested?

AW on 2/12/06 1:28PM [ Reply | Permalink | Report ]

version 2.0 works a lot better than 1.5 - it is not as slow and typing into writeable fields is improved. I can now see the vertical cursor, which I never could with 1.5.

Given that 100 people pledged 100 gbp to buy Select it should be possible for 100 people to fund Firefox - even if its a case of robbing Paul to pay Peter :-)

This is being typed in using Firefox.

JohnR on 2/12/06 2:07PM [ Reply | Permalink | Report ]

Excellent news, once again Peter delivers on his word.

I can only suggest Iyonix and A9home users who haven't contributed yet to do so. If you were holding out for Oregano 3, is it unwise to say this first release of Firefox 2 will only delay it further? Frankly I doubt if it will ever be released to the public. When was Oregano 3 announced? It seems like years ago... Indeed, Firefox 2 is here now.

I believe it's quite possible that Firefox 2 on RO can eventually enjoy an equally satisfactory level of RO integration as Oregano 3 (supposedly) has, and perhaps even a better quality of website rendering. This is one of those projects to support if you're depending on your RISC OS machine to remain capable in this day and age.

hEgelia on 2/12/06 4:09PM [ Reply | Permalink | Report ]

JohnR Yes the speed increase is welcome here. I didn't have a problem with the speed of rendering pages but the writeable fields speeds was irritating. Now FF2 is not worse than O2 in that respect actually seems a little faster.

Agreed FF2 integration is probably money better spent than Select for Iyonix.

blahsnr on 2/12/06 5:18PM [ Reply | Permalink | Report ]

Peter Naulls is to be congratulated in seeing things through with FireFox 2. When Peter first introduced ChoX11 back in June 2003 I said it would take him years before he had a real application working and I make that 2 1/2 years. He said at the time I did not know what I was talking about, but there you go thats Peter for you.

Jwoody on 2/12/06 8:07PM [ Reply | Permalink | Report ]

Well FireFox2 does contain a number of ARMv5 instructions which rules out it running on a Risc PC. Incdicentally ARMalayser required a 200MB wimpslot to analyse the 20MB firefox2 binary, and produced 340MB of dissassembly.

druck on 2/12/06 10:06PM [ Reply | Permalink | Report ]

Once again Peter delivers the goods. The ability to port and application like Firefox to RISC OS is testament to the technology Peter has developed and the work of the people behind GCC for RISC OS. A truly massive amount of work has been put in over the years to get to this point.

knutson on 2/12/06 11:06PM [ Reply | Permalink | Report ]

It is good to see there is a new better/faster port of Firefox for Iyonix users (those users are willing to pay for such a port, because Peter has uploaded the first version) I only can hope there will be enough A9home users to pay for a port for their machine. I will certainly donate for FF on A9Home

EasyKees on 2/12/06 11:35PM [ Reply | Permalink | Report ]

JWoody: thanks for your concern, but I think you'll find that ChoX11 was running "real" applications at least 2 years ago. It's just that most of those weren't very interesting to RISC OS users. For example, there was a game ported by Alan Buckley, the name of which escapes me presently. Even if that weren't true, the changes to ChoX11 in the last year have been very minimal - just minor bug fixes. Indeed, Firefox 1.5 was released more than a year ago.

mrchocky on 3/12/06 12:46AM [ Reply | Permalink | Report ]

Ah yes, looking back through drobe comments, you've made a number of confused and incorrect comments about porting and ChoX11, and seem to have some kind of unspecified beef about it. Why, I fear we will hear in too much detail shortly.

mrchocky on 3/12/06 10:25AM [ Reply | Permalink | Report ]

First of all, well done Peter N.. I have to use the middle mouse button when clicking on a link to open in a new tab rather than right clicking on the link to use the menu option which doesn't work here, but did with the previous FF1.5 versions. The icon for FF in the filer directory window is now that nondescript letter A made up of cog wheels. Are these other folks' experiences? In answer to Druck's post yesterday: Does Druck know that the police can arrest you for dissassembly especially if it's ARMed dissassembly. Ask Ian Chamberlain or any of us other RISC OS demonstrators. I am glad that bookmarks now works. Well done Peter.

DaveLane on 3/12/06 11:25AM [ Reply | Permalink | Report ]

Mr Chocky

No beef here. Just I thought FF1.5 was a beta and not complete in a lot of peoples eyes, so I stand by my statement of 2 1/2 years from ChoX11 to first significant real usable application FF 2.0

Jwoody on 3/12/06 11:28AM [ Reply | Permalink | Report ]

In reply to Jwoody: You appear to be forgetting Dillo. A webrowser port available from riscos.info. This uses the ChoX11 library and is a full application.

sa110 on 3/12/06 11:44AM [ Reply | Permalink | Report ]

And dillo is a fast little browser,that can be handy at times

highlandcattle on 3/12/06 12:32PM [ Reply | Permalink | Report ]

Jwoody: you can stand by it all you want, but it's clearly wrong for reasons I already stated. I'm glad you feel it's ok to dismiss those reasons out of hand. Meanwhile, the incompleteness of any version has little to do with ChoX11, which just provides raw X11 functionality and no more.

mrchocky on 3/12/06 3:48PM [ Reply | Permalink | Report ]

Basically Jwoody is trying to say is that it's been 2 1/2 years from when ChoX11 was released until FF2 was released. Which, to be fair, is probably correct.

Personally I'm impressed that people still have the inclination to produce these things and I can't belive that anyone would want to see anything other than positives.

I think the money Peter has collected speaks louder than any half baked criticism, but I thought I'd still pitch in as someone who doesn't really use RISC OS any more, but still an interest in the platform.

Keep it up!

john on 3/12/06 5:16PM [ Reply | Permalink | Report ]

I don't have an iyonix, and till i graduate, probably wont.

But fantastic news Peter!

Keep it up :@)

em2ac on 3/12/06 8:39PM [ Reply | Permalink | Report ]

For me this is a vast improvement on the 1.5 port. Smoother, faster and more stable. I have put it on my backdrop, meaning that it now qualifies as a browser I will make serious use of for sites that O2 or Netsurf can't handle. The days of being forced to use Linux or a PC for web browsing on some sites might be coming to an end. Well done Peter!

mrtd on 4/12/06 9:23AM [ Reply | Permalink | Report ]

Odd, I find it runs at the same speed as 1.5. If there is a difference it's not noticeable to me.

SimonC on 4/12/06 10:54AM [ Reply | Permalink | Report ]

A fantastic job by Peter, but unfortunately proving very much less stable than 1.5 on my machine, and I don't know why. Have reverted back to 1.5 for the time being.

DaveW on 6/12/06 7:15PM [ Reply | Permalink | Report ]

Just to clear up any confusion, Firefox2 does not need 340MB to run, thats the size of the dissassembly output, where each 4 byte instruction or data word in the 20MB executable can generate up to 160 characters of text, consisting of address, word value, assembler menemonic and with annotation and performance analysis comments.

druck on 9/12/06 11:47AM [ Reply | Permalink | Report ]

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

Create a new account
Forgot your password?

Search this website

This week's poll

Recent developments have left me feeling...

Assured ROS will appear on new hardware		55%
Pleased OS desktop features are being developed		10%
ROL and ROOL will eventually kiss and make up		5%
App developments are critical		10%
Dave Holden sleeps easy at night		19%

Discuss this | Archives