RISC OS London Show 2014
David Feugey (2125) 2709 posts |
I’m :) |
Steve Pampling (1551) 8155 posts |
Delete “high-spec” and you’re even closer to the target. AA? how long do you want it to last? I think AAA is still OTT.
“We” is not everyone. Have a check around newsgroups and see how many posts speak repeatedly about RO6 etc. It doesn’t run on modern hardware. |
Chris Hall (132) 3554 posts |
Is there a website for impressionX? Can non-attendees buy a copy yet? Presumably non attendees can buy a copy, but not from either of the two links above (Computer Concepts and X-Ample). Have you registered interest with Richard? He did invite offers of interest at the South West show. “To join the progress/release mailing list send a “subscribe” e-mail containing: you can also send general enquiries to that address" The price of £50 only applies until the end of October, from 1 November the price will be £60. |
Steffen Huber (91) 1949 posts |
With respect to the Risc PC I also do not understand the statement. The A9home, the IYONIX, the Raspberry Pi, the BeagleBoard, the PandaBoard – all easily surpass the Risc PC in nearly every way. Exceptions: PC card, number of extension cards supported, 26bit capability. Obviously, an i.MX6 board also fails here. With respect to the IYONIX pc, the statement however is 100% true. It had Gigabit ethernet and UDMA ATA support. None of the “successors” came close. So an i.MX6 system with a polished RISC OS port is certainly a big step forward. On the other hand, an OMAP5-based board is available for some time now – and does everything a lot better than the i.MX6. Higher clock speeds, faster core, USB3. So an OMAP5-based system with a polished RISC OS port would be another big step forward. |
Chris Hall (132) 3554 posts |
I would be cautious about the OMAP5. The OMAP family has lost support from phone type things – thus may be a dead end. Also the OMAP5 board is heavily subsidised by TI and so not necessarily a reliable commercial platform. All just my opinion though. |
Steffen Huber (91) 1949 posts |
It is true that TI has basically said “no more OMAP after OMAP5” and will change their focus from mobile platforms to automotive platforms (which may still be suitable as a RISC OS platform). However, the OMAP5 is available in quantities and is likely available for a very long time. So do we care if OMAP is a dead end? After all, nobody knows if Freescale will ever develop their i.MX range further.
Source? The price of the IGEPv5 is certainly high enough to cover the cost. |
Chris Hall (132) 3554 posts |
The TI OMAP3 and OMAP4 boards certainly said they were subsidised by TI so I assumed OMAP5 was as well. |
Rick Murray (539) 13806 posts |
I can probably rustle up an AA holder simpler than anything else1. Longevity not really an issue. It working, more an issue. ;-)
No number of new “awesome” boards is going to change that. Seriously, we could clone Jeffrey and get him to make versions of RISC OS for every single ARM board that exists. The only way this will happen is if the ROLtd source is merged into the ROOL source. Then the stuff people want that RO5 doesn’t do, it will. And those stuck with ancient hardware can move to something more modern. Everybody wins, right? <sigh>
PC card won’t work, but DOSbox can run a basic Windows 3.11 setup. It could possibly put in a reasonable attempt at running Word or such, but with no SHARE support in DOSbox, the software bombs before it even starts. Extension cards? Most stuff is USB these days. We lack drivers. 26 bit? Hardware issue, but there’s always Aemulor. Note that RO5 on a RiscPC is 32bit too.
I didn’t know it had Gigabit. Question is, though, is how fast did it actually run when running flat out? I’m not sure that spec would have been up to gigabit. Better than 100mbit? Yes. Gigabit? Mmmm… ATA is unlikely to appear on modern SoCs. Way too many I/O pins when SATA and USB3 can outperform with lesser hardware requirements. 1 If nothing else comes to mind, I think there’s one in the A310. |
Raik (463) 2059 posts |
Mmm, and PiWin? … Ok, very slow but faster on PandaES. |
Tennant Stuart (2505) 122 posts |
My friends & I really enjoyed the show, and well done to all concerned. However…
..which was very nice, though the mouse pointer disappeared 60% across the the width. It was being used to demonstrate the Open Street Map project, and yet this site rejects Netsurf with the error message… You are either using a browser that does not support JavaScript, or you have disabled JavaScript. OpenStreetMap uses JavaScript for its slippy map. |
Tennant Stuart (2505) 122 posts |
Somebody needs to tell Google Translate… Today is Sunday, 26th October. Aujourd’hui est dimanche, 26th Octobre. Oggi è Domenica 26 ottobre. |
Bryan Hogan (339) 589 posts |
Lucky?!? Do you think we just throw this show together with no planning? Don’t answer that :-) I have here the orginal stand layout, and you were at opposite ends of the hall.
Yes, that’s the point, if you have RiscOSM you don’t need to use the website. Plus you can export the maps as Draw files, so much more flexible than the bitmaps you get from the website. |
Rick Murray (539) 13806 posts |
You’ll need to come up with a more authoritative source than that… Try:
Also:
(^_^) |
Steve Pampling (1551) 8155 posts |
You might want to check the voltage limits on the RTC power so that it continues working.
True, unless your awesome board had a second processor slot that that took a 26 bit processor card. A sort of reverse PC card, running RO6 in a window or full screen. Even then the Castleness would upset some.1
“Castleness”
As above, RPC style second processor slot. Keep a few retro people happy.
Yes, with isolation good enough to stop a lightning strike doing the board any damage beyond the isolation. Question is, though, is how fast did it actually run when running flat out? I’m not sure that spec would have been up to gigabit. Better than 100mbit? Yes. Gigabit? Mmmm… Hard to tell anyway – gig network devices at home were not that common.
Lot’s of extremely cheap domestic items at £1 shops (Euro shops over there?) have fairly useless LED light fitments running from a pair of AA’s in a holder… 1 Yes, that’s a small dig at people that haven’t yet smelled the coffee |
Colin Ferris (399) 1809 posts |
What was driving the Large monitor at 3840×2160 resolution ie: What no of colours/Hz etc? |
Steve Pampling (1551) 8155 posts |
When you think about it, the card doesn’t currently work under RO5 irrespective of the new hardware having no capability of taking the elderly Gemini card. |
Chris Gransden (337) 1202 posts |
It was connected to a Raspberry Pi. The refresh rate was only 14Hz. |
Steve Fryatt (216) 2103 posts |
No it wasn’t: it was being used to demonstrate RiscOSM, which is a completely different entity. RiscOSM uses data from the Open Street Map Project, but is (AIUI) otherwise unconnected. and yet this site rejects Netsurf with the error message… Works fine in Firefox. You could always roll up your sleeves and help the NetSurf team work on Javascript support… :-) |
David Feugey (2125) 2709 posts |
I hope that FastDosBox will be updated soon for all computers (Dynarec works [worked?] only on the Pi). With Dynarec engine for all and a few patches (to use network and Win98), we could have something almost as fast as the fastest PC card. Just ask the author. I’m sure he is (almost) ready to make the needed changes. And I’m ready to pay for that :) http://gaming.capsule-sa.co.za/?gamepress_reviews=fastdosbox-for-raspberry-pi-raspbian-and-risc-os With a 128 MB patch for FastDosBox (and Win98 support), it could be possible to launch OpenOffice 2.4 and Office XP on RISC OS.
IMHO, there is no need. A tool like ArchiEmu is today very fast (about 40 MHz on a PandaBoard ES). I hope that Jan de Boer will be able to finish the fullscreen part, and then to go on something new : ARM7 emulation. A complete RISC PC emulation with a speed of – for example – 60 MHz, would be perfect for almost every needs. Perhaps that a full RISC PC emulation version of Aemulor could be OK too, or a port of RPCEmu. Once more, I would pay for that. And more than just a license. Today, we have enough power to emulate almost all the all setups. The very very good idea, would be to use the other cores for the emulation. One for ARM7 core, one for x86 core, one for video decompression, and no impact on the OS. ‘Just’ need to set up an AMP mode (reserve memory + launch a cut down version of RISC OS on the other cores) with a few API. You can put a bounty on that, I’ll bet a lot of money (really). Misc |
David Boddie (1934) 222 posts |
Steffen, Chris: I don’t know if TI have the same sort of statement about availability of parts, but Freescale has a Product Longevity page. Whether it’s easy or economic to get parts for 10 years is another matter, of course. |
Tennant Stuart (2505) 122 posts |
Please read again, I was criticising Google Translate. |
Steve Pampling (1551) 8155 posts |
Porting RPCEmu for use on multicore devices would be interesting – porting for use on single core being a first step. After all with an emulator for the old non-upgraded software and the host platform for the new software why would you need a real RPC (or older) except as a curiosity.
The current owner of RO6 rights is the same as the owner/author of VRPC.1
What is the point of having a commercial emulator for an old system that won’t run “old”2 software when a free emulator that does the same job? After all VRPC and RO4.3x (RO6) does run that old software. Be honest, don’t you think running VRPC in 32 bit mode diminishes its market? 1 If he wanted to modify VRPC to use RO5 I suspect the modified version would already exist. I wouldn’t hold my breath waiting for that outcome. 2 from the era of the real device. |
David Feugey (2125) 2709 posts |
That’s my opinion too.
Better speed, and good support of old computers (would be useful for me, as my Atom miniPC doesn’t like RPCEmu. RPCemu developers keep saying that there is no problem). And anyway, that’s a market opportunity for VRPC. Basically, we just need a 32bit version of hostfs. Give that to me, and I buy a licence tomorrow… But in fact, I was suggesting another thing: a version of VRPC for RISC OS 5. Of course, there is zero chance to see that one day :) |
Tennant Stuart (2505) 122 posts |
Really? Is there a link to where I can download a working version of Firefox? |
Steve Pampling (1551) 8155 posts |
I know. The point I was making was that VRPC only remains in existence as an available product because it runs 26 bit programs. If it was restricted to 32 bit only (RO5) then it is largely useless as genuine hardware exists to outpace the emulator. Diminish the market like that and the product disappears. Why would Aaron kill the remaining market? |