| |
RISC OS on The Register |
| |
bhtooefr (21:29 25/6/2010)
Phlamethrower (21:38 25/6/2010)
mac9 (11:46 27/6/2010)
bhtooefr (15:19 27/6/2010)
VincceH (22:28 25/6/2010)
tribbles (23:02 25/6/2010)
bhtooefr (11:24 26/6/2010)
VincceH (13:14 26/6/2010)
swirlythingy (09:38 26/6/2010)
VincceH (10:51 26/6/2010)
trevj (12:22 28/7/2010)
trevj (15:35 29/11/2013)
|
| |
|
Eric Rucker |
Message #114703, posted by bhtooefr at 21:29, 25/6/2010 |
Member
Posts: 337
|
It actually looks to me like Peter's talking about adding memory protection, not PMT... which is already there, just very weak between apps and the kernel, IIRC. |
| |
 [ Log in to reply ] |
| |
|
Jeffrey Lee |
Message #114704, posted by Phlamethrower at 21:38, 25/6/2010, in reply to message #114703 |
 Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot stuff
Posts: 15100
|
Nope, it's definitely PMT he's talking about!
And I can't see that it would be impossible to add interrupt driven multi-tasking to RISC OS. It's a programmed interrupt timer that will effectively call the same context switch code in the OS as is used in the co-operative system. May have to move some of the context save function into the interrupt handler, but that should be trivial. (OK, so he's actually saying that the act of moving some code into an interrupt handler is trivial, not the act of adding PMT as a whole - but I'll leave it as an exercise to the reader to work out why doing what he suggests won't leave us with a working PMT system) |
| |
[ Log in to reply ] |
| |
|
VinceH |
Message #114705, posted by VincceH at 22:28, 25/6/2010, in reply to message #114703 |
 Lowering the tone since the dawn of time
Posts: 1600
|
I'm slightly more amused by what the first commenter said:
This is all well and good, but when is someone going to get me a port of RISC OS for my iPad? After all, it's not so different to the Archimedes A420 I have kicking around somewhere...
Zarch/Lander would be fantastic with the iPad's accelerometers! It's not the concept of a port to the iPad that makes me laugh, it's the idea of a port to ANYTHING with the end result of being able to play Zarch/Lander on it*.
Far simpler would be for a publisher to write a Zarch/Lander/Virus-alike for the platform.
* He's right, though, the accelerometer would make the device a great platform for playing a game like that. (ditto my HTC Touch Diamond - but the iPad obviously has a bigger display).
[Edited by VincceH at 23:28, 25/6/2010] |
| |
[ Log in to reply ] |
| |
|
Jason Tribbeck |
Message #114706, posted by tribbles at 23:02, 25/6/2010, in reply to message #114705 |
 Captain Helix
Posts: 929
|
Damn - you've just nicked my idea  |
| |
[ Log in to reply ] |
| |
|
Martin Bazley |
Message #114707, posted by swirlythingy at 09:38, 26/6/2010, in reply to message #114703 |
Posts: 460
|
It's so sad that we're so desperate for attention that the moment an even remotely 'mainstream' news source (i.e. not RISC OS-oriented) breathes the merest mention of us the miracle is instantly reported by many different people, on comp.sys.acorn.misc, riscos.info, the Icon Bar forums, the Icon Bar news section, the RISC OS Open forums and the RISC OS Open news section. The only people not joining the party (as ever  ) seem to be My RISC OS!
I mean... come on... we're not that obscure, are we?  |
| |
[ Log in to reply ] |
| |
|
VinceH |
Message #114708, posted by VincceH at 10:51, 26/6/2010, in reply to message #114707 |
Lowering the tone since the dawn of time
Posts: 1600
|
I mean... come on... we're not that obscure, are we? Yes. Next question. |
| |
[ Log in to reply ] |
| |
|
Eric Rucker |
Message #114709, posted by bhtooefr at 11:24, 26/6/2010, in reply to message #114705 |
Member
Posts: 337
|
Nope, it's definitely PMT he's talking about!
And I can't see that it would be impossible to add interrupt driven multi-tasking to RISC OS. It's a programmed interrupt timer that will effectively call the same context switch code in the OS as is used in the co-operative system. May have to move some of the context save function into the interrupt handler, but that should be trivial. (OK, so he's actually saying that the act of moving some code into an interrupt handler is trivial, not the act of adding PMT as a whole - but I'll leave it as an exercise to the reader to work out why doing what he suggests won't leave us with a working PMT system) Fair enough, I missed that (not sure how,) although the rest of the post was talking about memory protection, it seemed.
* He's right, though, the accelerometer would make the device a great platform for playing a game like that. (ditto my HTC Touch Diamond - but the iPad obviously has a bigger display). But nobody develops for Windows Mobile any more, so your only hope is for Android to become stable on that hardware. (I've got a Touch Pro RAPH800 - CDMA version, so I'm even worse off than the RAPH100 and DIAM100 ports.) |
| |
[ Log in to reply ] |
| |
|
VinceH |
Message #114710, posted by VincceH at 13:14, 26/6/2010, in reply to message #114709 |
Lowering the tone since the dawn of time
Posts: 1600
|
* He's right, though, the accelerometer would make the device a great platform for playing a game like that. (ditto my HTC Touch Diamond - but the iPad obviously has a bigger display). But nobody develops for Windows Mobile any more, so your only hope is for Android to become stable on that hardware. (I've got a Touch Pro RAPH800 - CDMA version, so I'm even worse off than the RAPH100 and DIAM100 ports.)
It doesn't matter. I'll undoubtedly be looking for a replacement at some stage*. The point, really, was just to highlight that the iPad and iPhone aren't the only devices which have that feature.
* My choice will mainly centre around what navigation software I can run on it. My current choice is Mapyx Quo; it has its quirks, but overall I like it. |
| |
[ Log in to reply ] |
| |
|
Andrew McCarthy |
Message #114716, posted by mac9 at 11:46, 27/6/2010, in reply to message #114704 |
Member
Posts: 16
|
Funny, I seem to recall similar threads on the Acorn forums a long time ago. I wonder if over time these things are any more valid today, than they were then?
To explain, I'm going to use an example of multi-tasking from a non-techie point of view. My experience is based on a Kinetic RiscPC.
I used to be able to run in parallel about twelve !Replay movies, with one lip sync'd on my system. For me it was a great demonstration of RISC OS at that time. You should have seen my PC friends gawp at this demonstration. I'd always be amused at their puzzled faces when I opened the case and they couldn't see any cooling fans.
There are many other great real world examples, documents editing and printing, start-up times, etc, etc....
Anyway, back to the thread. It's good to see that RISC OS is in the news.
BTW I've always thought that the original "ARM OS" was brilliant, just a victim of FUD and maybe its name - RISC.
(c) :-P |
| |
[ Log in to reply ] |
| |
|
Eric Rucker |
Message #114719, posted by bhtooefr at 15:19, 27/6/2010, in reply to message #114716 |
Member
Posts: 337
|
In the case of twelve !Replay movies, that's purely a case of how well !Replay is written - if it plays well with eleven other sessions of itself, then CMT is obviously fine for that specific case.
The problem with CMT is that, fundamentally, it depends on the application passing control back to the OS, rather than the OS taking control from the application. Therefore, an application that takes control and never gives it up can essentially hang the system. (IIRC, some modules are already PMT - for example, AMPlayer keeps playing even after the system is hung - but that's another story, and doesn't translate to making apps PMT.)
So, applications have to be written to be extremely stable, and one buggy app can take down the whole system. Even if every app IS stable, if it's doing something CPU intensive, it has to stop doing what it's doing to do a Wimp_Poll, or the system is unresponsive. If it's PMT, that buggy or CPU intensive app may hog CPU time, but other stuff can still run.
Many of your other great examples either aren't true (CMT OSes are terrible about being unresponsive when printing,) or are because RISC OS isn't bloated.
A 6 meg PMT OS would be equally fast at booting, and would be smoother at every task that you've mentioned. Even the Replay playback.
(Also, Replay playback isn't necessarily a good benchmark. I suspect modern codecs are far better in bitrate for a given quality, and may even have lower CPU utilization if the quality is the same as Replay. Not to mention, there are other PC OSes that could do the same party tricks, and they were PMT. It was mainly Windows that had trouble with this one, not much else.)
[Edited by bhtooefr at 16:19, 27/6/2010] |
| |
[ Log in to reply ] |
| |
|
Trevor Johnson |
Message #114886, posted by trevj at 12:22, 28/7/2010, in reply to message #114703 |
Member
Posts: 660
|
...to see the comments from old hands such as Eddie Edwards... The Endgames thread just prompted a search, the results of which included this Eddie Edwards interview.
Thank-you for doing this interview, you will now be immortalised in the archives of Acorn Arcade! I don't quite know what to say.
How's that for nostalgia? |
| |
[ Log in to reply ] |
| |
|
Trevor Johnson |
Message #122870, posted by trevj at 15:35, 29/11/2013, in reply to message #114703 |
Member
Posts: 660
|
Here's a bit of Acorn/ARM history in the mainstream media again. And there's a Steve Furber video too. |
| |
[ Log in to reply ] |
| |
|
