[News] Fleecy on Amiga's direction | ANN.lu |
Posted on 06-Mar-2001 14:01 GMT by Christian Kemp | 28 comments View flat View list |
More information on Amiga's future plans will be displayed on amiga.com, says Fleecy, but they "don't want to pre-empt St Louis".
There "will be one direction that the owner provides, and that will be revealed at St Louis."
"As for the AROS group, it would be nice to roll a lot of their effort into any future direction", he adds.
Read the original posting on the amigaone group at Yahoo.
|
|
Fleecy on Amiga's direction : Comment 1 of 28 | ANN.lu |
Posted by Christophe Decanini on 05-Mar-2001 23:00 GMT | Since the start I think that the classic emulation layer will be AROS on the AmigaDE. Aros developpers were interested in portable binaries so AmigaDe should be the good choice.
Of course AROS is not finished, the HP 68k emulator will be a better short term emulation for PPC Amigas.
I don't know how Amiga will manage to bring all efforts together but I really hope they can achieve it. |
|
Fleecy on Amiga's direction : Comment 2 of 28 | ANN.lu |
|
Fleecy on Amiga's direction : Comment 3 of 28 | ANN.lu |
Posted by Christian Kemp on 05-Mar-2001 23:00 GMT | In reply to Comment 2 (NihilVor): Yes, but the world, its dog, and all Amiga sites, already covered that one, so I figured I'd try something remotely different. :) |
|
Fleecy on Amiga's direction : Comment 4 of 28 | ANN.lu |
Posted by CCRider on 05-Mar-2001 23:00 GMT | In reply to Comment 1 (Christophe Decanini): I don't remember anybody saying that the infamous H&P's 68k emulator
ever reached an usable state... |
|
Fleecy on Amiga's direction : Comment 5 of 28 | ANN.lu |
Posted by NihilVor on 05-Mar-2001 23:00 GMT | In reply to Comment 3 (Christian Kemp): "The World"? Well, that message was posted Saturday-- and I have not seen it anywhere. Oh well-- there is only one extra paragraph there (the rest is posted in this news item-- I just thought Fleecy's post obscured what was being quoted [formating prolem]).
As far as AROS-- the message is a little too ambiguous to make any objective observations (although I can think of many possible scenarios)-- which is the point. |
|
Fleecy on Amiga's direction : Comment 6 of 28 | ANN.lu |
Posted by Christophe Decanini on 05-Mar-2001 23:00 GMT | In reply to Comment 4 (CCRider): I think the H&P emulator is not infamous.
I saw it running 2 or 3 years ago at the koeln fair.
Sam Jordan did a demonstration and it was running fine. |
|
Fleecy on Amiga's direction : Comment 7 of 28 | ANN.lu |
Posted by Ralph on 06-Mar-2001 23:00 GMT | Basically a 68k emulator IS possible - MorphOS proves it. The question is, why H&P never
released their thingy (just wondering).
And - NO: I am NOT Ralph Schmidt ;-) |
|
Fleecy on Amiga's direction : Comment 8 of 28 | ANN.lu |
Posted by Christophe Decanini on 06-Mar-2001 23:00 GMT | In reply to Comment 7 (Ralph): I think that Sam Jordan stated many times that the average speed of the 68k emulator would give 040 level. It does not make any sense for H&P to market the emulator as the only PPC card for now are dual CPU configuration.
The gain of speed would be not noticable so it is not very usefull.
With new products such as the Amigaone or the SharkPPC it will make a lot of sense. Then H&P could have an agreement with the hardware manufacturer to sell its emulator with every hardware. |
|
Fleecy on Amiga's direction : Comment 9 of 28 | ANN.lu |
Posted by Anonymous on 06-Mar-2001 23:00 GMT | I wonder if they will clear up their threats of legal action against the MorphOS team at St Louis?
Or will they continue in a Microsoftian direction, using all tactics, fair and foul, to kill off any perceived threat to their Chosen Path? :( |
|
Fleecy on Amiga's direction : Comment 10 of 28 | ANN.lu |
Posted by Mark Olsen on 06-Mar-2001 23:00 GMT | In reply to Comment 8 (Christophe Decanini): Sure, it would be useful. Besides speeding things up a lot, it would make coding easier, as you wouldn't have to care for cache flushes, how many contextswitches you make, etc. |
|
Fleecy on Amiga's direction : Comment 11 of 28 | ANN.lu |
Posted by Martin Baute on 06-Mar-2001 23:00 GMT | In reply to Comment 9 (Anonymous): Both sides of the story... |
|
Fleecy on Amiga's direction : Comment 12 of 28 | ANN.lu |
Posted by Martin Baute on 06-Mar-2001 23:00 GMT | In reply to Comment 10 (Mark Olsen): > Sure, it would be useful. Besides speeding things up a lot
...erm, but you read the comment about the 68k emu being up par with a 040? So where´s the speed gain?
> it would make coding easier
How so?
> as you wouldn't have to care for cache flushes, how many contextswitches
> you make, etc.
Ah, I see. That is, if your target user also has the 68k emu installed. If not, your careless use of context switches will slow *him* down considerably. Moreover, if I get the whole architecture right, the context switches would still be there, but emulated? The 68k emu is made to remove the necessity of two CPUs (and bus systems) in the system, but cache flushes and context switches would still have to be emulated to keep compatibility, no?!? |
|
Fleecy on Amiga's direction : Comment 13 of 28 | ANN.lu |
Posted by J on 06-Mar-2001 23:00 GMT | In reply to Comment 12 (Martin Baute): >> Sure, it would be useful. Besides speeding things up a lot
> ...erm, but you read the comment about the 68k emu being up par with a 040? So where´s the speed gain?
AFAIK, *current* speed - ie on a 604e PowerPC - runs at about 040/25 speed. So if you also have an 060 cpu as well there is no performance gain.
However, new systems use G3/G4 PPC so you are already gaining an increase in speed. Add to that the backside Cache and SD-RAM and you are looking at a lot faster performance. Personally, I'd guess that it would be now be running at 060 type speeds.
J |
|
Fleecy on Amiga's direction : Comment 14 of 28 | ANN.lu |
Posted by MAS on 06-Mar-2001 23:00 GMT | I was under the impression that the H&P 68k emu comes
with 3.9. If this is the case, will the emulation run
on the older P5/DCE dual 68k/PPC cards (and would such
a test be very informative if it was tried)? |
|
Fleecy on Amiga's direction : Comment 15 of 28 | ANN.lu |
Posted by Mark Olsen on 06-Mar-2001 23:00 GMT | In reply to Comment 12 (Martin Baute): Read before you reply?
> ...erm, but you read the comment about the 68k emu being up par with a 040? So where´s the speed gain?
I'm not saying that 68k apps will be faster, but PPC apps will be:
1. No cache flushes needed when doing a 68k<->PPC switch
2. 68k<->PPC switches are next to instant
On my 603@200MHz running MorphOS, Frank Wille's QuakeWOS gets as fast as it is in native mode on a 603@240MHz/060@66MHz.
Enough proof?
> Ah, I see. That is, if your target user also has the 68k emu installed.
> If not, your careless use of context switches will slow *him* down
Sure, my WarpOS programs also depend on people having WarpOS installed, they are completely useless on a PowerUp system. What's your point?
> considerably. Moreover, if I get the whole architecture right, the context
> switches would still be there, but emulated? The 68k emu is made to remove
Yes, but they are next to instant, which is not the case in a real 68k&PPC system.
> the necessity of two CPUs (and bus systems) in the system, but cache flushes
> and context switches would still have to be emulated to keep compatibility,
> no?!?
No, cache flushes on 68k<->PPC switches are not needed in MorphOS, as the emulated 68k and the PPC shares the same cache. |
|
Fleecy on Amiga's direction : Comment 16 of 28 | ANN.lu |
Posted by Nicholai Benalal on 06-Mar-2001 23:00 GMT | In reply to Comment 15 (Mark Olsen): > considerably. Moreover, if I get the whole architecture right, the context
> switches would still be there, but emulated? The 68k emu is made to remove
> Yes, but they are next to instant, which is not the case in a real 68k&PPC system.
You are both wrong. 68k and ppc code runs in the same context on MorphOS so there is no context switch.
Friendly,
Nicholai Benalal |
|
Fleecy on Amiga's direction : Comment 17 of 28 | ANN.lu |
Posted by Nicolas Sallin on 07-Mar-2001 23:00 GMT | In reply to Comment 16 (Nicholai Benalal): And this is not the case with WarpOS + his 68k emulator. |
|
Fleecy on Amiga's direction : Comment 18 of 28 | ANN.lu |
Posted by Martin Baute on 07-Mar-2001 23:00 GMT | In reply to Comment 17 (Nicolas Sallin): Exactly - I was under the impression we were talking about the H&P 68k emu, which is not affiliated with MorphOS, is it? |
|
Fleecy on Amiga's direction : Comment 19 of 28 | ANN.lu |
Posted by Paul Heams on 07-Mar-2001 23:00 GMT | In reply to Comment 17 (Nicolas Sallin): Yes, but there is no physical context switch between processors - As everything is running on the PPC.
IE: There is no interruption while there is a switch in processors - as the OS and Apps are only running on ONE processor (whether that is PPC Native or 68K emulated on the PPC). |
|
Fleecy on Amiga's direction : Comment 20 of 28 | ANN.lu |
Posted by Martin Baute on 07-Mar-2001 23:00 GMT | In reply to Comment 19 (Paul Heams): Sorry if I am insisting, but I want to get this straight...
The reason because a context switch was/is slow is that the CPU has to flush it´s caches to RAM before the other CPU can take over, to avoid inconsistencies due to unflushed write cache.
Now, are the existing 68k emulations emulating the command set, or the whole of the chip hardware (including caches)? I am not that much into Assembler, but wouldn´t a instruction-only emulation break if the code does some arcane tricks? And if the cache is emulated, doesn´t this also mean the emulated cache has to be flushed?
Please, this is genuine interest. I am a Software Engineer, not a Hardware Designer. |
|
Fleecy on Amiga's direction : Comment 21 of 28 | ANN.lu |
Posted by Anonymous on 07-Mar-2001 23:00 GMT | In reply to Comment 11 (Martin Baute): Go on then, tell us what the extra information you imply you know is.
We know Fleecy has repeatedly threatened to launch legal action against the MorphOS team in order to tie them up in legal battles toprevent them from releasing MorphOS for ages.
Does that not sound exactly like the sort of think Micro$oft do?
So why is it not acceptible for Microsoft to do this but acceptible for AInc to do it? |
|
Fleecy on Amiga's direction : Comment 22 of 28 | ANN.lu |
Posted by nOw2 on 07-Mar-2001 23:00 GMT | In reply to Comment 20 (Martin Baute): > Now, are the existing 68k emulations emulating the command set,
> or the whole of the chip hardware (including caches)?
Think about it, you only want software to run in the emulation and software has absolutely no access to the caches or other hardware beyond what is provided by the CPU's instruction set. As long as the cache bits operate correctly, you're okay.
The emulated CPU doesn't need to provide a 'virtual' cache as the host CPU running the emulation will cache stuff for you. |
|
Fleecy on Amiga's direction : Comment 23 of 28 | ANN.lu |
Posted by Björn Hagström on 07-Mar-2001 23:00 GMT | In reply to Comment 21 (Anonymous): "We know Fleecy has repeatedly threatened to launch legal action against
the MorphOS team"
This part may or may not be true.
"in order to tie them up in legal battles toprevent them
from releasing MorphOS for ages. "
But this part is invented by you.
/Björn |
|
Fleecy on Amiga's direction : Comment 24 of 28 | ANN.lu |
Posted by Anonymous on 07-Mar-2001 23:00 GMT | In reply to Comment 23 (Björn Hagström): Actually, you know, I think (s)he may be correct.
ISTR reading an exchange between Mr Moss and some other people (unfortunately, I cannot remember if it was an IRC log or a message board or what :( ).
They were talking about MorphOS and how much of a threat it could be to AmigaDE.
Fleecy said that all they needed to do was launch a lawsuit claiming that MorphOS was based in part on AmigaOS sources, and they could then tie up the MorphOS team of years in legal wrangles.
Sadly, I cannot remember where the hell that was, although I am pretty sure it was linked to from ANN, but I haven't been able to find it, so I can't really back it up, but just thought I'd point out that the other poster may be correct :/. |
|
Fleecy on Amiga's direction : Comment 25 of 28 | ANN.lu |
Posted by anonymous on 08-Mar-2001 23:00 GMT | In reply to Comment 24 (Anonymous): Actually, Amiga Inc. Had chosen Morphos as it's base for emulation in the Dec./Jan.
timeframe, but due to efforts from other amiga factions (Hyperion/Haage), that decision
has been reversed, and it will be announced in St. Louis. OS4.0 which will be basically
an very rough version of WarpOS w/68k emulation, rather then a more mature MorphOS.
This is why you haven't heard Fleecy spurting off about MorhpOS, until recently, now
the threats are starting again, he expects the MorphOS team to just step aside, I
don't believe this will be the case. |
|
Fleecy on Amiga's direction : Comment 26 of 28 | ANN.lu |
Posted by the man in the shadows on 08-Mar-2001 23:00 GMT | From what I have seen, Amiga is doing nothing more than protecting their intellectual properties (i.e. 3.1 roms) from being used without licensing on MorphOS. IIRC Ralph & co do not have the cash flow needed to license the roms just yet. One question is whether or not Amiga is willing to offer licensing of the roms to Ralph & co. After having tested the latest betas of Morph, I would love to see the implementation through however it looks interesting to see what is happening on the front of the dispute though. Ralph is working with bplan in order to get some of the hardware/software settings right... bplan is releasing their own AmigaOne system. It is possible that Ralph & co could do some interesting twists on the platform of the AmigaOne from bplan allowing MorphOS to run without being in violation of the intellectual property licensing from Amiga. Who knows. I hope they do well... both Amiga and MorphOS. |
|
Fleecy on Amiga's direction : Comment 27 of 28 | ANN.lu |
Posted by anonymous on 08-Mar-2001 23:00 GMT | In reply to Comment 26 (the man in the shadows): If they plan on going after MorphOS good luck, they will have to go after half of
the aminet authors too who have replaced Amiga API's with their own too. To me
MorphOS is the true successor OS to the Amiga, not this AmigaDE thing, and frankly
the people behind MorphOS have done much more for the Amiga then the current Amiga
owners have done or will do. |
|
Fleecy on Amiga's direction : Comment 28 of 28 | ANN.lu |
Posted by Luca on 09-Mar-2001 23:00 GMT | In reply to Comment 27 (anonymous): Are you too scared to show your name? Maybe because we know who you are? (well... I know) ;-) |
|
Anonymous, there are 28 items in your selection |
|
- User Menu
-
- About ANN archives
- The ANN archives is powered by #AmigaZeux. It was updated daily (news last: 22-Oct-2004; comments last: 18-May-2005).
ANN.lu was created, previously owned and maintained by Christian Kemp, www.ckemp.com.
- Contribute
- Not possible at this time!
- Search ANN archives
- Advanced search
- Hosting
- ANN.lu was hosted by Dreamhost. Sign up through this link, mention "ckemp" as referrer and he will get a 10% commission on any account you purchase.
Please show your appreciation for any past, present and future work on ANN.lu by making a contribution via PayPal.
|