@hazydave
iam currently working for a company still using a VMS Alpha cluster for their backend. We're just in the process of swicthing to the new VMS Release for IA64 (really, it's funny to see VMS running on IA64 now), so it seems Intel got a new niche to place their IA64s. I have to admit that HP (and former Compaq employees) did a good job with porting to IA64 - if you have not written any very low-level code all the code can be ported with a simple recompile (Cobol,C or Forthran ... ), from the OS side and the behaviour almost nil changed. You can even use an image translator if you lack source-code for Alpha executables to convert to IA64 executables... Incredible, really. Of course, they implemented HAL (some parts already with vax->alpha but much more with this transition) as well as that is the only sensible way to guarantee consistent OS behaviour on different architectures without making porting a nightmare...

Concerning Hyperion, you are right about the timescale, it is way beyond schedule. As someone already said, afaik development had to be halted several times... This is not an excuse, rather sad reality.

Hyperion surely lacks resources to perform better (on a timscale), but apart from that iam glad at least something is done and I have an replacement for my a1200 (my a1000 / phoenix is setup and still used, just busy putting a HD in i) with definetly better performance. Sure, this is not near cutting edge, this is also not sufficient to "establish" AOS again in the desktop market - afaik this was not the proclaimed goal of hyperion with OS4 anyway. But it gives me and a bunch of others a fairly fast and more-up-to-date amigaOS system to play and work with. I hope Hyperion can succeed with their ambitions in the embedded area (i dont know any details so i cannot really judge these ambitions) - if this will enable them to maintain a PPC dekstop version of the OS this would be better than having to result to emulators on x86 for performance.

You said Amiga should strive for excellence, and i generally second that. I would wish that also the HW for a new AOS is excellent, not standard x86 either. Concentrate on what Amiga was excelling in (e.g. multimedia) and have AOS on CELL really exploiting the possibilities of this chip in the multimedia area and we would have something unique again. I don't know yet if this will be possible since as it stands there is no openly available CELL desktop / workstation although at least the statements from STI suggest that they will be available (which would also allow AOS to be ported, but surely not bundled with it).

@T_Bone

Quote:

T_Bone wrote: Console manufacturers don't want an OS on these things, the only reason they want to sell them is so they can sell games, they definately do NOT want people buying these things for any other porpose, they lose money on 'em.

It's true, but remember that Sony sold a PS2 Linux kit which, as the name suggests, allowed Playstation2 users to run that OS (and to develop games too with a somewhat lowcost, limited devkit) so I wouldn't say it's impossibile, on the other hand Microsoft has... well... at least ONE reason to prevent people from doing anything not Windows (or X360) flavoured on their game console

I vaguely remember that there was some sort of tax incentive for claiming that PS3 was a computer and not a console/toy.
Does anyone know what I'm talking about?
Is it an EU law?

Ok. Going back to my happy place now.
It's a hotdog, not a toy!!!

_________________
Leo Nigro, CTO Commodore USA, LLC
Opinions expressed are my own and not those of C= USA.
Commodore/AMIGA "Beautiful, High-Performance, Home Computers for Creativity and Entertainment."

@Hammer
Quote:

Quote: It should be noted, however, that SheepShaver is only a PPC *usermode* emulator. Most significantly, it doesn't support the PPC MMU; This would make its code a good choice to run existing PPC apps on a non-PPC OS, but it won't be any good for running any non-obsolete PPC OS on a non-PPC machine. Well, Rosetta doesn't emulate the entire PPC machine e.g. it doesn't support any kernal level extensions.

That's because Rosetta is for running PPC code on non-PPC OS. PearPC is for running PPC OS on non-PPC machine.

Sheepshaver is, essentially, comparable to Rosetta, except that it *does* run an obsolete Version of PPC MacOS which doesn't absolutely need an MMU.

If you want OS4/PPC to run on x86, you need something like PearPC. If you want OS4 PPC apps to run on a hypothetical OS4 x86 OS, you need something like Rosetta or SheepShaver.

@hazydave

Yeah, we agree the XBox looses money on every unit, but I'm saying the XBox GAME sales aren't even bringing in the money lost. It's not 5-6 games sold per unit to make up the loss, it's a lot more. For the 360 they're talking $300 sales price and $720 cost. That's $420 to be split up over 5-6 games at $50. How would you do that? Interesting math

So Microsoft would be rather silly for bringing up this with the Customs Office

The Playstation 2 is a pitiful machine, hardware wise. So the cost per unit really isn't high. It's comparable to any DVD player, actually. I can buy a DVD player for 499NOK, I doubt Phillips are loosing money on that, so why would Sony be loosing money on a slightly beefed up DVD player for 1200NOK? No, I think Sony are thruthful and they really do not loose any money on the units (they might not bring in much when R&D is taken into account, but they need that R&D just as much to sell the game licenses as they need to sell the hardware. So it makes sense to subtract the R&D cost from the total sum, rather than as a percentage from each unit).

I do not know right now how much R&D is calculated into that $720 number I came across for the Xbox360, might be interesting to find out, though.

Edit: ok, might not be that good a number, other sources claim it's loosing 120 or so USD per unit.

And Sony DID loose money on each unit until rather recently. The difference is that Microsoft NEVER got out of that situation. Nintendo are just way too good at making money, and have never sold at a loss, from what I gather.

Last edited by olegil on 18-Jan-2006 at 02:10 PM.

_________________
This weeks pet peeve:
Using "voltage" instead of "potential", which leads to inventing new words like "amperage" instead of "current" (I, measured in A) or possible "charge" (amperehours, Ah or Coulomb, C). Sometimes I don't even know what people mean.

@olegil

Quote:

olegil wrote: @hazydave Yeah, we agree the XBox looses money on every unit, but I'm saying the XBox GAME sales aren't even bringing in the money lost. It's not 5-6 games sold per unit to make up the loss, it's a lot more. For the 360 they're talking $300 sales price and $720 cost. That's $420 to be split up over 5-6 games at $50. How would you do that? Interesting math

Screw people on the upgrade hardware. (the people buying the $300 Xbox 360 would be better off paying for the deluxe version up front)

They've also been rumored to be working on a monthly-$$ MMORPG for the 360.

_________________
"If you want to tell people the truth, make them laugh, otherwise they'll kill you." - Oscar Wilde

@umisef

Quote:

umisef wrote: If you want OS4 PPC apps to run on a hypothetical OS4 x86 OS, you need something like Rosetta or SheepShaver.

But then binaries (and the OS) for x86 must be compiled using a compiler that flips endianess which results is speed loss. As you know yourself amigaish systems do *not* have a *seggregated* kernel space. That ties them quite close to big endian maschines or to non elegant tricks.

Only other option would be that the x86 AmigaOS would break with binary compatibility.

_________________
My programs: via.bckrs.de
MorphOS user since V0.4 (2001)

@hazydave

Quote:

Here we have $500 Macs, too. They boosted the Mac's popularity a little, but it's still a tiny, tiny segment of the market. Apple has kept it interesting and financially worthwhile by taking over much of the Mac software market themselves. This gets more money from each user for Apple, and ensures such applications will contiune to exist. However, it also makes the Mac market a dangerous one for any developer -- the investment you make in building a Mac product is at risk if, at any time, Apple can come along and clobber you with their own, in that small market. I'm not sure Apple had a real choice here; I know Hyperion and Amiga, Inc. don't have similar resources.

A mini mac is not too distant as specific from an A1. Is suerly most affidable than an A1 or a Micro nothing else. The fact is that Apple has many other resources (like IPOD and so on) that can spend in the desktop market. I'm agree with you that the desktop market seem closed to us. But i think that without any good company is too difficult. we have no money and so our market now is that. no x86, no other. a choice has been made and now we can only hope that in a future we can grow a little. nothing else and if one day apple we crunch all.. ok we have tried..

Quote:

Unfortunately, a market that doesn't attract users isn't one being kept alive. We'll see... assuming the product is ever officially in general release.

The only market that attract now are Phone cell and Console. But try to fight with Symbian and Sony is to difficult.. so what we can do?

Quote:

The DVD player has a Sigma Designs EM8620 video processor, which plays MPEG-4/DivX, Windows Media 9, and MPEG-2 in high definition. That's the way you get such players cheap enough -- throwing general purpose computing at a relative simple display device would be overkill.

That is surely less powerful than a dual x86 athlon or p4

Quote:

My next motherboard will certainly be for the 64-bit Athlon... or perhaps two of them.

no, i mean a new motherboard for us..

_________________
http://www.amigasoft.net

@afxgroup

Quote:

afxgroup wrote: Quote: The DVD player has a Sigma Designs EM8620 video processor, which plays MPEG-4/DivX, Windows Media 9, and MPEG-2 in high definition. That's the way you get such players cheap enough -- throwing general purpose computing at a relative simple display device would be overkill. That is surely less powerful than a dual x86 athlon or p4

That isn't the issue; the Sigma Designs EM8620 is an embedded processor that doesn't do anything else other than decode video streams. It has no overhead to speak of. All it's designed to do, what it's optimized to do, is decode video and audio. It's a display device, that's all. Yeah, it's less powerful than a dual x86, but it absolutely doesn't need to be.

A general computing sort of CPU has to be powerful to decode those streams because it's doing lots of other things at that time, too.

_________________
The problem with AmigaOS on PPC isn't that PPC is big-endian. The problem with AmigaOS on PPC is that PPC is dead-endian.

@TheDungeonDelver

i know, but this is a demonstration that we don't needed GHZ and GHZ only because the market want so. We (i) need an OS that could be a replacement of another if we (i) like it. And we (i) don't need a 3GHZ cpu if we (i) must write a document with word.

_________________
http://www.amigasoft.net

@hazydave

Quote:

For a PC, it's a question: you're always in competition with Windows/PC, simply because everyone knows The PC. For a "personal computing device", you'd have to come up with something that's interestingly different than a PC. For example, put AmigaOS, Cell, a DVD drive, and a fast network interface in a DVD/CDTV style box, and if you support the various multimedia stuff popular in today's advanced networked DVD players, you might also effectively sneak that Amiga personal computer functionality along with it. This is a very large niche that Microsoft's been gunning for, from above with their MediaPC idea, from below with the X-Box. Sony, too, puts the PS3 right there, designed as well to advance the new Blu-Ray format.

Exactly, this is where I was going previously with the 'get on PS3 or XBox' comment. Again, I'm totally ignoring IF any licensing fees exist to make an app for either of these systems. I don't work in the games industry, but we have some here that do that may know. How much up front $ vs royalties for licensing would also make a big difference here in the cash strapped (seemingly) 'Amiga market'/companies.

Look at what Apple is aiming for with their new media app and remote built into todays iMacs- they're giving you the start of the 'living room media center' as a 'side benefit' of buying an iMac, instead of going the MS route and saying, "Oh, here's a new OS, sort of, Windows Media Center Edition." I'm not sure which approach will win long-term, but think something similar to Apple- in their case, they 'give you' the media center bits on top of the OS. An AmigaOS approach would be the opposite- here's an app to manage your media, play some nostalgic games, oh, and BTW, you can also install the entire OS if you decide to, and make your XBox360/PS3 a 'real' computer. It might wind up being priced higher than the $60 new games by a little bit, but it gets some revenue in, and satisfies many of 'us' that want the OS functionality, plus some free marketing for AmigaOS at the same time in something _tangible_.

Note that for the media side, obviously it would need to do better than either of MS or Sony's built-in software, which could be a challenge, so that may drop things back to targetting nostalgia/UAE ready to run type setup as the primary functionality.

Quote:

But that's a big, big project (for Amiga-industry). You would need 30-something people, and a year or more. And in the end, you wind up in competition with the PS3 and X-Box, unless you're careful.

Yes, but perhaps not if you can do it ON either of those platforms- no competition, but maybe license fees.

Unfortunately, aside from that option, I'm not seeing inexpensive Cell chips coming out for the next year at least. I may be wrong on that, as I know IBM's desired model, but I expect them to be pretty near capacity with Sony for a while yet.

It might be possible to do a nostalgia/UAE/OS4 box with the supposedly forthcoming ACK card (plugin to A1200 or run standalone, 400MHz, DDR RAM), but once it hits over the $300 price point _for the complete box_, I think it's pricing itself out of the 'general consumer toys' category.

Short of those options, I'm back to 'run on commodity hardware, period,' because without significant $$ being invested, I see no way to even come up to parity on either price OR speed/performance, let alone both, compared to the $400 Dell etc...

Unless of course, there's some other 'embedded' market I'm simply ignorant on, which may be the case, but I keep getting to 'err, some sort of STB,' and besides that, simply am drawing blanks.

_________________
Are we not done with the same silly arguments and flames yet??!

@hazydave





Quote:

PC-133?!? That's a memory module not terribly long for this world. I was supporting 64-bit single data rate SDRAM back in the late 80s on the Metabox stuff. Still available, but the prices are on the rise, as no new systems use these modules. Even DDR is being replaced, slowly, with DDR2 (which, along with speed improvements, eases power requirements).

Well, the AmigaOne's northbridge Articia S is very sensitive about what PC-133 dimms will work.
It can handle registered modules in pairs, but not all of them, and can handle single unregistered dimms too, but not all of them either.
So what I wished for, was a northbridge that was a bit more tolerant, so it would be both easier and cheaper to find a memory-dimm that works 100%.

Agree that both DDR or DDR2 would be more desirable, not at least for the lower prices.

Last edited by tomazkid on 18-Jan-2006 at 05:09 PM.

_________________
Site admins are people too..pooff!

Quote:

Five years! Again, the orginal didn't take five years, why should a port.

The original was tied to the Amiga hardware. There were no extensions such as PCI, AGP,... to support. No retargetable sound, graphics,...

I don't think this justifies all the extra time it took (and it is taking) to finish OS4, but at least it may be a reason why it took more time.

Leo.

_________________
http://www.warpdesign.fr/

@hazydave

Quote:

ntel's sticking with the archaic shared bus architecture, on IA32 and IA64, is also certainly part of the problem. Particularly in SMP systems, even if the IA64 FPUs are fast enough, you just can't keep them fed off that congested single bus. AMD, meanwhile, did it right, even if that meant abandoning pure SMP for NUMA. And yet, that's the architecture that Cray wanted, and as well, perfect for the kind of huge, loosely coupled showcase machines Intel's been making all these years.

I've seen several N-way IA64 boxes in action as prototypes. (N way in this case I believe was 4 way, might have been 8, been a few years)... I don't know the architectural differences with x86/ia32 offhand, but from a usability standpoint, all I can say is the performance was NOT impressive. Things may have changed since then, but this became a commercial product from a big company that's already been mentioned in this thread. It did not fare very well, ia64/Itanium as far as I can tell, regardless of HP's pushing them, has for years now since initial availability, pretty much been a dismal failure.

Note Itanium and EMT64 (newish term for amd64/intel clones) are different (I know you know this, Dave, clarification in general)..buying quite a bit of server hardware each year, I've yet to come across a _need_ for Itanium boxes, period. AMD64 is pretty competitive, and EMT64 effectively a clone of AMD64...I'm just not seeing Itanium gaining traction at this point.
As it is, with Sun coming out with < $1000 1U AMD64 boxes, it's hard enough to even buy Intel EMT64 boxes in the low and mid-range for servers. If AMD became able to price match to OEs or better (Intel still holds the Centrino card for some..), I'd be hard-pressed to see any reason to even buy Intel EMT64, let alone Itanium.

_________________
Are we not done with the same silly arguments and flames yet??!

(Damnit! Can't Xoops handle more than two levels of quoting?)

@hazydave

Quote:

hazydave wrote: Quote: How do they run Linux "at MS's expense"? He's technically right about that. Microsoft sold the X-Box at a substantial loss (the estimated retail, if sold for-profit, runs over $700, according to at least one detailed estimate I read, once you factor in advertising , retail markups, etc). ...

Yes, I know this. But I don't think the loss made on people running Linux just-because-they-can (and who also don't buy any XBox games) even compares to what MS pays in a week to get the MS campus lawns mowed. :)

Quote:

I think it's more the obscurity of a so-far undocumented new hardware platform, rather than simply "repackaged PC" that gets Microsoft some proof against such hacks. The PowerPC itself isn't obscure enough to matter. Of course, they may just be raising the bar for the system hackers -- with the right tools, someone fill figure out the X-Box360 and put Linux on it, too. Because of the challenge, and probably becau

Hey, it seems you were interrupted! :)
I agree. The protection and lock-in is basically a result of the purposely designed DRM and secrecy. Not because it's running a PPC derivative instead of x86 or whatever.

Quote:

Quote: Your XBox example confused me. When you talk about "lock-in", you're really referring to the chosen CPU? ... I took his "lock-in" to mean that the OS and the hardware are locked-together: you can't run the OS elsewhere, you can't run anything else on the hardware.

Yeah, but that would make sense. :) It was the earlier musings about "lock-in-because-of-PPC" that were confusing me.

Quote:

Quote: Are you perhaps hinting at the old "running on PPC will protect us from competition" myth? I do believe he's citing that particular myth. That logic only prevails in a world where each person is allowed only one computer. Otherwise, locking hardware and OS together in both directions simply makes both less appealing, resulting in fewer sales. Whether that's significant or not (eg, is it many fewer sales, or negligable) has lots to do with the usefulness of that locked OS and the attractiveness of the hardware. In this case, locking the OS to the hardware pretty much guarantees that substantially fewer people will be running AmigaOS, than were the OS available on an existing hardware platform. Or even a new commodity priced hardware platform that would also run other existing OSs (eg, the tack that Apple's taken).

QFT (quoted for truth), as the kids say these days! :)

IMO, locking your OS to "special" hardware from "special" vendors could be a definition of "counterproductive", unless you also have some substantial weight to throw around on the hardware market (which AInc/Hyperion don't have, they make a niche OS, and that's it). And that's regardless of CPU.

_________________
Oh, bother.

Quote:

Samwel wrote: Rogue has said a desktop system is needed for developers and also for his own preference Who would want to program and use a OS through a PDA???

But why is a whole desktop system needed to develop on?
I didn't have to buy a 68k desktop to code for the old Palms, PalmSource had a
SDK and emulator/debugger that ran on the PC.
A mate of mine does stuff for Windows Mobile PDAs, he didn't have to buy an ARM powered
desktop.

Or you could think lateraly..put a big fresnel lens over the PDA and use a USB keyboard

_________________

@Doobrey

Quote:

Doobrey wrote: Quote: Samwel wrote: Rogue has said a desktop system is needed for developers and also for his own preference Who would want to program and use a OS through a PDA??? But why is a whole desktop system needed to develop on?

How many people would develop for OS4 if there was no desktop machine and no desktop AmigaOS users? I imagine the only OS4 ports that would happen are the ones Hyperion did themselves, or by people working directly with Hyperion and it's embedded customers.

_________________
"If you want to tell people the truth, make them laugh, otherwise they'll kill you." - Oscar Wilde

About Amiga on those gameconsoles and such. I read in the developer section at aros.org and found out they´ve even improved the API for better hardware independence. Aros was intended to be portable from scratch.
It´s probobly more likely aros will run on those machines than AOS4.
While OS4 is further developed than aros I see more problems ahead in its path. Aros flyes free with its community at the wheel. No obsticles aside from a very small community. I think that could change with time.

What will be here first, AROS HD installer, hardware for AOS4 or KHTML for MorphOS. It´s defenately gonna be a good amiga year!

@hazydave
Quote:

At the right price, Cell would be pretty cool. It's a nearly perfect processor to put into an advanced set-top box: cheaper (in theory) and cooler (also in theory) than an x86, more power, potentially eliminating any need for video acceleration chips (eg, if I can do H.264 HD in something reprogrammable, no need to buy a dedicated decoder; the power I'm paying for can be used for other purposes when not playing back next-gen video, and there's the ability to issue software upgrades, as new audio/video formats come along). For a PC, it's a question: you're always in competition with Windows/PC, simply because everyone knows The PC. For a "personal computing device", you'd have to come up with something that's interestingly different than a PC. For example, put AmigaOS, Cell, a DVD drive, and a fast network interface in a DVD/CDTV style box, and if you support the various multimedia stuff popular in today's advanced networked DVD players, you might also effectively sneak that Amiga personal computer functionality along with it....

Now the discussion is getting interesting.

Depending on budget and target market, why not include a PWRficient CPU into the mix with the Cell.
I suppose you've seen all this before, but for others, here are some links.

PASemi.
http://www.pasemi.com/downloads/index.html

CELL
http://www-128.ibm.com/developerworks/power/cell/
http://cell.scei.co.jp/index_e.html


From Hyperion's OS4.0 site. (Notice the mention of the Cell in the H/W "Box-out")
Quote:

If you are evaluating Operating Systems for an OEM product, the chances are high that AmigaOS4.0 can offer a unique advantage to your product, too

http://os4.hyperion-entertainment.biz/index.php?option=content&task=view&id=16&Itemid=

Cheers,
Tim

@hazydave

Quote:

Hey man, you're entitled to your opinion, and don't let anyone knock you down for it. But hey, you're also WRONG!

I may be, but don't bet on it just yet (keep reading).

Quote:

You should. But just in case you don't know, I've been programming since 1973, I have a degree in Mathematics/CS from Carnegie Mellon (it's a double degree, I also did Electrical Engineering, and nearly made a triple with Cognative Psychology, only I would have had to stay an extra semester to schedule things, versus what actually happened: underloaded last semester, just graduate Robotics and Compiler Design).

No, obviously, I didn't know. mea culpa.

Quote:

I wasn't in High School in 1972, or using computers.

Very few people were in 1972. 'Computing' for most was still batch entry by punch cards. My school was the first I'd ever heard of to have an in-house computer for the students. The 1130 was taken away a year later and replaced by punched-card batch entry to the district's NCR 7000.
Quote:

A year later, I started teaching myself programming, first on a programmable HP desktop calculator, next on a CDC Cyber 72 on my Dad's account (he was a department head at Bell Labs).

Lucky you. I didn't have the luxury of a family situation that allowed for a home terminal. I did manage to teach myself 8080 assembler, though, without a computer.

Quote:

Progamming lets you get close, but make no mistake -- OS Systems programming is unlike anything else. For one, it's the only place in programming where the hardware is a real issue.

No kidding.

(Looks up at bookshelf containing an Intel manual with -preliminary- data on 8085 and 432, reprint of PE article on CDP1802, book on KIM-1, NatSemi docs on NS32000 series, The S-100 Handbook, multiple binders full of documents on the H-8 and HDOS, The Soul of CP/M, the wonderful Que book Insider's Guide to PowerPC Computing, oh yeah, and Mapping the Atari).


[Chest-beating, mine's at least as big as your's, mode]
Did I mention I'm self-taught in 8080, 6502, 1802, NS32000, 1401, 68000, and 8086 assemblers? That college taught me Fortran, Cobol, RPG, and S/360 assembler (and aced every one of them)? That I picked up C all on my own? That I can read -most- languages the first time I see them (except APL - that's just wrong!) That I'm not just some VB noob who knows -just- -where- to stick an "Exit Sub"? Bits and registers don't scare me. I've done hardware banging.
(N.B. I've also managed to figure out the (entirely useless) bytecode for the Tandy Pocket Computer/Sharp PC-1211)
[/chest beating - *cough*cough*cough*]

[quote]I have lived it; probably more than you, unless you actually have been designing operating systems...

Last edited by syrtran on 19-Jan-2006 at 04:14 AM.
Last edited by syrtran on 19-Jan-2006 at 04:11 AM.

_________________
Tony T.

People who generalize are always wrong.


1989 - 500 / 1991 - 3000 / 1997 - Genesis Flyer 1200T / 2003 - A1XE