@-pekr-

Quote:

It would surely be interesting to compare those two business plans.

My estimation is 1 milion EUR budget and two years of work for x86 port. You can start bounty.

@KingKong

Quote:

AmigaOS needs only to be stable and secure (memory protection would be nice but should be possible) and must have a good Linux emulation (MSwindows emulation would be nice too) - that's nearly all for starters.

None of that can/will happen without at least memory protection, and you won't get sufficient performance to compete without also adding SMP. That's a large part of why I've mentioned both. It requires a *massive* lift in terms of OS features.

Quote:

The price of the X1000 may be high - let it be 2000 ¤ - but some hardcore gamers, hifi-freaks, etc. pay more (though not for the X1000)

They pay more because they want hardware that's perhaps 10 times as fast as the X1000 will be.

I intend to buy the X1000, but I want to buy it because I have a special relationship to Amiga and AmigaOS. Most people don't, and won't be willing to pay for something that costs far more than a typical PC without performing accordingly.

Quote:

In mass production the X1000 can be as cheap as any PC.

That doesn't help when there's no possible way to get the kind of demand that would be required to ship the many tens of thousands of units a quarter that would be necessary to get the kind of economies of scale required to bring the price down that low.

If we're lucky we'll see volumes sufficient to start making tiny dents in the price, but there's no way we'll see volumes that can bring the price down towards PC territory anytime soon (as in: not unless we manage several years of rapid growth, which is at best highly improbable).

Quote:

The advantage of AmigaOS is that it's much smaller than MSwindows and therefore (potential) faster, less power consumptive, less bug-infested, more usable for small devices and real-time computing.

And all of these are irrelevant in a day when my *phone* has a 1GHz ARM CPU with 512MB RAM that can run *java* apps fast enough for me not to notice any latency.

The world has moved on. While having a small compact OS is still great, it isn't sufficient of a differentiating factor when hundreds of MB of memory cost as little as it does, and the systems needs a fast CPU to handle things like video and javascript and flash anyway - the performance and memory benefits of AmigaOS vs., say, Android or plain Linux are lost in the noise for most users.

Quote:

I see the chance for AmigaOS to kill MSwindows - one (the EU) must only want.

Sorry, but that's pure fantasy all the time it doesn't have a feature set that would make it possible. It'd take dozens of full time developers for several years before it could hope to have the features needed to even be an option. Who would fund that when they could invest the same amount in Linux - a system that's far more mature, secure and feature complete - and that is well known and understood?

I'd love for us to get AmigaOS (or any derivative) to that stage some day. Stranger things has happened. But even if *everything* goes 100% perfectly in our favor, and investors magically line up to help grow the market, it'd still take many years of hard work before it'd even be possible.

It's more realistic to think it won't happen, but that perhaps we will be able to grow the community to a few tens of thousands over a few years. Even that will require a lot of things to go just right and a lot of people working really hard.

_________________
Wiki for new/returning Amiga users - Projects: ACE basic compiler / FrexxEd / Git

@pavlor

Quote:

My estimation is 1 milion EUR budget and two years of work for x86 port. You can start bounty

That's hugely overestimated. Most of OS4 is written in C. If basically one man can port AROS to ARM in a course of one or two years in his spare time I'd like to think Hyperion could do a x86 port much faster.

Last edited by WolfToTheMoon on 18-Apr-2011 at 09:46 AM.
Last edited by WolfToTheMoon on 18-Apr-2011 at 09:46 AM.

_________________

@WolfToTheMoon

Quote:

That's hugely overestimated. Most of OS4 is written in C. If basically one man can port AROS to ARM in a course of one or two years in his spare time I'd like to think Hyperion could do a x86 port much faster.

No, if you want to maintain backwards compatibility.

Sure, plain x86 port would be much faster (and less expensive) - but it will be as useable as AROS for PowerPC (or ARM).

@OlafS25

You're right that software decides, but you're also right that no company will write new games and apps for a few hundred users. That means we're going to be largely dependent on hobby projects and ports.

For ports, features like memory protection would make things *massively* simpler, even if it's not enabled by default for existing applications. It would allow far better emulation of Linux/Unix API's, for example - sufficient MP support would even make it possible to write a compatibility API to let PPC Linux apps run unmodified - for Windows there's for example CoLinux that allows you to run Linux; having those kind of capabilities would be good temporary measures to plug "holes" in application availability.

But MP is also pretty much a must to be taken seriously. Having your OS crash because of badly behaving applications just isn't acceptable any more. My Linux boxes at work have uptimes measured in *years* even though we do development work on them. That is what is expected of a modern OS.

_________________
Wiki for new/returning Amiga users - Projects: ACE basic compiler / FrexxEd / Git

@pavlor

Quote:

It is clear that OS is not important

No, it is clear that good, stable OS feels unimportant. I'm talking from current perspecitve. In the 90s we've lived in a different computer world. It just can't be compared.

_________________
OpenWindows Initiative. Port PS3 hardware to bananas. For free. Join today and receive expired $50 cupon from AI!

@WolfToTheMoon

Quote:

That's hugely overestimated. Most of OS4 is written in C. If basically one man can port AROS to ARM in a course of one or two years in his spare time I'd like to think Hyperion could do a x86 port much faster.

How many would be interested in OS4 with no backwards compatibility (apart from UAE) with M68k or PPC apps? Effectively starting from scratch in terms of applications.

Consider how many people seems to be regularly using x86 AROS... Granted, OS4 on PPC is more mature than x86 AROS, but nevertheless application availability is a big enough problem for OS4 on PPC. A port without JIT support for both m68k and probably PPC too would likely be DOA.

That drives the cost up a lot.

Then there's additional features. A straight port would leave potential new users totally underwhelmed, and alienate the portion of existing users that wants to stay PPC.

You're also ignoring the issue of moving to a platform with different endianness. Having compiled a few old Amiga programs for AROS I can tell you it is likely to bite you in a lot of unexpected ways.

_________________
Wiki for new/returning Amiga users - Projects: ACE basic compiler / FrexxEd / Git

@opi

Quote:

No, it is clear that good, stable OS feels unimportant

Even AmigaOS can be stable enough for daily work. Sometimes I even use my WinUAE box for hours without crash.

@pavlor

Since Commodore went under in 1994, Windows 3.x is what Amiga was competing against. Even though 386 PCs with VGA and Sound Blaster cards were superior to AGA Amigas, Workbench was so far ahead of Win3.1 that overall things were in Amiga's favour. Windows 95 took away a lot of that advantage, and Win 2000/XP obliterated it (plus PC hardware kept improving). It shows how much ahead of its time AmigaOS was in the 1980s when it was written, that it took an OS written in 1998/1999 to overtake it (most of the AmigaOS changes in the late '80s and early '90s were either for hardware support or cosmetic - the core OS remained unchanged since about 1987). But it's well and truly overtaken - a lot of the basic aspects of AmigaOS's design are very obviously rooted in the late 1970s / early 1980s (BeOS is an example of an OS that has a more modern design).

@vidarh

the problem is that MP breaks with old software so there is some kind of sandbox concept needed that is integrated in the Operating system (in best case without knowledge of the user). But besides of that I think that Software decides what you can do with a OS. So we should better think where are niches for the different platforms and what is needed.

@CodeSmith

Quote:

BeOS is an example of an OS that has a more modern design

Yes, BeOS is my favourite UI. However, I´m used to very limited UIs (like Windows 3.1) so AmigaOS is good enough for me.

@pavlor

Quote:

pavlor wrote: @opi Quote: No, it is clear that good, stable OS feels unimportant Even AmigaOS can be stable enough for daily work. Sometimes I even use my WinUAE box for hours without crash.

That's the thing - the OS itself is fairly stable, but you're completely at the mercy of the programs you run. If all the software you run is well written, you'll never have a problem. Unfortunately, a lot of software isn't well written, and a lot of Amiga software is written in C and assembler which are very susceptible to pointer errors - the ones that MP helps guard against. Software written in interpreted languages like AMOS also works a lot better in such an environment because the interpreter protects you against those kind of errors. Buggy AMOS software just dies with an error message instead of corrupting some other program you had running at the same time.

@pavlor

Quote:

pavlor wrote: @CodeSmith Quote: BeOS is an example of an OS that has a more modern design Yes, BeOS is my favourite UI. However, I´m used to very limited UIs (like Windows 3.1) so AmigaOS is good enough for me.

Actually I was referring to how the OS works internally, but I guess that too

I don't live far from the ARM headquarters in Cambs. I used to live right 'round the corner! Amazingly profitable as they get others to buy the designs and let others get involved with all the stress of manufacturing - a great strategy.

I have air conditioning in my office now - I keep a couple of branch servers in here so they justified it. The atmosphere will take some getting used to.

_________________
Amiga A1200, 3.1 ROMs, Blizzard 1230 MKIV 64MB & FPU, 4GB DoM SSD, Workbench 3.1

@vidarh

At least the ideas from developers and people are right. Hyperion has talked about plans for memory protection, SMP and so on. Since they are so silent and seem to work like crazy I'm wondering how much they are planing to do for the X1000 release? I hope that they don't work too hard on it meaning that they get burned.

I think that AmigaOS shall do things that people need most first and it seems that everybody agree on that. Timberwolf are being ported and Open Office is on its way too for example. A well working videoplayer or two, the OS features mentioned and maybe it will begin to grow.

And then there is the hardware issue. What hardware must be available? What price? What design? Quality... and so on...

I'm kind of thinking that it won't happen just to not get too disapointed should it fail but I'm following it with interest.

I won't say intel isn't useful to us anymore!

As I just reported THIS I now recognize that there is more than meets the eye regarding that decision.

It would seem the September announcement was following THIS he he

It would seem that performance of Freescale CPUs is bound to rise a lot due to competition at the higher-end, which in turn should have the nice side effect of providing Amiga with strong PPC CPUs in the future...

Besides, older Altivec from Freescale had 162 instructions, while their PR said the new one is an improved rendition (which seems to be proven by the fact the next one will have 180 instructions instead).

_________________
SamFlex Complete 800Mhz System + AmigaOS 4.1 Update 4
Amiga 2000 DKB 2MB ChipRam GVP G-Force040 Picasso 2 OS3.9 BB2
AmigaCD 32

Its quite likely that war is possible between the USA and The EU over AmigaOS as Kingkong rightly points out.

I know i will be labeled a conspiracy theorist has the USA been paying Hyperion to slow down the development of AmigaOS to hide its true power from the EU? Obviously Intel and MS pose a grave and present danger to the EU, I mean it would be a betrayal of the motherland if the new Panzer tanks have USA Intel technology. Now if they only had the superiour AmigaOS with PowerPC onboard i think kingkong and the rest of our fatherland would feel a lot more secure in the future of the amiga.

_________________
"ICE CREAM, ICE CREAM!" - Speedball 2.

"Look behind you, a three-headed monkey!" - Monkey Island

With a million $ you'd better off starting a complete new OS from scratch than porting a 25 year old non-secure outdated OS to x86... my two cents. Why insisting on staying compatible ?

_________________
http://www.warpdesign.fr/

@pavlor

Quote:

pavlor wrote: My estimation is 1 milion EUR budget and two years of work for x86 port. You can start bounty.

Your estimation imo does not hold any water Let's forget that 1 million EUR. What you said though is - it would take 2 years to acomplish. Two years = two years of salary for programmers, to acomplish the job. OK - how long are OS4 devs paid to bring us OS4 on X1000? How long are HW developers paid to bring us X1000?

It's just the same imo, because imo when thinking of ARM based devices out there (e.g. Beagle/Panda boards), those are already being developed by stand-alone community = no need to carry HW design expenses. I know we can't compare their performance to X1000, but Beagle based upon OMAP 5 might be coming soon enough.

Other than that - I don't want to argue here It is just my point of view, nothing more. IMO ARM is a good compromise to not enter desktop world, and yet being HW safe = having rather cheap HW available for your OS. There are also other advantages - with X1000 HW like design, you can't create power efficient and small enough devices, so it depends, where you want to go. ARM surely brings you much more options, e.g. for home automatition, etc. ...

-pekr-

@KingKong

Your post 314:
Quote:

KingKong wrote: I have absolutly no interest in a hobby-OS for a bunch of loonies -

loonies - ......LOONIES - ?

Yes! You do understand. Loonies, geeks, flakes, nerds, hobbists, non-conformists,
left-handers, nuts, tinkerers, dreamers,..............etc....

Sane people buy and use Windows boxes at work, at home, in their vehicle, at the
restaurant, during conversations or while sitting on the pot.

This is a different kind of party. It requires a different, unique, insanity. Some
people have it,, most people do not.

Lets P-A-R-T-Y--------Oh Yea!




- - -