@ChrisH

Why do you pursue this false notion that I already showed is completely false? The plan to eventually put AmigaOS4.x in a sandbox for backwards compatibility hasn't changed since 2001. You're just confusing the plans for the AmigaOS4.x series with the AmigaOS5.x series.

_________________
Sammy Nordström, A.K.A. "Samface"

MINDRELEASE.net - The Non-Commercial Network of Digital Arts.

Samworks D & C - Professional Web Development (in Swedish)

@ChrisH

Quote:

ChrisH wrote: They might as well call it OS5...

Spot on!

_________________
Sammy Nordström, A.K.A. "Samface"

MINDRELEASE.net - The Non-Commercial Network of Digital Arts.

Samworks D & C - Professional Web Development (in Swedish)

@Hans

Quote:

These problems could all be fixed with relatively minor changes to the Amiga OS API.

I would not call them relatively minor fixes. Certainly sandboxing is only solution to get full MP and have ability to run current software. But implementing full memory protection on the top of Amiga is still difficult even when you break binary compatibility (and source compatibility too).

Just count how many public system structures are there, then consider that message passing system must copy messages. Then there are other design flaws in Amiga like LockDosList(), hooks and what not. Amiga as a whole does not support MP very well. MP is not just isolated address space. You certainly wish flaws introduced in 2.0 get fixed, too.

_________________
Amiga Developer
Amiga 500, Efika, Mac Mini and PowerBook

@ChrisH

Quote:

I'm not stunned because this wasn't inevitable, because many said it was (just ask Ralph or Bernie...), but rather: I am stunned because the whole *design* of OS4 so far has been aimed at maintaining compatibility withOUT using a sand-box. It was also a key differentiator between OS4 & MOS, in the stupid red vs blue war.

While it is only way to reach full memory protection sandboxing is not holy grail either.

@debrun

I could agree...

Last edited by itix on 15-Aug-2007 at 07:47 AM.

_________________
Amiga Developer
Amiga 500, Efika, Mac Mini and PowerBook

Argument often turns on terminology. A couple of more quotes, in itallics, from Rogue in the same thread:

Quote:

I wouldn't call it a sandbox, rather a virtual environment. The old system will run in its own address space as a separate task. It's not like a UAE or anything like that. It might or might not be able to become aware of the outside world, we still need to clear up the details. Billsey wrote: So, what you're talking about is really like running two OS's at once on the same machine, with one acting as "overseer", so to speak. Basically yes. It's not actually a new concept, the Mach kernel can do this already.

Noel

@All,

Ok ok, try to keep it civil, no need for snipe remarks or MOS vs AOS / Blue vs Red comments.

_________________
~
Everything you say will be misquoted and used against you..
~

And all that has to run on a 400Mhz SAM ? or on the now old A1's ... ?

_________________
- Proud Member Of The Belgian Amigaclub Since 2003 -

The Belgian Amiga Club on FACEBOOK !

The Belgian Amiga Club Website

@Aminicle

You're assuming that its slow, or difficult to do. Neither of which is necessarily the case. Take running Shapeshifter on a classic Amiga. Thats another OS running in a sandbox. When 030 accelerators came out for Amiga 2000s, they became the fastest macs on earth.

_________________
Playstation Network ID: xeron6

@xeron

Both OS's in the same time ?

_________________
- Proud Member Of The Belgian Amigaclub Since 2003 -

The Belgian Amiga Club on FACEBOOK !

The Belgian Amiga Club Website

@Aminicle

Quote:

Aminicle wrote: And all that has to run on a 400Mhz SAM ? or on the now old A1's ... ?

Or under the Hypervisor of the PS3... ;)
Just think of it, the "old system" is running in a virtualized environment which then has access to the Hypervisor which arbitrates/virtualizes the low level hardware. Hmm...
At least it might provoke discussion about what "Amiga design philosophy" should be about.

Last edited by hatschi on 15-Aug-2007 at 08:54 AM.

@Aminicle

Yes. Shapeshifter runs MacOS classic as a task on top of exec.

_________________
Playstation Network ID: xeron6

@xeron

Quote:

Take running Shapeshifter on a classic Amiga. Thats another OS running in a sandbox.

Sort of sanbox. When Shapeshifter crashes it takes Amiga with it. If Amiga had memory protection there would be more overhead, obviously. (Not to mention problems with PrepareEmul patch needed by SS.)

Last edited by itix on 15-Aug-2007 at 09:28 AM.

_________________
Amiga Developer
Amiga 500, Efika, Mac Mini and PowerBook

@ChrisH

Quote:

saying that OS4 (as it exists now) is a dead-end design

Not really. This isn't a bad thing. In order to go forward we will need significant changes to the OS. The design elements of Amiga OS would surely be carried across but structurally it will have to change.

There is no bad thing in this. It's not like Amiga OS has some massively popular back catalogue that needs to run at full native speed. It can already all be run in a more than usable speed in UAE anyway.

To be honest I think if anyone wants to go forward with the Amiga principles from this stage a structural redesign is inevitable.

_________________
Sam

@Toaks

Quote:

So OS4.x will become morphos now?

"The compatibility break is inevitable" ... "moving ahead will include breaking compatibility"...

Hmm, maybe not after all...

After spending so much time, money and work to obtain a certain level of backwards compatibility ("Amiga compatibility" some would say), including spending *years* on creating a 68k JIT emulator, they will now throw all that away? All that time, money, work hours, tough nuts to crack, etc, turned out to have been completely *in vain*?

To me it seems like there has been no strategy, no long-term plan for this OS development, no management, only ad-hoc decisions that seems just as spontaneous as they are fundamental.

*This* is the "shocking OS4 news from Rogue" IMHO...

@takemehomegrandma

i don't think you're so dumb by saying that.Do you honestly think that improving the API to allow for memory protection while keeping all the actual thinks that work thanks to such *wasted effort* inside a controllated environment means dropping all the work they did? petunia can still run inside this environment,as well as all this other *wasted efforts* and tons of working hours made since they started working on it.
As if your beloved Morphos had advanced much that Quark concept other than the 3.x Layer since its conception.

Somehow your smilies don't make justice to your comments.

_________________
Cuida tus piedras gordas.

A1200/030 32Mb
A4000D
A1-X1000.

@debrun

Yes you're right. Let me quote Hen Bermans:"It's a well known fact
that the AOS4 people are using MorphOS source code"

OM(F)G!!

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

Anyone that is familiar with VMWare/VPC/Parallels will probably understand that this is a very general idea thats been around for years - and not something that is a copy of anything in particular.

For example - if you run Vista but use an XP VM does that mean you are copying MorphOS?

There are some cool tools on OSX now that go one step further and just let you run a virtualized application as if it was running natively rather than in a separate window as if it was run on a separate computer.

Its a great way to keep compatibility without sacrificing your ability to move forwards.

@ChrisH

The original AmigaOS design is old and outdated, this has been known for ages. Sooner or later it will have to be replaced by a clean new design, and old applications will have to be running in a sandbox. OS4 is a huge step forward, it gives us a PPC-native AmigaOS with many new features, but it's not the end of the road, it's an intermediate step. This is how OS development goes, as soon as they finish a particular version, they start working on a new version, throw out old outdated stuff and add new modern features/functionality, and making the necessary changes to the software architecture as they go along. What is shocking about this?

@Ami603

Quote:

i don't think you're so dumb by saying that.Do you honestly think that improving the API to allow for memory protection while keeping all the actual thinks that work thanks to such *wasted effort* inside a controllated environment means dropping all the work they did? petunia can still run inside this environment,as well as all this other *wasted efforts* and tons of working hours made since they started working on it. As if your beloved Morphos had advanced much that Quark concept other than the 3.x Layer since its conception. Somehow your smilies don't make justice to your comments.

Go easy on him, hes not a programmer.

@Ami603

Quote:

i don't think you're so dumb by saying that.Do you honestly think that improving the API to allow for memory protection while keeping all the actual thinks that work thanks to such *wasted effort* inside a controllated environment means dropping all the work they did? petunia can still run inside this environment,as well as all this other *wasted efforts* and tons of working hours made since they started working on it. As if your beloved Morphos had advanced much that Quark concept other than the 3.x Layer since its conception.

You can't just pour certain things into the Amiga environment without breaking compatibility. That's the whole point with having this "not-a-sandbox-and-nothing-alike-MorphOS" box in the first place, to keep the classic environment clean and intact while moving new ideas outside. Right? Problem is that they seem to mix those things into the classic environment nevertheless? Rogue: "The compatibility break is inevitable. We've seen it again and again, last time with the pager support - the current (beta) kernel supports swapping to disk. However, there are a lot of exceptions of memory that can simply not swapped (mostly because of the memory protection issues again)". I'm not discussing individual/particular technical issues here, rather the implications at large of these statements. The way they are working, by introducing certain stuff into the Amiga environment, will break Amiga compatibility and they have seen this time after time. And if they are giving up the compatibility, what was the point with all that efforts of creating the current level of backwards compatibility in the first place? That's what I meant with "ad-hoc development without a strategy"! Then what's the point in having a 68k JIT compiler at all for example, unless you plan to compile OS4 specific programs in 68k? You make comparisons to MorphOS - well, in MorphOS, the Amiga environment is running intact inside the "A-box". Compatibility has been prioritized, at least this far, which has been the core point with the "box" design from the start. Rogue obviously envision a different approach for the future: "moving ahead will include breaking compatibility, taking what we think is good and defining in AmigaOS, and leave the rest behind". To me that last quote looks like they pretty much will create a new OS, based on some AmigaOS philosophy (unclear though what the OS will be called, given Amiga Inc's stance regarding the Amiga brand in a Hyperion context), that will be inspired by AmigaOS but not really compatible. At least that is how I read mr Frieden. Or am I missing something?