@KimmoK

I'd rather see something entirely new. The window analogy is slowly being replaced by surfaces, which are slightly more abstract but still basically the same thing. At least a surface lends itself to thinking outside the box re: input devices: tablets, paper, tabletops, spheres, water vapor, etc. EDIT: (... an input device that is also a display surface.)

Last edited by Trev on 21-Apr-2009 at 09:09 PM.

_________________
Sam440ep-flex 733 MHz/1 GB RAM/Radeon 9250/AmigaOS4.1 Update 2
borked A1200/Blizzard1260+SCSI-IV/Z4+MediatorZIV/Deneb/Voodoo3/CatweaselMk3
more borked A1200/MBX1200z/Indivision
A500/clockport/RRNet
A600/A603

I wanted to start a new thread because this one is so off topic but I don't bother.

I hope that ambitious project means porting to Cell/PowerXCell. It's compatible with PPC, more unique than x86 and I hope many many ex-Amigans would be very interested in it.

Making many big changes into the AmigaOS have been in the roadmap all the time. It takes time, of course.

@KimmoK
How do you like your Sam ?

Last edited by Tomppeli on 21-Apr-2009 at 10:00 PM.

_________________
Rock lobster bit me. My Workbench has always preferences. X1000 + AmigaOS4.1 FE
"Anyone can build a fast CPU. The trick is to build a fast system." -Seymour Cray

@Trev Quote:

That troubles me (see my previous posts re: innovation), but it makes sense to stick with something that's well supported within the industry. Making OpenGL the native interface might be interesting, as long as it remains open (it's in the name, after all) and extensible.

The problem with big innovation is that it makes it very hard to use existing solutions. Unless you have massive momentum behind you, that puts you in a worse position than not being very innovative.

For example, an OS based on true capabilities (not the POSIX kind) would automatically be totally virus secure, but it also requires all software to be written from scratch, so it hasn't taken-off yet (*).

edit: In this particular case, having fast native OpenGL makes porting many games & other apps (e.g. Blender) sooooo much easier. What has Warp3D got us?

This is basically a version of "worse is better". If you haven't heard of this before, it was an essay explaining why C won over Lisp, despite C initially being inferior in every way (and despite it taking 20 years for C to get features that Lisp took for granted early on, like real-time debugging).


(* = although I think we are heading in the direction where people might actually prefer a limited-but-secure OS over being instantly infected by a virus as soon as they view a web site & having their internet bank account emptied. My current hope for a capability-based "OS" is Genode.)

Last edited by ChrisH on 21-Apr-2009 at 10:06 PM.
Last edited by ChrisH on 21-Apr-2009 at 10:05 PM.
Last edited by ChrisH on 21-Apr-2009 at 10:04 PM.

_________________
Author of the PortablE programming language.
It is pitch black. You are likely to be eaten by a grue...

@Trev

Quote:

Trev wrote: @ChrisH Quote: Hyperion (or the brothers anyway) have made it quite clear that they will NOT create a new API from scratch That troubles me (see my previous posts re: innovation), but it makes sense to stick with something that's well supported within the industry. Making OpenGL the native interface might be interesting, as long as it remains open (it's in the name, after all) and extensible.

Standardization is a good thing, and I don't think that creating a custom 3D API would be worth the effort, or even particularly innovative. What it has to achieve would remain the same. All it would do is reduce the amount of software that's available.

I don't think that this would be there "most ambitious project yet" because it's not all that ambitious. Granted, it is a big task, but it's not a huge one compared to the original Amiga OS 4 project.

Hans

_________________
http://hdrlab.org.nz/ - Amiga OS 4 projects, programming articles and more. Home of the RadeonHD driver for Amiga OS 4.x project.
https://keasigmadelta.com/ - More of my work.

@ChrisH

Quote:

For example, an OS based on true capabilities (not the POSIX kind) would automatically be totally virus secure, but it also requires all software to be written from scratch, so it hasn't taken-off yet (*).

There's no reason why the kernel couldn't be written this way, with the operating environment running in a free-for-all compatibility mode, but that would require a separation that I guess just doesn't exist in AmigaOS today.

@Hans, too

I agree that standards are good, but would a standards-compliant OpenGL actually be implemented, or would we end up with some kind of backwards compatible half-way house? All that compatibility could be shoved off into a separate layer, but will it happen? My own observation--admittedly naive, without owning OS4--is that the community wants the old and not the new. No one would buy into dropping m68k emulation, for example. That obviously limits what Hyperion can do while continuing to maintain a successful product. Maybe a branch is in order: AmigaOS Classic becomes all things that AmigaOS 3 was and AmigaOS 4 sustained; AmigaOS x becomes all things that AmigaOS can and should be (whatever those things are).

_________________
Sam440ep-flex 733 MHz/1 GB RAM/Radeon 9250/AmigaOS4.1 Update 2
borked A1200/Blizzard1260+SCSI-IV/Z4+MediatorZIV/Deneb/Voodoo3/CatweaselMk3
more borked A1200/MBX1200z/Indivision
A500/clockport/RRNet
A600/A603

@Hans
The argument against OpenGL is that it is a nasty mess of semi-compatible extensions (according to the developer of Infinity, who is an incredibly clever chap).

_________________
Author of the PortablE programming language.
It is pitch black. You are likely to be eaten by a grue...

@Trev Quote:

There's no reason why the kernel couldn't be written this way, with the operating environment running in a free-for-all compatibility mode, but that would require a separation that I guess just doesn't exist in AmigaOS today.

Genode aims to solve this conundrum by (a) running ON TOP of an existing OS, and (b) by having a "free-for-all" compatibility mode like you said.

BTW, if Hyperion made a brand-new OpenGL/Mesa implementation, I would imagine that they would look at adding a Warp3D wrapper on-top of it (given time). Even the current Warp3D does not run all existing Warp3D games, so arguably compatibility here is not too important.

Last edited by ChrisH on 21-Apr-2009 at 11:24 PM.

_________________
Author of the PortablE programming language.
It is pitch black. You are likely to be eaten by a grue...

@ChrisH

Quote:

ChrisH wrote: @Hans The argument against OpenGL is that it is a nasty mess of semi-compatible extensions (according to the developer of Infinity, who is an incredibly clever chap).

That used to be the case when the OpenGL committee moved so slowly that graphics vendors added their own custom extensions. With OpenGL 2 and 3, most things are officially part of the standard and the need to use extensions has decreased dramatically.

Microsoft could afford to create its own thing (DirectX); we can't.

Hans

_________________
http://hdrlab.org.nz/ - Amiga OS 4 projects, programming articles and more. Home of the RadeonHD driver for Amiga OS 4.x project.
https://keasigmadelta.com/ - More of my work.

@ChrisH

Quote:

The argument against OpenGL is that it is a nasty mess of semi-compatible extensions (according to the developer of Infinity, who is an incredibly clever chap).

Wasn't the 3.0 track aiming to fix that? The mess, I mean, by having more of the features developers use become part of the standard and keeping a closer eye on governance and adoption of new standards going forward.

EDIT: Hans beat me to it.

Last edited by Trev on 21-Apr-2009 at 11:43 PM.

_________________
Sam440ep-flex 733 MHz/1 GB RAM/Radeon 9250/AmigaOS4.1 Update 2
borked A1200/Blizzard1260+SCSI-IV/Z4+MediatorZIV/Deneb/Voodoo3/CatweaselMk3
more borked A1200/MBX1200z/Indivision
A500/clockport/RRNet
A600/A603

@ChrisH

Quote:

Genode aims to solve this conundrum by (a) running ON TOP of an existing OS, and (b) by having a "free-for-all" compatibility mode like you said.

After a quick read, sounds like a great place to start exploring a new shell/workbench replacment.

_________________
Sam440ep-flex 733 MHz/1 GB RAM/Radeon 9250/AmigaOS4.1 Update 2
borked A1200/Blizzard1260+SCSI-IV/Z4+MediatorZIV/Deneb/Voodoo3/CatweaselMk3
more borked A1200/MBX1200z/Indivision
A500/clockport/RRNet
A600/A603

@Trev & @Hans
This is getting rather off-topic, so I will just point you to his rants about OpenGL:
2005
2006
(AFAIK, he stick with OpenGL, despite being unhappy with it)

_________________
Author of the PortablE programming language.
It is pitch black. You are likely to be eaten by a grue...

@Tomppeli

I've had too little time for hobbies. And I made a mistake by starting to implement custom case for it, it takes too much time, even with proper tools.

But in one of these days it's AOS4's turn to boot up.


About the topic matter:

I wish AOS goes forward instead of making sidesteps to totally different CPUs.

Last edited by KimmoK on 22-Apr-2009 at 09:47 AM.

_________________
- KimmoK
// For freedom, for honor, for AMIGA
//
// Thing that I should find more time for: CC64 - 64bit Community Computer?

@ChrisH

Quote:

ChrisH wrote: @Trev & @Hans This is getting rather off-topic, so I will just point you to his rants about OpenGL: 2005 2006 (AFAIK, he stick with OpenGL, despite being unhappy with it)

We're not in 2006 any more. I don't doubt his gripes were true, but the OpenGL board has been working to make things more uniform. Besides, what other option do we have? DirectX 10? I don't think so. Using MESA leverages years of development; creating a new API from scratch means catching up with DirectX 10 and OpenGL whilst being incompatible with the dominant standards. That doesn't sound like a good idea to me.

Hans

_________________
http://hdrlab.org.nz/ - Amiga OS 4 projects, programming articles and more. Home of the RadeonHD driver for Amiga OS 4.x project.
https://keasigmadelta.com/ - More of my work.

Quote:

My own observation--admittedly naive, without owning OS4--is that the community wants the old and not the new. No one would buy into dropping m68k emulation, for example.

Sounds true to me.

Quote:

That obviously limits what Hyperion can do while continuing to maintain a successful product.

Why ? If the community drives the development, seeing so many "go arm, go mips, go cell, go whatever,... I want OS4 on 68k, etc...", I don't think they would go anywhere! :)

Now, seeing maybe one thousand people in the whole world bought OS4 (and I'm optimistic), would you call a 7 years long development like OS4... successful ??

I don't.

Going against the community may mean they could lose some (actual) users. So what ? Since it also means they may target a *lot more* users (and we're not talking about thousand users, but way more) and have a truely successful product...
More users mean more developers, more feedback, so a better product with more applications...

That was my two cents.

Last edited by Leo on 22-Apr-2009 at 12:30 PM.

_________________
http://www.warpdesign.fr/

@Tomppeli

Quote:

Tomppeli wrote: I hope that ambitious project means porting to Cell/PowerXCell. It's compatible with PPC, more unique than x86 and I hope many many ex-Amigans would be very interested in it. Making many big changes into the AmigaOS have been in the roadmap all the time. It takes time, of course.

I share your opinion. SMP and 64-bit are next big step.
Cell support was considered before and Rogue said that is the way AmigaOS should go.
He was very interested in cell, and I hope they decided to go for it.

PS3 port would be nice (since I have one ) but is not realistic.
Completely new Amiga Cell board would be much more serious project, but who will manufacture it?

Let's wait (again), that's all we can do!

_________________
I don't want a bubble gum, I want change!

@Kluz

"Completely new Amiga Cell board would be much more serious project, but who will manufacture it?"

How pitiless you are. Don't spread sentence like this.

A computer like this would be a Dream.

Now i can't brake my fantasies...



@hyperion
Ahem...
could we have a little info, please

/me fly to mars

Last edited by AmigaBlitter on 22-Apr-2009 at 07:41 PM.

_________________
retired

@ChrisH

Quote:

This is getting rather off-topic ...

If we can't discuss the pros and cons of the submissions re: Hyperion's "most ambitious project" then what's the point of posing the question in the first place? This is a bulletin board / discussion forum, right? I guess I don't see the point of posting ideas without others being able to acknowledge, discuss, and debate them. Maybe there's a way to lock threads so the original poster can grandstand and soapbox without rebuttal? Anxiously awaiting this post to be censored and replaced with "REMOVED: Off topic."

Anyhow, OpenGL is a Good Thing. OpenGL on top of an innovative API is also a Good Thing. If Hyperion could streamline the graphics subsystem to the point where OpenGL *is* the primary graphics API, including a standard set of supported extensions and a way for developers to provide their own extensions, both at the driver and application level, then that would be a Really Good Thing. It would certainly simplify the porting of OpenGL applications and may even encourage new development.

_________________
Sam440ep-flex 733 MHz/1 GB RAM/Radeon 9250/AmigaOS4.1 Update 2
borked A1200/Blizzard1260+SCSI-IV/Z4+MediatorZIV/Deneb/Voodoo3/CatweaselMk3
more borked A1200/MBX1200z/Indivision
A500/clockport/RRNet
A600/A603

@Trev

Port AmigaOS to this processor would be very, very ambitious:
http://www.physorg.com/news159548117.html

_________________
retired

@AmigaBlitter

Phew.. caught up. It'd be nice to be one of the inner circle and be in the know :D

Oh well. I thought before that AOS could only move forward in the direction of memory protection / SMP etc... but then they brought out the new 3D accelerated workbench (compiz based?) and cairo etc., so I'm not going to try to imagine what it could be.

tiffers

@Tomppeli

Quote:

Tomppeli wrote: I wanted to start a new thread because this one is so off topic but I don't bother. I hope that ambitious project means porting to Cell/PowerXCell. It's compatible with PPC, more unique than x86 and I hope many many ex-Amigans would be very interested in it. Making many big changes into the AmigaOS have been in the roadmap all the time. It takes time, of course.

But assuming for a second we still have the same roadblocks of having a PS3 version as before, if Cell is the target, what actual end-user hardware would we be talking about here?

_________________
Pegasos2 G3 running AOS 4.1 and MorphOS 2.0
Amikit user, tinkering with Icaros VM (AROS)
EFIKA owner
Amiga 1200