@NutsAboutAmiga

Quote:

NutsAboutAmiga wrote: @cdimauro So you finally admit it can be done

It can be done. If you sacrifice binary backwards compatibility. The same with SMP, "Silly SMP" showed how it probably could be done, but since it will still require a recompile of the S/W base and you won't be getting a "proper" 2015 SMP but something hacked to fit into the Amiga context, it's just... well, silly!

There are tons and tons of old Amiga 68k programs. They all have their own peculiarities, their individual certain prerequisites and assumptions regarding the environment they run in. Combined, they are a plain mess of unpredictable hazards, and the only way to be completely sure they will work as supposed, is to provide an environment as close to the original as possible. This means 31 bit unified memory space, single CPU, etc, etc. It's the only way to be sure.

But of course, if binary backwards compatibility isn't important to you, you are free to go ahead and do whatever you propose. But in that case, why stop at some half-arsed hacks and work-arounds, when you can aim for the real stuff instead? If you drop the compatibility requirements, go for proper 64-bit, proper SMP etc. And more importantly, go for a proper ISA/Architecture instead of holding on to the cursed PPC platform that has done nothing for us than denying us a chance of a future?

Question is; how thin can you bunch keep splitting those hairs?

Its our hobbies (which some actually even makes a living from), and is "logic" required if you have fun?

And with FPGA accelerators hopefully arriving, maybe developers like Daniel can find incentive to support the WB 3.x platform since people like me either buy his games if they are for Windows, ios OR wb 3.x.

If not, ill still enjoy wb 3.x without caring about what the politically correct name is.

@pavlor

Quote:

by pavlor on 22-Dec-2015 8:57:50 Quote: @cdimauro Quote: the first OS4 version was running with only Exec ported to PowerPC (mostly C code, and some PowerPC assembly to handle the usual stuff: interrupts & task switching, signals. I suppose), and the rest was running with 68K code. Note port did take 3 years. Developement started in 2001 with 68k components, ExecSG ran on AmigaOne at the latest in late 2002. We don?t know exact 68k code share in initial 4.0 pre-release (April 2004), but PowerPC versions of substantial parts (Picasso96, MUI) were released with the next update (October 2004).

I don't know what cdmaurio wants to count as the 'first OS4 version' but the first publicly available version ('Developer' prerelase April 2004) was 80% PPC

Virtual all of kickstart was PPC except graphics.library and P96
80% of C:
90% of Libs:

_________________
BroadBlues On Blues BroadBlues On Amiga Walker Broad

@TRIPOS

Quote:

Hehe, well, your contributions to this thread has been to counter verifyable facts and logic with fairy tales and personal "want-to-be's", so why change your way now, eh?

I don´t know in what world you live, but I´m tempted to make my
vacation there.

As an interest user what is "WB 3,2"`?

Not a long time ago bought 3,9 just for the fun of it to use the old and good stuff with a hd but without too many turbos and gfx cards for compability.

Another question: Why is there a disk-drive in an AmigaOne XE? Any use without the Catweasel?
(Never tried from 2002;) Having a few thousand disks could try how compatible the =4.1 is.

Third and last question: Does AOS4,2 Support the XE to fullfill our Really Really long wait?

@PR

Quote:

As an interest user what is "WB 3,2"`?

3.2 release was planned by Escom/Amiga Technologies back in 1996 (Walker has beta Kickstart 3.2) and later became what we know today as 3.5. What Tripos means is another question...

Quote:

Another question: Why is there a disk-drive in an AmigaOne XE? Any use without the Catweasel?

You may format DD/HD disks in FFS, but these will be not readable on Amiga (without proper mountlist, I think). So for Amiga purposes mostly useless without Catweasel.

Quote:

Third and last question: Does AOS4,2 Support the XE to fullfill our Really Really long wait?

I don´t think 4.2 was released...

@pavlor

Quote:

You may format DD/HD disks in FFS, but these will be not readable on Amiga (without proper mountlist, I think). So for Amiga purposes mostly useless without Catweasel.

AmigaONE-XE has PC floppy controller, so you get the typical 9 (DD) or 18 (HD) sector per track.

So does not support the 11 (DD) and 22 (HD) sector per track.

You can format the disk with what file system you like, but you also are forced to use PC format MFM encoding, making the disk sort off incompatible with Amiga500/1200/4000.

_________________
http://lifeofliveforit.blogspot.no/
Facebook::LiveForIt Software for AmigaOS

@TRIPOS

Quote:

There are tons and tons of old Amiga 68k programs. They all have their own peculiarities, their individual certain prerequisites and assumptions regarding the environment they run in. Combined, they are a plain mess of unpredictable hazards, and the only way to be completely sure they will work as supposed, is to provide an environment as close to the original as possible. This means 31 bit unified memory space, single CPU, etc, etc. It's the only way to be sure.

Or by decompiling, patching and hacking, crackers used to skip copy protection and do all sort of magic to get things running, if someone with the skills, where interested and did care about consequences, spent some hours hacking the software to work, I'm sure a lot more might run. Sadly most this people started to crack PC games instead, or should we be happy about that

Quote:

why stop at some half-arsed hacks and work-arounds, when you can aim for the real stuff instead?

Sure it be better to recompile 32bit programs, so they become 64bit programs, on case where you did have not choice does make sense to spend that time, working on virtual box, I guess.

That was not what we argued over, we argued if it was possible or not, not if was worth doing.



Anyway having a virtual box, might be nice to run in sort of protected sandbox, for testing software, with having to reboot, every time things crash. Can also be nice to test low memory conditions and so on. Not that you can't do that in UAE.

Last edited by NutsAboutAmiga on 29-Dec-2015 at 07:10 AM.
Last edited by NutsAboutAmiga on 22-Dec-2015 at 05:01 PM.

_________________
http://lifeofliveforit.blogspot.no/
Facebook::LiveForIt Software for AmigaOS

@PR

Quote:

As an interest user what is "WB 3,2"`?

The quotation marks around "3.2" does in this case mean "kind of", in the sense that it's really not, but could/should have been.

The November 2014 release of Workbench 3.1 is not a straight-off re-release of 3.1, it’s updated:

* Updated C/Version command (Y2K patch)

* Addition of Libs/workbench.library (for A-4000T 3.1 ROMs and 3.X ROMs)

* Updated S/Startup-Sequence (conditional SetPatch, for 3.X ROMs)

* Increased MultiView stack to 32768

* Installer script improvements

* Updated Installer (43.3 or 44.10, depending on CPU)

* Updated FastFileSystem to 45.9 (to support larger disks with 3.X ROMs)

* Installer itself is now part of the system installation (inside the Utilities directory)

Thank You for the answers. So Long time ago when bought 3,5 was it and 3,9 just some add-ons.

The disk drive is useless in the XE as suggested to swap with the A500/A1200. Back then was dissapointed. There is some hack around these and it's stupid to share files via e-mail with side-byside computers. Luckily could stick the multi CF reader to this pc and the A1200 but the XE is or could be shared by usb stick. Complicated as the usb is, well You know even with a hub.

Waiting for AOS 4,2 if it's for XE too or only for the newer ones. We Made it happen thinks so as 4 + 1 bought and a real lot of meddling around + help after Linux.

Last edited by PR on 22-Dec-2015 at 07:02 PM.
Last edited by PR on 22-Dec-2015 at 07:01 PM.

@pavlor

Quote:

pavlor wrote: @cdimauro Quote: the first OS4 version was running with only Exec ported to PowerPC (mostly C code, and some PowerPC assembly to handle the usual stuff: interrupts & task switching, signals. I suppose), and the rest was running with 68K code. Note port did take 3 years. Developement started in 2001 with 68k components, ExecSG ran on AmigaOne at the latest in late 2002. We don´t know exact 68k code share in initial 4.0 pre-release (April 2004), but PowerPC versions of substantial parts (Picasso96, MUI) were released with the next update (October 2004).

Thanks. Unfortunately it doesn't help to clear the question.

Quote:

pavlor wrote: Well, I don´t think anything without MMU is best way for future.

Unfortunately having an MMU means nothing, if you don't use it... appropriately, to achieve the goals discussed before.

Even some Amiga machines had an MMU, but it was of no practical use for normal users (developers appreciated, but they are the little minority).


@TRIPOS

Quote:

The image is wrong. AFAIK, AROS contributed both to MorphOS and AmigaOS 3.5.


@broadblues

Quote:

broadblues wrote: I don't know what cdmaurio wants to count as the 'first OS4 version' but the first publicly available version ('Developer' prerelase April 2004) was 80% PPC Virtual all of kickstart was PPC except graphics.library and P96 80% of C: 90% of Libs:

There was nothing important, Andy.

I was just trying to figure-out how much of the Amiga o.s. 3.1 code was written in assembly, and how much in in C. For this, I used some information which pavlor gave, albeit it wasn't enough.

Have you some detail about it? It's quite interesting to know. Thanks.


@NutsAboutAmiga

Quote:

NutsAboutAmiga wrote: @pavlor Quote: You may format DD/HD disks in FFS, but these will be not readable on Amiga (without proper mountlist, I think). So for Amiga purposes mostly useless without Catweasel. AmigaONE-XE has PC floppy controller, so you get the typical 9 (DD) or 18 (HD) sector per track. So does not support the 11 (DD) and 22 (HD) sector per track.

The PC floppy controller can reach at least 10 (20 HD) sectors per track. That's for sure. I don't remember now if someone achieved 11 (22 HD).
Quote:

You can format the disk with what file system you like, but you also are forced to use PC format MFM encoding, making the disk sort off incompatible with Amiga500/1200/4000.

The MFM encoding is exactly the same for any controller. What changes between the different systems is the track layout (the data structure used to store a sector's header/metadata plus the real data).

Only Macs had a slightly different controller (Apple, as usually, wanted to differentiate with all others), which changed the rotation speed on specific parts of the disk, making only part of it readable by the Amiga. But it was enough for exchanging some data between the two systems.


@NutsAboutAmiga

Quote:

NutsAboutAmiga wrote: @TRIPOS Quote: There are tons and tons of old Amiga 68k programs. They all have their own peculiarities, their individual certain prerequisites and assumptions regarding the environment they run in. Combined, they are a plain mess of unpredictable hazards, and the only way to be completely sure they will work as supposed, is to provide an environment as close to the original as possible. This means 31 bit unified memory space, single CPU, etc, etc. It's the only way to be sure. Or by decompiling, patching and hacking, crackers used to skip copy protection and do all sort of magic to get things running, if someone with the skills, where interested and did case about consequences, spent some hours hacking the software to work, I'm sure a lot more might run. Sadly most this people started to crack PC games instead, or should we be happy about that

This way you can change how memory allocation is marked, in order to correctly configure the MEMF_PUBLIC usage. This allows more RAM to be marked as private, and then swapped out. It also allows the usage of a partial resource tracking.

You can also intercept task & interrupts counters accesses, and convert them in more polite code in order to make a better usage of multiple cores (but we aren't yet talking of SMP).

But those changes will not solve all cited problems and, which is worse, require an enormous amount of time. So much time that it's easier and faster to rewrite the application from scratch.

A smarter approach can be used to avoid a full reverse engineering, saving a lot of time. But, similarly, it doesn't solve all issues.
Quote:

Quote: why stop at some half-arsed hacks and work-arounds, when you can aim for the real stuff instead? Sure it be better to recompile 32bit programs, so they become 64bit programs, on case where you did have not choice does make sense to spend that time, working on virtual box, I guess. That was not what we argued over, we argued if it was possible or not, not if was worth doing.

But now you're talking of modifying the existing binaries to patch them and fix some issues, whereas retrocompatibility means that you get the old binaries, and let them run as they are.

However, and as I stated, this doesn't solve all issues. In fact, you still cannot have the hybrid execution of 32-bit and 64-bit software.

Unless you cripple the whole system on a 31-bit address space...
Quote:

Anyway having a virtual box, might be nice to run in sort of protected sandbox, for testing software, with having to reboot, every time things crash. Can also be nice to test low memory conditions and so on. Not that you can't do that in UAE.

Well, that's a normal thing which a developer wishes, and this is the reason why having OS4 running on WinUAE can greatly help here.

Typing with the XE. Tried a disk and nothing happens. Better that way?

Maybe put off from the Kicklayout+wires.

Just interested could there be a card reader in that place?

Better PR not touch a corrected factory working mb. (Even as has changed some G4 but still)

If an card reader is and was cheaply I could try it and: in black;)

Offcourse =OS4.1 Update6 should recognize it ass mass storage from a pc or camera or Nokia.. Very normal in the pc land.

@PR

Quote:

Typing with the XE. Tried a disk and nothing happens. Better that way?

If I remember correct, you needed to use a program to mount the disk, as the PC floppy controller chip, can't sense the disk.

Last edited by NutsAboutAmiga on 23-Dec-2015 at 03:00 PM.

_________________
http://lifeofliveforit.blogspot.no/
Facebook::LiveForIt Software for AmigaOS

Not read all through this but I really wish OS4 could be ported to ARM. With the likes of the Raspberry Pi and many others now to choose from and most being very cheap. The ARM microcode seems to be bi-endian too so I guess that would make porting a little easier (I remember that being an issue with going X86).

I know this is an old argument but I really can't afford the PPC motherboards that are coming out for OS4 (nor the deskspace for another full machine).

Anyway this is all wishful thinking on my part ;) I would gladly support a Kickstarter to raise funds to port it to ARM to. I can dream can't I? Hehe :D

@Sky7

Quote:

Sky7 wrote: I have always wondered why AmigaOS 4.X was only developed for PowerPC based Amiga's and not for the Motorola 680xx Amiga's? I can fully understand the concept of trying to move the Amiga platform to a new processor family but the core Amiga user community back then (and now) was of the Motorola 680xx variety. From the perspective of potential sales, it seems like they limited themselves to a very, very small market (PowerPC based Amiga's) versus the general Motorola 680xx market. Was this a combination Motorola 680xx/Amiga AGA limitation thing that drove AmigaOS 4.x to be PowerPC only? -S7

FYI, I purchased AmigaForever 2016 with AmigaOS 4.1 FE for my Windows 10 laptop.

My point, a Wintel PC user can run AmigaOS 4.1 FE PPC by official Amiga Forever 2016 package.

Last edited by Hammer on 29-Dec-2015 at 04:12 AM.
Last edited by Hammer on 29-Dec-2015 at 04:08 AM.

_________________
Ryzen 9 7900X, DDR5-6000 64 GB RAM, GeForce RTX 4080 16 GB
Amiga 1200 (Rev 1D1, KS 3.2, PiStorm32lite/RPi 4B 4GB/Emu68)
Amiga 500 (Rev 6A, KS 3.2, PiStorm/RPi 3a/Emu68)

@iggy

Quote:

iggy wrote: @fishy_fis Quote: old, clunky CPU like a g5 The problem with your posts is you aren't really that informed about cpu architecture but you like to make broad statements. The cpus in the PS3 and Xbox360 were specifically designed with games in mind (and are still ahead of the cpu in the WiiU) and they are in order cpus. At 2.7 GHz the 970 is built on older technology (but is an out of order system, but is considerably more powerful per core than the Cell BE or the Xenon (with, I might add, better math capability than X86s of its period)..

From http://beyond3d.com/showpost.php?p=1788668&postcount=693

bkilian: Everyone talks about this 100GFLOPS, but all I've ever said is that the entire 360 CPU would not be able to run Shape at 100%. Remember that the 360 CPU gets, on average, an IPC of 0.2. So while the chip is _technically_ 100GFlops, you will never see that in real running code. Jaguar cores have an IPC of close to 1. A single jaguar core could almost outperform the entire 360 CPU on production code.


Here's Digitalfoundry's take on the encoder, not surprisingly the view is fairly negative.

http://www.eurogamer.net/articles/codecsys-video-encoding-on-ps3-article

That said, the speed is the key here, and make no mistake, CodecSys CE-10 can be stupendously fast. In my tests I found I could encode 720p at 35-40FPS, while 1080p came in at around 20-24FPS. That isn't exactly real-time, but it's still very, very fast. But speed counts for nothing in the greater scheme of things, if the quality of the encode you get out of the other end isn't particularly good. The freeware x264 encoder can pump out 1080p at a sustained 24FPS using just one core of a 3.0GHz Core 2 Duo if you turn off enough encoding features. Speed is important, but the whole point of h264 is that quality is king.



PS3 was using 5 SPEs at 3.2
PS4 was using 6 CPUs at 1.6 Ghz, PS4's games has access to 7 CPU cores
XBO was using 6 CPUs at 1.75 Ghz, XBO's games has access to 7 CPU cores.
X360 was using 3 CPUs at 3.2 Ghz

Last edited by Hammer on 29-Dec-2015 at 04:27 AM.
Last edited by Hammer on 29-Dec-2015 at 04:26 AM.
Last edited by Hammer on 29-Dec-2015 at 04:16 AM.

_________________
Ryzen 9 7900X, DDR5-6000 64 GB RAM, GeForce RTX 4080 16 GB
Amiga 1200 (Rev 1D1, KS 3.2, PiStorm32lite/RPi 4B 4GB/Emu68)
Amiga 500 (Rev 6A, KS 3.2, PiStorm/RPi 3a/Emu68)

@Hammer

the cell can fire today too if was with new gpu . ps3 is a great machine .

_________________
I love Amiga and new hope by AmigaNG
A 500 + ; CDTV; CD32;
PowerMac G5 Quad 8GB,SSD,SSHD,7800gtx,Radeon R5 230 2GB;
MacBook Pro Retina I7 2.3ghz;
#nomorea-eoninmyhome

The Cell had no GPU inside. It was the PS3 which had one.

Anyway, Cell wasn't a good processor. The PowerPC core wasn't good for general purpose computing, and the SPE's weren't as good as GPU stream cores.

@cdimauro

i know the cell is a cpu without gpu ...(dont make your troll way like your usual) ... but im sure if the ps3 have a new gpu with the cell that it already have you will not see difference in gaming like the ps4 or xbox one.

if you dont belive (like your usual way) ask to him if im not true

... yes .. the cell... was not good processor and for sure was not good for gaming ...i know ... just say this to all the games devs who make masterpiece for ps3..

just for remember some :God of war , heavy Rain, Uncharted , kill zone, resistance, the last of us ... etc .. etc ..

some comparisons
tomb raider

uncharted

Call of duty

Last edited by tlosm on 29-Dec-2015 at 08:04 AM.
Last edited by tlosm on 29-Dec-2015 at 08:03 AM.
Last edited by tlosm on 29-Dec-2015 at 07:54 AM.

_________________
I love Amiga and new hope by AmigaNG
A 500 + ; CDTV; CD32;
PowerMac G5 Quad 8GB,SSD,SSHD,7800gtx,Radeon R5 230 2GB;
MacBook Pro Retina I7 2.3ghz;
#nomorea-eoninmyhome

@tlosm

Quote:

tlosm wrote: @cdimauro i know the cell is a cpu without gpu ...

Then the next time write a correct statement.
Quote:

(dont make your troll way like your usual) ...

I reported the correct data, without trolling. Just plain FACTs which anybody can verify.

But, as usual, you have to go against me by principle. Typical of the troll that you are.
Quote:

but im sure if the ps3 have a new gpu with the cell that it already have you will not see difference in gaming like the ps4 or xbox one.

The problem is that a modern GPU can do much, much better than what Cell does. So, it doesn't make sense to upgrade the GPU and keep the legacy Cell.
Quote:

if you dont belive (like your usual way) ask to him if im not true

I don't know what you're talking about. You made such statement, so you have to report sources / facts for it.
Quote:

... yes .. the cell... was not good processor and for sure was not good for gaming ...i know ... just say this to all the games devs who make masterpiece for ps3.. just for remember some :God of war , heavy Rain, Uncharted , kill zone, resistance, the last of us ... etc .. etc .. some comparisons tomb raider uncharted Call of duty

So what? It doesn't change the fact that Cell wasn't good for gaming. Such developers had to make A LOT of work for creating such AAA titles, and it isn't an hidden secret that the PS3 teams inside a software house were much bigger than the XBox360 ones, due to the difficulties on developing on such crap hardware.