@Crumb

I use 68k ext2 all the time on my AmigaOne OS4. I used to need a TD64 patch but I haven't needed that for a long time, so there must be some kind of TD64 support somewhere.

@dangerman

Quote:

I use 68k ext2 all the time on my AmigaOne OS4. I used to need a TD64 patch but I haven't needed that for a long time, so there must be some kind of TD64 support somewhere.

It's part of NSDPatch (which is integrated into SetPatch) - it patches TD64 devices to be NSD compliant.

_________________
"Miracles we do at once, the impossible takes a little longer" - AJS on Hyperion
Avatar is Tabitha by Eric W Schwartz

@Chris_Y

I didn't think there was a setpatch on OS4 - I'll have to have a look at that when I get back home.

@Crumb

Last time I used it NTFS and ext2 worked out of the box with AmigaOS 4.1

@dangerman

Of course there's a SetPatch command present in AmigaOS 4.1.

Last edited by cha05e90 on 13-Mar-2013 at 06:23 PM.

_________________
X1000|II/G4|440ep|2000/060|2000/040|1000

@cha05e90

Quote:

Of course there's a SetPatch command present in AmigaOS 4.1.

Indeed there is. I wonder why I didn't think there was. I must be going mad.

@dangerman

If you can recognise madness in yourself, then you're quite sane.

That's what I tell myself, anyway...

_________________
cheers
tony

Hyperion Support Forum: http://forum.hyperion-entertainment.biz/index.php

@thread

Development News

#6

_________________
This posting, in its entirety, represents solely the perspective of the author.
*Secrecy has served us so well*

@number6

64 bit address space awareness

Sounds good!

@number6

Good news.

Scheduler for multicore :)

Pluggable scheduler. Alltough executive will not sport in os4, it made huge difference in os3. Same type of possibilities...

PowerISA imlementation may be must have nowadays.

Gallium, I didnt get point at all :D

@mlehto


Quote:

Pluggable scheduler. Alltough executive will not sport in os4, it made huge difference in os3. Same type of possibilities...

executive, despite possible benefits (i have not been able to notice any in real life), has gained no foothold on genuine amigas. its good that the kernel gets improved, but two things;

1. it doesnt look like multiprocessing is any close. the features of the kernel are being prepared for that, but the actual implementation apparently has not even started.

2. the kernel has been "forked into separate branch" shich sounds pretty much like loss of backwards compatibility. and what of separate address spaces?

since it seems a major break with amiga philosophy towards that of linux&co, wouldnt it be easier to take a ready kernel that already supports mentioned features out of the box and build upon that? because otherwise it might end up like running after departing train.

Last edited by wawa on 31-Mar-2013 at 01:59 PM.

@wawa

I don´t share your point of view, but it is for sure possible explanation. On the other hand, if they maintain compatibility with OS3/OS4, it could be really big step forward in comparison to other Amiga-like OSs.

@pavlor

im a noob, but i thought giving up common address space means a breech of legacy compatibility. what remains is sandboxing legacy apps. how is this different to running legacy amiga stuff on any other system under emulation, like separate instances of uae (what has always been complained about aros compatibility solution comes to mind)? imho breaking compatibility (if at all) should happen once for good, in favour of any really serious benefits (fill in this space).

@wawa

In real life executive made OS3.1 more responsive and more multitasking friendly. I and many others noticed difference.

It is made for OS2.04 and is most likely full of 68k asm and is non smp, so it is only analogical to mention it with os4.

As for separated branch. It may bee in stage, where kernel is unstable. It is said experimental and stable. Nothing more or less. We are free to speculate why, but I see it as experimental/unstable and stable. Pretty normal in sw-development.

As for is multiprocessing possible or not, it is discussed to death. I think, that it is possible. Wich degree, without breaking backwards compability, I dont know.

It may, that old programs cant use SMP and runs as before and new programs can use SMP. Or then not. Anyway SMP-conversation has so much signal-noise, that I am pretty open about its possibility.

Many different parties saying nono, whereas there is politics behind. Dont want to start other flame-bait, so that is solely my own opinion :)

And for linux&co, if my amiga start to look like posix under the hood, I jump from wagon :) Thats main reason, why I dont have Apple. Even if userbase is pretty well isolated from posix, it is there and if one have problems, you have to dig there. I have enought that at work, so I dont need it at home :)

There is allso many different way to make posix-compability. OS2 from IBM got something from workbench, is win32 and posix-compilant and still looks and likes as OS2. AFAIK. Not used it for years now.

As for is multiprocessing coding started or not, world is full of specifications about situation. If/when exec APIs are prepared and tested, it is not traveling to mars implement SMP cabable scheduler. Of course it needs talent but I think that Friedens have that :)

It is not surprise, that it breaks system stability from start.

What wasnt said... Seems that mesa itself is more or less made. What holds radeon drivers, is not said anywhere. Should be intresting to know.

Last edited by mlehto on 31-Mar-2013 at 03:06 PM.

@mlehto

Quote:

PowerISA imlementation may be must have nowadays.

ISA is an abbreviation of Instruction Set Architecture the current Power ISA revision is 2.06b. The PA6T in the X1000 for example, is PowerISA 2.04 compliant.

"Work is in progress on yet another Power ISA implementation." is a roundabout way of saying we currently have new hardware you guys aren't yet aware of. Most likely something E5500 based although E6500 would be most welcome.

@mlehto

Quote:

Seems that mesa itself is more or less made.

software mode. mesa has been reported working in software mode a year ago already.. no wait, it is two years now:
http://blog.hyperion-entertainment.biz/?paged=4

go to blog entry from 5 april 2011.

thats what i mean by "running after departed train". same applies to updating timberwolf to latest firefox source.

@Rob

Quote:

"Work is in progress on yet another Power ISA implementation." is a roundabout way of saying we currently have new hardware you guys aren't yet aware of.

sure, ive got the message without your explanation as probably everybody else. i mean we are living of such vague hints and speculations since years. hopefully it isnt same story as the netbook one. at least another batch of x1k to be supplied till the end of the year doesnt indicate support for new devices is any advanced.

Last edited by wawa on 31-Mar-2013 at 03:29 PM.

@Rob

One can read, that it is new implementation.

@wawa

The reason the kernel was "forked" into two separate branches it to test backwards compatibility of the kernel with new functionality with both legacy and AmigaOS 4.x applications and software so bugs can be rooted out and compatibility issues can be identified and addressed.

There is a clear concept defined for multiprocessing since quite some time which is now being implemented. In fact, on an X1000, it is already possible to boot two independent instances of AmigaOS each running on a different core.

Separate address spaces: this was already mentioned numerous times, there IS already support for that in the existing kernel of AmigaOS 4.1, it will only be accessible for "aware" i.e. "new" applications.

_________________

@mlehto

Thanks. This is just great news, that OS 4.2 is work in progress.

Quote:

PowerISA imlementation may be must have nowadays. Gallium, I didnt get point at all :D

Gallium will be "3D for all" (hopefully Radeons 9700, X1600 and HD4000 series that are not in current Warp3D driver plan).

If I do understand, PowerISA is Hardware platform specs, so this basically means support for some new board.

Looking at current PowerISA specs, it might be a great high end.

Quote:

In fact, on an X1000, it is already possible to boot two independent instances of AmigaOS each running on a different core.

Sounds great for X1000 owners, will there be a promised ability to run Linux on other core from AmigaOS 4, removing need to dual boot?

Last edited by vox on 31-Mar-2013 at 05:08 PM.

_________________
Future Acube and MOS supporter, fi di good, nothing fi di unprofessionals. Learn it harder way!

@vox

A-EON has some Linux developers working on Linux distro's for the X1000, you'd better ask them. In theory it is possible but it would be a waste of a core for AmigaOS 4.2 if you ask me ;)

_________________