OS 4.0 comes to New York! - Updated: New Address

[magnetic]

@KurtGrach
    I have been demoing Pegasos and Morphos since APRIL. I also have written a complete review of Pegasos that was finished 2 months ago. Our website has the full review and downloadable MPEG videos of Pegasos running Morphos 1.3, Debian Linux, and Mac On Linux.  Please go to our site:
www.magneticsystems.com for more info!
   I did a demonstration of the PEG in late April early May at the very same place you are doing your demo! (NYU)  I have been doing demos privately and publically for MONTHS. You had NO interest in seeing the Pegasos in action, but now that I see you do I will put you on my mailing list and inform you of the next showing.
  NYCPUG (New York City Pegasos User Group) and Magnetic s will be announcing a public demonstration of the New MorphOS 1.4 Operating System, The Superbundle Software Package of NATIVE MorphOS apps such as Pro Station Audio and FX PAINT, as well as classic Amiga programs running Flawlessly on the Pegasos and Morphos... such as Image Fx, Amiga Amp, Photogenics, Pagestream, MOS MAME and more! That would be a good time for you to see how much power the PEG and MOS have!  Kurt:I'm glad you see the light
of the Pegasos in flight!
:-D

  I would love to see the Beta of OS4 you have, as I've only seen pics online of it. I am surprised its still in the 68k state. I thought it was going to be done by October..... what has improved since Euro Road Show? What Amigas apps run on it?You say there is a JIT 68k emu but Hyperion themselves said it is not running yet.  Why on earth are you running a Cyberstorm with the IDE controller? Doesn't OS4 support the UWSCSI of the CYPPC? Its a great controller and no one in their right mind would want to run off the old and slow IDE on the 4000....
   I wish you luck with your demo and I hope to be there to see it for myself. However, I'm in Cape Cod and may not be back to NYC in time. Lou Vidal of QAUG mayb e there to tape it..
   Kurt, If you or anyone elese in the area, or country for that matter, want a Pegasos Board or System or Systems we can provide it. We will be stocking Pegasos2 boards and turn key setups!!  How many would you like?
** Pegasos 1 boards are available NOW if you want one but ACT FAST!

magnetic :-o

[Skyraker]

I'm sick of seeing this OS4 on tour bollocks, I just wish they'd stop pissing about and release the bloody thing....

I remember when I used to care..... honestly it's a joke.

sKy

[Rogue]

I'm sick of seeing this OS4 on tour bollocks, I just wish they'd stop pissing about and release the bloody thing....

Very nice and uninflameable comment. Congratulations.

For your info, the individual showings around the globe have all been organized by dealers and/or user groups. Neither Hyperion nor Amiga had anything to do with it Believe it or not, there *are* people interested, and for those interested people the organizers have done their (excellent) job. If you don't like it, just ignore it.

I am sick of seeing people slander the organizer's work.

I remember when I used to care..... honestly it's a joke.

If you are not interested, why don't you stay away from these news items, then? (No need to answer, that was a rethorical question).

You are free to go or not go visit one of the showings.

[Rogue]

I am surprised its still in the 68k state.

It's not stricly 68k. Some modules are PPC, some 68k. From the top of my head here's a list of all PPC modules

intuition
gadtools
input.device
cia.resource
timer.device
SmartFileSystem
CDFileSystem
icon.library
diskfont.library
ft2.library
layers.library
con-handler
exec
expansion
ram-handler
utilty.library
dos
...

I don't get them all now :-) I might have been easier to name the 68k modules...

[z5]

@Rogue:

With things moving to PPC, is there a big difference in speed? Show reports have indicated that OS4 is slow. But with more and more PPC modules, is the speed up allready noticable?

(One wonders how fast it can be with a CPPC card as it is old hardware).

Aren't you afraid that ppl will compare MorphOS on Pegasos and OS4 on cyberstorm (which is not a fair comparison). Or are you confident that, once completed, it will be fast even on CPPC?

[alx]

@Rogue

IIRC wasn't intuition 68k at Amiwest (that's why the windows wern't moving as fast as they could) - I'd have imagined that PPC intuition would make GUI far more responsive :-)  And is DOS newly converted?

BTW I see that the news has been edited and the JIT isn't mentioned now - I guess that means it isn't running :-(

Keep up the good work with OS4 :-D

[Rogue]

With things moving to PPC, is there a big difference in speed? Show reports have indicated that OS4 is slow. But with more and more PPC modules, is the speed up allready noticable?

Sure there is a big difference. Right now most of the rendering code used to be 68k (especially intuition and reaction). Intuition is PPC now, reaction is following.

Aren't you afraid that ppl will compare MorphOS on Pegasos and OS4 on cyberstorm (which is not a fair comparison). Or are you confident that, once completed, it will be fast even on CPPC?

Well, it takes a huge amount of ignorance to compare a CSPPC against any modern machine and saying "hey, the Pegasos is faster than that". I mean, there isn't only a differently clocked chip at work, the whole architecture has changed dramatically.

Not that I wouldn't expect these comparisons  :-D

[Rogue]

IIRC wasn't intuition 68k at Amiwest (that's why the windows wern't moving as fast as they could) - I'd have imagined that PPC intuition would make GUI far more responsive

Yes, the Amiwest version was already a week or two old. We never show the latest stuff because we don't like surprises :-) So the Intuition at Amiwest was 68k, as was diskfont and ft2 (both of them make a big difference in rendering speed and font loading).

And is DOS newly converted?

Yes, DOS is only a few days old on the PPC. I never dreamed it was possible, with all the BCPL legacy stuff to be thrown out. Colin did a tremendous job at cleaning the code out and still staying compatible to old startup codes... Porting to PPC, after he finalized the code, was just a matter of hours. Cheers to Colin and Jörg for this.

BTW I see that the news has been edited and the JIT isn't mentioned now - I guess that means it isn't running

Well, like I said we never really show all the latest stuff, only if we know it runs stable. This is the reason why most of the time OS 4 didn't crash on these shows. (and yes, on the europe shows, people could actually play with it)

In the case of the JIT, it's got a few minor issues to be taken care of. Petunia is finished, it's just a matter of the integration into OS 4.

[Nickman]

Thank you Rogue for all the comments.. It's always nice to hear someone say there is progress.. Not just knowing there is.. If you know what i mean.  ;-)

[kgrach]

@Tom
Yes it supports the UWSCSI controller but I have lots of spare IDE drives. So it was easier not to mention cheaper. To Stick the SCSI.device V50 into the new kick. Then to break open a NEW SCSI drive. I might transfer things over to the SCSI bus for the demo. I have to see how things work out. I Just got some more PPC parts to replace the 68K code. Much of the code I believe has been ported from assembler to C.
So to do the switch from the current emulated 68K to full blown PPC is just a simple recompile of the existing source.
 
I did not mean to say I will be showing the JIT all I meant to say is that when released  AOS V4 will have both types.

The AOS V4 I will be showing is a mixture of PPC code and Emulated 68k code all running on a 200 Mhz 604. The A4000 is not my Tower but an A4000 desktop provided to me from Mr Hardware Computers. I just helped him out by intstalling  AOS V4 on the A4000. I now currently testing the system out and installing a few programs to demo.  


Tom I have played with MorphOS. I have always had  an interested in MorphOS. I did not make your two earlier demos because like you. I used to travel allot for my job. This is now no longer the case. I am also well conversant in what MorphOS is capable of.

All I wanted to say is if you want to see a
MorphOS demo is contact Tom and he will arrange one for you.

[amigatv]

magneticsnyc is not out there, magneticsystems domain is for sale, but this one is registered:

Registrant:
magnetic systems (LXTKXLZYND)
   918 metropolitan ave
   brooklyn, NY 11211
   US

   Domain Name: MAGNETICSYSTEMSNYC.COM

   Administrative Contact:
      Morris, Thomas  (35352375P)      metro.realize@verizon.net
      918 metropolitan ave
      brooklyn, NY 11211
      US
      718.599.0001

[boing]

What i'm curious to see is how the various PPC system emulating a real Amiga compare to the x86-based Amiga emulators.  In terms of speed, flexibility and reliability.  If one can get x86 boards for cheaper and run an Amiga emulator on it that's as fast or faster, then I hate to say it, but there goes a big reason to go with a PPC solution.  Word that OS4.0 is kinda slow- even on a 66MHz PPC, let alone more modern PPCs... well that's just depressing.

Now if the new "Amiga" had some special hardware that gave it clear advantages over the emulations- even just a way to assist AGA compatibility or emulation, then we'd have a good reason to stick with the corporate plan.  

 As it stands, we have no compatible hardware (at all! neither processor nor chipset),  When it comes to that point, one has to ask if there's any benefit in going with AI's plan, as opposed to say, X86 running an emulator.  Sad but true.  Nobody hates x86 more than me.  But one has to ask where the benefit is in paying more for a system that's no faster  at running existing Amiga programs than wintel running UAE or similar.

[boing]

>Much of the code I believe has been ported from assembler to C.


Cripes.  Weren't these Amiga Inc guys around in 1990 to remember how the transition from Assem to C caused 2.04 to be slower than 1.3?  C has proven to be nowhere near as portable as the hype of the early 80's promised, and it sure as heck is harder to deal with, and the compilers out there aren't even close to what could theoretically be reached (in terms of creating executables that are lean and fast).

Come on people.  Get yourselves some cross-assemblers if you have to.  Relying on a compiler to create fast code has always been a mistake.  Sure if you could trade speedy executables for faster development time, it might be worth it in a pinch.  But it's been over a decade and everybody is still struggling with C portability issues.  C is crap.  Stick a fork in it.

BTW the Apple community has several fast 68K emulators.  So why are we reinventing the wheel instead of licensing one of them?

[MagicSN]

Hi!

>Cripes. Weren't these Amiga Inc guys around in >1990 to remember how the transition from Assem >to C caused 2.04 to be slower than 1.3? C has

Most of the OS was still ASM at that point.
Believe me. Maybe you talk about that some
BCPL (predecessor of C...) code was replaced
with C-Code ?

>Come on people. Get yourselves some cross->assemblers if you have to. Relying on a compiler to

Nonsense. Do you know how DIFFERENT various
ASM (for different CPU families) are For example
on 68k you have an extremely fast 16 Bit shift
instruction (which leaves the other half-word unchanged). On PPC such an instruction does not
exist, and has to be emulated by using a whole
bunch of commands, being much slower.
The only reason for ASM is to be very close
to a SPECIFIC hardware. If you want cross-platform
that's an argument for C, not ASM.

ASM is not for Cross-platform developement, get
the message. If you translate an ASM-Source for
CPU-Family 1 to ASM-Code for CPU-Family 2
you might even end up (under certain circumstances) with something slower than a C-Rewrite. Of course you could do a complete rewrite (not 1:1... instead a complete reimplementation for the new hardware) in the new ASM-language. But why no Rewrite in C, then ? It has only advantages.

Ah, on a modern CPU there is no such huge speed-difference between ASM and C-Compiled code anymore (while on the 68k there is a huge difference...). This is because modern CPUs (like
the PowerPC) were designed with taking into account that very efficient C-Compiled-Code should
be possible.

>struggling with C portability issues. C is crap. Stick >a fork in it.

 Fine, you like ASM. But still
this does not require a C-is-crap attitude. Try to
stick to the reality - and check out the facts first
BTW: Sometimes I am using ASM too, so *I know*
the facts...

>BTW the Apple community has several fast 68K >emulators. So why are we reinventing the wheel >instead of licensing one of them?

Why spend money for an Emulator if we have an Emulator of our own ? Be also aware that
"just a command-by-command Emualtion" - like
in UAE - is not possible to use in an OS. The
Emulator needs to operate on a very low level,
and I am not aware of Emulators who provide this
publicly available on the Mac... the only one I am
aware of is the Emulator from Apple, and I do not
think it is available for licencing...

Steffen

[amigatv]

Thanks to all for coming - we had nearly 30 of you for the meeting, considerably more than is usual these days.  Some photos are here:

AMUSE - photos of OS4 meeting on Aug 5