@JimIgou

Quote:

I haven't been impressed with Apple since the days of the Apple II when a machine that was based on a $5.00 processor could cost upwards of $2000 or more.

LOL! Wasn't that $25 a piece for a 6502?

Regardless you make a good pont. This cancels out any old canard about Apples being expensive because the CPU cost so much.

Later they moved to 68K and then they controlled the PPC market. Which likely would have reduced the CPU price. Though Motorola still had some kind of monopoly on PPC at the time. Apart form IBM.

Quote:

Tabor does decrease the likelyhood of a 64 bit Amiga OS, and its slow.

I read it has a 36-bit address bus of sorts.

Quote:

Also, this is an open platform (unlike X64). Want to license it? IBM is supporting third party manufacturers.

Most of us wouldn't care if it's open or not, we just want it in available quantities and on an available motherboard for a resonable price.

Quote:

An eight core/32 thread Blackbird costs little more than a 2core/2 thread X5000, and is SO much more powerful.

Okay, so one core would have four hardware threads? Yes, OS4 running on four hardware threads, would be more powerful than just running on two.

Last edited by Hypex on 08-Jan-2019 at 02:18 PM.

@NutsAboutAmiga

Quote:

You can't simplet compile it for little endian and expect it to work. nothing will work.ÜÜÜ

It would be userful if the "-mlittle-endian" switch in GCC actually got it to compile code for little endian. I don't know what it does but it would be useful if did what it implied and used the reversing instructions on memory read/writes. I tried it once but didn't see any difference when running some code. Should have compared the ASM output. Of course little endian on PPC means it is running in LE mode from the get go. So all code regardless is running as LE.

If the compiler could be told to use reversing instructions on memory read/writes this would make porting LE code a lot easier. Provided it is isolated to code modules running in a closes system it should work. Parameters are passed in registers which by their very nature are endian agnostic. All it needs to do is carry out its work, while calling any external C functions with registers only, and only reading/writing memory words isolated to that one module.

Would be a lot easier than fixing up LE only code. Including finding where the breaks are. Or wrapping a BE macro around every pointer reference to compensate.

@Kronos

Quote:

The ST was to 68k what the Amstrad/SchneiderCPC was to the Z80: Just enough off the shelf HW to make it an attractive option in the lower range of the markets.

Bingo.
My main concern was that it had a 68K cpu, and not a crippled 68008.
And I only ran the Atari supplied OS and software to use commercial software.
Most of our software was still text oriented at that point, although we did eventually port a GUI to the alternative OS we used (one that Gespac had been using created by a programmer named Steve Adams).

I'm not that much of an Amiga chipset fanatic.

Just a Motorola fanatic.

VGA graphics were better than AGA, Paula is only 8 bit while Soundblaster was16 bit, and I've never liked SCSI (and even our early 68K systems, outside of the ST, were IDE).

Amiga was very original, the first consumer oriented multi-media computer.

Pity they didn't devote more funds to further development.

@Hypex

Quote:

Okay, so one core would have four hardware threads? Yes, OS4 running on four hardware threads, would be more powerful than just running on two.

Actually...the real comparison is 2 threads on Tabor or the X5000/20, four threads on the X5000/40, eight threads on Roberto Innocenti's proposed four cour T2080 based laptop, or sixteen threads on a four core Power9 based Raptor Blackbird (OR thirty-two threads on the eight core variant).

And no, the 6502 did not cost $25, it cost $5.
Which is why you saw it used more often than the much better 6809, which cost about $9.

Also Apple was SO cheap, they bought their floppy drives for the Apple II without the controller board usually mounted on the back of such drives and had Woz create a simplified replacement they could make on their own.

BTW - I am NOT anti-Aeon. It's SO nice to be able to buy new hardware that will run MorphOS (and for that matter OS4). The old crap is failing, be it from Eyetech or Crapple.

Last edited by JimIgou on 08-Jan-2019 at 03:13 PM.

As if by design, the irrelevant and obsolete are presented in this thread as some hot topic with any meaning, discussed by those who are just doing their same old same old.

You, whatever your current name is why do you even come to a thread like this is if to you this is about a processor? not having enough peeps to argue with on Morph Zone? Not only are you talking about some bogus preference to a hardware that died which no one gives an eff about, you do it in a place where some still care about a certain hardware and software connection which obviously doesn't say anything to you. Why don't you go look at some processor charts or something? Ahhh Jimbo?

@Srtest

Interesting, but irrelevant.
I never left the market, but if I did,it would be because of the negative attitudes and psychobabble coming from a few of the remaining posters like you.

Funny, I just talked to Bill Buck via cellphone a few days ago, and he is still saying "If everybody could get on the same page...".

Never happening.

So, if YOU'RE pissed off, why don't YOU leave?

I still contributing, and its in more than just words.

@JimIgou
The rifts in the Amiga community have very little to do with posters who are in it to slag.

We can't all get on the "same page" as there is a lack of agreement on the fundamental issues:
1. CPU architecture - You have the Power ISA holdouts, the x86-64 pragmatists, the ARM futurists, and the 68k in FPGA romantics.

2. Hardware architecture - Custom hardware vs. off-the-shelf components vs. SoC boards vs. Other OEM's ready-made systems

3. OS - Hyperion vs. MorphOS vs. Classic OS vs. AROS

4. System architecture - Backward compatibility with 68k apps and games should not be broken vs. Catch up with the rest of the industry and sacrifice compatibility

5. Market - People who want a high-end Amiga, people who want a hobby machine, people who want a nostalgia fix

A hell of a lot more money than all the Amiga related companies have put in over the past two decades is required, in order to create a single page for most of that mess to get on.

_________________
All the way, with 68k

Quote:

JimIgou said: VGA graphics were better than AGA,

I think your memory is a bit selective. Yes, PC hardware eventually surpassed the Amiga, but not until several years after the A4000 was released. SVGA was better in some ways, but VGA wasn't. When the A1200 came out, popular 386SX PCs with ISA VGA cards were slow as molassas in Windows 3.1.

Quote:

Paula is only 8 bit while Soundblaster was16 bit

Sound Blaster was crap. Fewer D/A channels, limited DMA capability, couldn't loop samples properly - and that's if you could get it to work with your game at all! (clones were even worse). Most music was synthesized and sounded like a cheap Yamaha keyboard.

Quote:

and I've never liked SCSI (and even our early 68K systems, outside of the ST, were IDE).

Later Amigas had IDE too. But when the first Amigas came out there was no such thing as IDE. SCSI was much better than the alternatives, and the only choice for external drives (plus other devices such as scanners etc.). At that time I was maintaining PCs with ST506 interface drives, and they sucked! The first 'IDE' drives had an 8 bit proprietry interface, and they sucked too! Later on we had lots of trouble when IDE drives changed to a 3.3V interface - and wouldn't work with some motherboards! Then there were flaky VL bus cards, driver incompatibilities, confusion over drive geometries etc. Man, those were frustrating times for PC assembers...

@JimIgou

Quote:

VGA graphics were better than AGA, Paula is only 8 bit while Soundblaster was16 bit, and I've never liked SCSI (and even our early 68K systems, outside of the ST, were IDE).

I wouldn't say VGA was better than AGA. The only thing VGA had above AGA was a chunky mode that AGA should have built into the Amiga chipset. Plus text modes and native monitor support.

But it was limited to 18-bit RGB colour palette. 256 max colours. No copper. No sprites. No dual playfields that I know of.

SVGA, when it came about, improved things slightly. The SB might have been 16-bit but did it have multiple hardware channels with variable volume and frequency control? The sprites of the sound world.

As to IDE, in the Amigas time, did it have DMA support? SCSI always had DMA on my Amigas. Even today the latest IDE cards for the Amiga are still old fashioned; I have never seen one with DMA. And IDE is limited to two devices per bus. SCSI could do 8 devices (including controler.)

Quote:

And no, the 6502 did not cost $25, it cost $5.

IIRC I get the $25 figure from a Commodore book. But it also turns up in a Google search. If it is incorrect then a lot of sites need to be fixed.

http://www.google.com/search?q=6502+price

@Hypex

Their own advertising is propably wrong too.

Quote:

Hypex wrote: I wouldn't say VGA was better than AGA. The only thing VGA had above AGA was a chunky mode that AGA should have built into the Amiga chipset. Plus text modes and native monitor support. But it was limited to 18-bit RGB colour palette. 256 max colours. No copper. No sprites. No dual playfields that I know of.

AGA looks better than VGA but it was dreadfully slow because C= used the slowest chip memory possible until the end. Jay Miner proposed using VMEM (dual ported chip memory has twice the bandwidth) in the Ranger chip set before AGA even came out. Even AGA memory bandwidth improvements were minimal and cheap. Faster chip memory could have been used in the high end Amigas followed by upgrading the low end machines to that spec when the high end is upgraded again. C= failing to upgrade the CPU performance made matters worse as gfx performance is dependent on CPU performance (C= accelerators usually had the slowest memory also). Doom would have run adequately on AGA with better chip memory and CPU performance (FPGA AGA Amiga hardware can have several times the chip memory and CPU performance today). Chunky wasn't that much of advantage over planar until 16 bit chunky when the quake in the Amiga market resulted in its doom.

Quote:

As to IDE, in the Amigas time, did it have DMA support? SCSI always had DMA on my Amigas. Even today the latest IDE cards for the Amiga are still old fashioned; I have never seen one with DMA. And IDE is limited to two devices per bus. SCSI could do 8 devices (including controler.)

Amiga IDE wasn't even buffered until the Amiga 4000T. This meant a slow CPU was overworked servicing the IDE interrupts. C= failing to upgrade the Amiga hardware fast enough resulted in the hardware turning from their biggest asset to their biggest liability. The "standard" Amiga hardware spec stood still while the Moore's Law PC world took off.

Last edited by matthey on 12-Jan-2019 at 08:50 PM.

@Hypex

Quote:

The only thing VGA had above AGA was a chunky mode that AGA should have built into the Amiga chipset. Plus text modes and native monitor support.

And the undocumented, non-BIOS Mode X with square pixels, one byte-per-pixel and page flipping.

_________________
"Unix is supposed to fix that." -- Jay Miner

Quote:

matthey wrote: C= failing to upgrade the Amiga hardware fast enough resulted in the hardware turning from their biggest asset to their biggest liability. The "standard" Amiga hardware spec stood still while the Moore's Law PC world took off.

It wasn't Moore's law that killed the Amiga, but simply that it never had sufficient market share to drive the technology. Even if Commodore had managed to stay on the 'bleeding edge' it wouldn't have mattered. The Amiga had a fundamental flaw that could not be fixed - it wasn't a PC.

Not that they didn't try. Remember the Sidecar? An IBM XT clone in a box, and it sucked. Then 'big box' Amigas with ISA bus slots, and various Bridgeboards to make use of them - which also sucked. Finally they gave up and just made straight PC clones (which still didn't save them). For the Amiga to be truly competitive Commodore would have needed to give it:-

1. PCI slots.
2. An Intel CPU
3. Microsoft Windows.

Nobody could stem the tide of PC clones, not even IBM itself. Remember Micro Channel? That was when IBM tried to take back control of the market - and failed.

Moore's Law says that the number of transistors that can be put in an IC doubles every 2 years. OCS Agnus had ~21000 transistors. The AAA chipset (prototyped in 1994) was to have 1 Million transistors, which matches the increase predicted by Moore's law. But Commodore went bankrupt before they could get AAA out the door. Was it because PC hardware was so much better, or did people choose them for some other reason?

I was selling Commodore 386sx-16 PCs alongside an equivalent A1200 package for the same price. The PCs came with Windows 3.0, 2MB of RAM and a crappy standard VGA monitor. Compared to the A1200 they sucked, but people bought them because Microsoft. Those machines are actually quite rare today, probably because most them soon got tossed in the trash.

@AmigaBlitter

The only way to "save" somthing was to build up a good AmigaHW emulator and someway develop OS on it...

No new real HW was usefull imho...only software development...hw wothout software was USELESS!

So better good OS/softrware on emulator imho...

Last edited by Tuxedo on 13-Jan-2019 at 12:45 PM.

_________________
Simone"Tuxedo"Monsignori, Perugia, ITALY.

@bhabbott

Quote:

The Amiga had a fundamental flaw that could not be fixed - it wasn't a PC.

That could be argued for the USA but the rest of the world WOULD have accepted another home computer or better business machine in 1985-1987. With an AA Amiga in 1989 and AAA in 1992 this could even have gained traction in the US home computer market IMHO. PCs whatever the suits thought were rubbish until Windows 98 and then still behind the Amiga (for home use) until Windows XP.

_________________
"Art challenges technology. Technology inspires the art."
John Lasseter, Co-Founder of Pixar Animation Studios

@bhabbott

Quote:

Moore's Law says that the number of transistors that can be put in an IC doubles every 2 years. OCS Agnus had ~21000 transistors. The AAA chipset (prototyped in 1994) was to have 1 Million transistors, which matches the increase predicted by Moore's law.

Erm, no it doesn't

You compare Agnus to the whole AAA chipset, what about Denise and Paula? Maybe even have to add Gary and the CIA to that count....

And then there is the fact that it not just about transistor count but also how effective they can be used.

_________________
- We don't need good ideas, we haven't run out on bad ones yet
- blame Canada

Quote:

bhabbott wrote: It wasn't Moore's law that killed the Amiga, but simply that it never had sufficient market share to drive the technology. Even if Commodore had managed to stay on the 'bleeding edge' it wouldn't have mattered. The Amiga had a fundamental flaw that could not be fixed - it wasn't a PC.

I agree with your point about the Amiga having insufficient market share to "drive the technology". C= could have adopted good standards like PCI (Dave Haynie wanted to as once again the developers had a better vision for the Amiga than the management) and cheap standards like IDE to reduce costs (C= messed this up by not using IDE sooner in the Amiga 500, buffered IDE and cheaper 3.5" IDE drives in some Amigas). The Amiga would likely have fallen behind the capitalistic competitive force of the PC with Moore's Law but I believe C= could have survived with competent management. C= became profitable under Thomas Rattigan but he probably cut too deep adversely affecting future R&D. He was smarter than former and later CEOs at C= though and may have been able to keep it profitable. Cutting costs and getting a cost reduced Amiga (500) out were important priorities. I would have tried to keep the original Amiga team (Commodore had visionaries Jay Miner and Chuck Peddle but how much did upper management listen?) and upgrade the Amiga 2000 with 68020 and Ranger chip set even if it caused some delays. Management kept canceling important R&D projects and only managed minor improvements (ECS instead of Ranger, late AGA instead of AAA, etc.). Management also failed to make important deals including licensing PA-RISC for Hombre, licensing the AT&T DSP technology, creating perfect Amigas for the Toaster, licensing the 68k for a SoC, licensing the Amiga for embedded use, licensing the Amiga for clones, etc.

Quote:

Not that they didn't try. Remember the Sidecar? An IBM XT clone in a box, and it sucked. Then 'big box' Amigas with ISA bus slots, and various Bridgeboards to make use of them - which also sucked. Finally they gave up and just made straight PC clones (which still didn't save them). For the Amiga to be truly competitive Commodore would have needed to give it:- 1. PCI slots. 2. An Intel CPU 3. Microsoft Windows.

C= could have upgraded the Amiga 2000 with a 68020 and created optimized 8088 and 6502 emulators instead. These CPUs could be emulated at close to full performance and the extra performance could have been put to good use when not needing 8 bit compatibility.

Quote:

Nobody could stem the tide of PC clones, not even IBM itself. Remember Micro Channel? That was when IBM tried to take back control of the market - and failed.

True. Even IBM could not regain control of the PC market they had created by being too open because of the U.S. DoJ breathing down their necks. Ironically, the PC was born because of IBM's good reputation which didn't matter shortly later.

Quote:

Moore's Law says that the number of transistors that can be put in an IC doubles every 2 years. OCS Agnus had ~21000 transistors. The AAA chipset (prototyped in 1994) was to have 1 Million transistors, which matches the increase predicted by Moore's law. But Commodore went bankrupt before they could get AAA out the door. Was it because PC hardware was so much better, or did people choose them for some other reason?

It's a moot point as C= couldn't get major hardware upgrades out the door.

Quote:

I was selling Commodore 386sx-16 PCs alongside an equivalent A1200 package for the same price. The PCs came with Windows 3.0, 2MB of RAM and a crappy standard VGA monitor. Compared to the A1200 they sucked, but people bought them because Microsoft. Those machines are actually quite rare today, probably because most them soon got tossed in the trash.

The PC market quickly became saturated and unprofitable for all but the biggest players. C= had a much better hand with the Amiga than PC clones but blew it.

@hth313

FFS was an absolutely horrible filesystem. The by far most fragile and slowest of them all. I just cannot believe some people are still using it on NG Amigas!!

I switched to the journalling AFS->PFS->SFS as fast as I could and never suffered any data loss since. I even had some massive corruption dure to hardware failure. The one FFS partition I had got so broken it dissappeared completely. Could never be roecovered. On the four PFS-partition I lost 3 files in total. Otherwise all good.

_________________
- Don't get fooled by my avatar, I'm not like that (anymore, mostly... maybe only sometimes)
> Amiga Classic and OS4 developer for OnyxSoft.

To be perfectly honest, I never have nor will I ever "leave" the Amiga. Yes, I may only turn on my Amigas a couple of times a year, and for the rest of the time I'm using a modern smartphone, desktop, laptop, and gaming console as part of my day-to-day computing and gaming wants and needs.

My use of other platforms is not a de facto endorsement of those platforms. These are tools for being productive in the modern landscape. Windows is still the most dominant consumer and business OS, macOS has grown in business acceptance over the past decade and is also used by a lot of developers and content creators. I'm typing this on macOS at work. The same machine we use when we are live-streaming a webinar with up to 3 1080p camera feeds, mixed audio from lapel mics, with chroma key, and switching with slides and overlays using mimoLive.

When I use my Windows 10 VR gaming rig to jump into a virtual game space with my brother who lives 900km away, or when I'm designing web sites on my macOS desktop or laptop, I constantly find things that I dislike about those platforms and ecosystems. They certainly have their issues, but currency is not one of them.

For me Amiga is more than a fond memory. I relish the idea that in the not too distant future I will have made the kind of money that is required to give the Amiga computing philosophy one more attempt at achieving currency.

_________________
All the way, with 68k

@Kronos

Quote:

Kronos wrote: You compare Agnus to the whole AAA chipset, what about Denise and Paula? Maybe even have to add Gary and the CIA to that count....

Unfortunately I couldn't find any information on transistor counts for the other OCS chips. However I suspect they are much lower than Agnus (certainly the CIA count would be insignificant).

Quote:

And then there is the fact that it not just about transistor count but also how effective they can be used.

Agreed. But then the Mooore's law argument looses weight anyway, so...

One argument that some Commodore bashers often make is that the Amiga's chipset was too tightly integrated, which made it harder to advance. IOW, it was too good! From a certain perspective they are right. Early PC chipsets were poor to non-existent, so they had less 'baggage' to carry forward.

But the lack of a good base standard for PC hardware created massive compatibility issues, made life difficult for developers, and fragmented the market. Remember when PC games started coming out that demanded a fast CPU, VGA graphics and a '100% Sound Blaster compatible' sound card, limiting their sales to a tiny fraction of PC installations? PC developers did that because supporting a miriad of different hardware variations was just too difficult. They got away with it because the PC market was so huge, but this would never have worked for the Amiga. Having a constantly changing hardware spec would have fragmented the (already much smaller) market too much, as well as upsetting users.

I am glad that Commodore didn't keep bringing out new models with more advanced chipsets, because it means there are now only a few different hardware variations to cater for, and compatibility between chipsets is very good. Had they kept going we probably would have seen the Amiga go the way of PCs, with new 'Amigas' bearing little resemblance to the original design.