Interview with Trevor Dickinson

[Karlos]

Especially in the UK, where it is widely acknowledged that house prices have risen disproportionately in recent decades.

A different indicator of value is called for, I think.

[Karlos]

Anyway, a nice interview. That's quite a collection Trevor has :-o

[Karlos]

At the time of writing the kernel is up and running and the A1-X1000 can already boot up the AmigaOS 4.1 Update Workbench/install CD. This is the first time that an AmigaOS has run on a 64-bit Power CPU.

Running on a 64-bit processor is not quite the same thing as running 64-bit natively. I'm assuming that this is still a 32-bit build of the OS?

I wonder if a PAE kernel is in the pipeline to bridge the 32->64-bit gap? OS4 native apps that rely on MEMF_PRIVATE for most of their needs could presumably benefit from it.

[Karlos]

AFAIK PAE is something x86 specific.
Staf.

I did wonder about that, I thought it might have been a "general" method of large memory model support for 32-bit systems. It's been a while since I've looked into this sort of stuff.

[Karlos]

I'm presently facing the possibility of being made redundant in the near future, so I won't be buying any new hardware until that's clearer.

[Karlos]

@Karlos

Understood, but the point is, is this a lot for a high end computer.

http://store.apple.com/uk-business/browse/home/shop_mac/family/mac_pro

From £1940, I know which I would prefer and it is not named after fruit.

Quad-Core

One 2.66GHz Quad-Core Intel Xeon "Nehalem" processor
3GB (three 1GB) memory
640GB hard drive
18x double-layer SuperDrive
NVIDIA GeForce GT 120 with 512MB

That's disgraceful. Aside from the CPU/memory specs, my current PC stands up well to that and cost considerably less almost 2 years ago.

Comparable specs, when originally built:

2.66GHz Quad Core Q9450 processor
4GB DDR3 (dual channel rated 6-6-6 at 1600MHz, running at 1333MHz (CPU FSB) with 5-5-5 timing)
500GB SATA 2 HD
24x DVD RW (not blu ray though)
NVidia GTX260 with 896MB

Total cost was just over half the cost of the above Mac and that includes the Coolermaster Cosmos-S case, 850W CoolerMaster ultra-high efficiency modular PSU, Gigabyte GA-X48T-DQ6 motherboard (intel express X48 chipset, 2x PCIe 16 - both capable of full speed operation when occupied simultaneously) as well as the fact the CPU and graphics card were virutally fresh out of their respective stables at the time. Never mind the cost of DDR3 then!

Since then, I've replaced the gfx card (the hazard of buying bleeding edge) and added more storage and it still doesn't put much of a dent in the remaining price difference.

[Karlos]

The truth is that Windows, Mac OS X and even Linux beat the crap out of any NG amigoid system in any given task.

Are you sure you didn't you mean:

The truth is that Linux, Windows and even Mac OS X beat the crap out of any NG amigoid system in any given task.

Sorry, couldn't resist.

[Karlos]

Talking about the justification of the price tag is a bit misleading. Unless someone is trying to assert that the manufacturer of the system will be making a huge profit on each unit sold, there isn't really much to justify, only something to explain. A minority platform using non-mainstream hardware is subject to cost escalation that common mainstream hardware is not. With those factors taken into account, the system will cost whatever it costs.

Precisely why, I might add, the cost of the Mac system above does need justifying since it is now using "mainstream" hardware. The Xeon processor is phenomenally expensive for x86 hardware and, unless you are running a server, it's hard to see what specific advantage it gives you for desktop computing over say a Core 2 / Core i7, both of which are significantly less expensive.

[Karlos]

Also, the initial OS4 will use only one of the cores, and is likely to run in 32bit mode, and there won't be 3d acceleration for the included ATI card. That's somewhat of a bummer if you compare it against the Mac MorphOS offerings.

That presupposes that the above situation will not change. I would be surprised if someone had gone to the effort of building new hardware, in a market as fragile as this one, without making sure the hardware could be properly utilised by the OS either on release, or without a serious feasibility study as to whether or not the OS can move in that direction.

You can cite the USB as a precedent, of course, but underutilizing USB is a far less serious waste of resources than running a 32-bit only uniprocessor OS on a dual core 64-bit CPU with 4 DIMM slots available to it. At least one of those issues has been mentioned previously (SMP).

[Karlos]

They'll easily outlive any custom amiga HW offerings. Try to find AmigaONE from Ebay. Now try to find Mac Mini G4 from Ebay.

In all fairness, this isn't really about hardware longevity as it is numbers, is it?

You can find a deluge of second hand PPC macs simply because they were turned out in far larger numbers (than any PPC amiga hardware) to begin with and then rendered obsolete amongst their user base overnight by an architecture switch.

In comparison, PPC-based amiga machines (A1 - all flavours, Peg, Peg 2, Efika, Sam etc) have been produced in much smaller numbers. Most of the people that have bought them tend to keep them and whilst there has been a fair attrition rate on some of the old A1 / Peg 1 machines, it's probably fair to say most of them are still working. They just aren't for sale.

[Karlos]

I wont buy MorphOS on dead end hardware. The Mac people would just laugh at me taking their scraps.

This parasitic route is not the way for me.

Just as with the the "I bet they'll sell a dozen units" type remarks we've heard, I think that's equally a bit uncalled for. I think you are really trying to say is that you wouldn't buy MorphOS, period. Without a hardware vendor, releasing the OS for PPC Mac was probably the smartest move available for a software team. Either that, or go x86, for which there already is a free alternative.

[Karlos]

Following this argument you won't buy OS4 on X1000 either, as it's even more dead end.

Can you even have degrees of dead-endedness? If you take it as meaning "lacking opportunities for development or advancement", then something is either dead-end or it isn't *shrug*

As long as a piece of hardware is in production, it has the potential to be developed further.

PPC Macs are not in active production or development. Viewed purely as a piece of hardware, it is, according to the above definition, dead-end. As are the old A1, Pegasos et al, not to mention classic Amiga models. The only opportunity for any form of advancement they all have are the ongoing development of software (including OS).

I have several dead-end machines and their dead-endedness has not proven to be a barrier to having any fun with them

I'd be quite happy to own more, finances permitting

[Karlos]

Change "elitist" to "religious" and I'll agree with you both 100%, instead of just 99.9% . The Amiga is a religion anymore, with acolytes, heretics and a few agnostics here and there; the pragmatists all fled a long time ago...

I don't quite know where I fit into that. I have a fairly modern PC (2 years old, nearly, but a decent enough spec), I use other OSes, but I still retain several classics, an A1 and I was hoping to get a PPC mac for MOS2 and contemplating this X1000 machine too.

Alas, the old job security took a kick lately and such plans are now on hold for the moment.

[Karlos]

But it isn't a classic Amiga, it runs classic software through UAE, the same way a Mac or PC does.  Only it is far more costly than either of those....

Only the last part of your observation is correct. The first part is not. Neither OS4 or MOS run classic software in the same way a "Mac or PC" running UAE does. In both OS4 and MOS, system-friendly 68k apps run translated within the host operating system, wherein OS calls are ultimately routed to native code. Furthermore whilst UAE exists for both platforms to run software that does require classic hardware emulation, even that has been integrated rather more seamlessly of late. With software like RunInUAE you can just run such a legacy app from the workbench and have it pretty much transparently open in UAE.

[Karlos]

I like PPC for the same reason I like most things: variation. I don't regard it as a "must-have". As long as it has the performance to run old apps "better" and new ones adequately I'm not really bothered if there are x86 processors that are far faster. I can, and do, use those just as happily with a different OS.

What I find rather dull is the thought of the day when there is only one hardware platform left and only a choice of OS to differentiate systems.