Thoughts on the A1Lite and mainstream markets

[Hammer]

I don't think you're going to be able to fit an Athlon64 plus all its cooling equipment into one of those neat little mini ITX cases.

The closest to ITX would be SFF. Note that they manage to fit in the cooling devices in an Athlon 64-M 3000+ laptops.  

[Hammer]

@GregS

You cannot make the Intel family run cool, it just won't do it,  

Recall Intel’s “Pentium M” for thin and light markets.

you also have all that legacy design, plus one way or another in the end the Intel type overclocked CISC design will have to go.
 

One of the key aspects of RISC is the fix length instructions.

Ever since the AMD's K6 processor, X86 variable length instructions is decoded/translated to fix length instructions.

X86 legacy support (a.k.a software investment protection) is actually the boat anchor for the current situation.  

As we all know Windoze is slow and unreliable and not likely to get better.

Relative to what?

Has it got a market, indeed the A1 and OS4.0 - well I think it has, history has a way of catching up - the thing to remember is that INTEL/WINDOZE IS A DEAD END

Realistically, it would take more than just A1 and OS4.0 to reenter the mainstream PC market i.e. software.

Both Microsoft and Intel still dominate the desktop PC market. Even without Intel there’s always AMD, Transmeta and VIA to take the slack. Microsoft still has non-X86 Windows editions for expansion.

“Software investment protection” is a powerful aspect for the current situation. Even the might of Intel has not significantly move the market to non-X86 solution.

What will be increasingly apparent is that in terms of a mass CPU market the PPC design is well ahead (not denying there are better CPUs) .

Note that PowerPC 970’s transistors count is similar to Athlon XP (Barton core).

Here the Linux family have the server side already established,

Mostly X86 Linux distros i.e. Linux X86 edition has the ability to leverage the influential “Software investment protection” aspect e.g. WINE/WINEX/VMWARE.

 MacOSx has a place,

Barely 2.5 percent of the desktop PC market.

[Hammer]

@Hammer - so let me summize - on OS which nobody loves (Windoze) which has never worked well ?

"Love" is not an intangible/rational factor.
To illustrate  an example, i.e. it works well enough and earning sustainable cash flow for www.locus.com.au**, solution6.com.au**, myob.com.au** and 'etc'.

and has always been sluggish,
 

Please quantify “Sluggish”.  

running on a hot-chip design is nevertheless unbeatable?

Note that MS Windows NT 4.0(up to SP3) runs PowerPC.

I just don't understand the logic, just where and how do you propose Windoze moves in the future - each move destroys its immense base.

In regards to “Windoze moves in the future - each move destroys its immense base”, have you tried MS Windows Longhorn** X86-32 build 40xx or have you tried MS Windows XP AMD64 Edition** (with WOW64)?

**Beta releases.

Within these product scope, the legacy (includes skills, tool chain, development and etc) is mostly preserved. Other Windows editions are an extension to the main product lines.

Note that MS has VirtualPC and FX32 style technology (via Intel) for future non-X86 Windows XP and Server editions.

Mac bit the bullet (finnally) and remerged as Mac OSx and in doing so fragmnted even its tiny market share - do you think MS could do a similar thing and not fall apart at the seams.

Unlike Apple, DotNET framework and VirtualPC technology is designed to unify the code base.  

The fact is that people have had a gut full of MS, anything that looks half good at the moment has a running chance

IF that was the realistic case then the OS/2 Warp 4.5 with Win32 compatibility(via third party add-on) may have a chance for a come back. But sadly, this is not the case.

in my opinion the A1 and OS4.0 look a lot better than Half-good!.

In regards to “looks”, an integrated DirOpus 5 with AOS4.0 style icon bar per windows would have been nice i.e. integrating some modern GUI elements.

Secondly, one of the main problems with AOS is its middleware support for application development.

From your comments it seems that you have the idea that things just keep on going on, that MS by some act of God has already planned an escape route and that the right time old Billy-boy will make his move

I’m still waiting for a creditable catalyst for change.

MS is dead - its just that it has a livily corpse.

Such statement is in the realms of "reality distortion field" (TM).  

To ram my point through;
 
Can AOS4.0 match Windows and Sound Storm/SB Audigy 2 ZS software suite?  
Can AOS4.0 match DirectX 9.0b?
Can AOS4.0 match Windows’s OBDC functionality?
Can AOS4.0 match Windows’s OLE functionality?
Is Oracle 9i available for AOS (middleware issues)?
Are there object oriented case tools for AOS?
Are there object oriented visual development environment for AOS (in the level Borland C Builder/J Builder/Delphi/Visual Studio dotNET 2003)?
Is there even a fully implemented Java VM for AOS?
Is there a Windows 2003 Server** level based on AOS? (**near brain dead server maintenance)
Can AOS deliver multi-user support?
Can AOS deliver roaming profiles?    
Can AOS drive wireless LAN?

Should I add some more?

Amiga doesn’t need to compete with Windows/X86 Linux i.e. Alan’s “leisure computing” targeted market should be satisfactory enough.

[Hammer]

Well obviously not. But is there any reason why this is fundementally beyond the Amiga?

In regards to light weight OS, one can apply a similar reasoning to Microsoft (e.g. Windows XP Embedded) .

Amiga’s “fundamental” issues are not the real issues. The real issues are the matter of economics (includes market*) issues.

*National issues in some countries i.e Japan, South Korea, China, European Union, Russia and ‘etc’ (the list is made up of basically the countries who missed out of  the Information Revolution).      
 

My point is that MS has no future despite its apparent strength.

“To control the future one has to control the present”.

 On the otherhand amiga is tight and light, robust and (hopefully) very stable.

Stability is usually proportional to the features it provides.

The future, recompiling the amiga on Taos, is something of a permanent solution for software development.

Do they have a reasonable level of Integrated Development Environment?

Do they have a reasonable level of middleware support and ecosystems?

Note that the distribution dotNET framework is now here. Could you point me to the download of runtime ecosystem for the competing product?

It may not be the Amiga, my point was that MS is in a historic fix and over the next period the opportunity to shove it into the dustbin is already here.

Define "here".  The market doesn’t like empty void.
The world operated without MS’s market domination a few years ago, thus there’s nothing new under the sun.

Well how about hate - that is how I would describe the MS experience in terms of those that have to use it

The some reasoning can be applied on all operating system products. No operating system is perfect. It’s a matter economics and dollar signs.

To illustrate an example with Linux i.e. dependency issues and flawed driver modelling. Needs a higher competency skills to maintain this particular solution.  

(I am talking of people who have never heard of the Amiga and barely know the Mac exists). Intangible, perhaps - irrational - I don't think so.

Have you done a statistical count? Have you done any SWOT analysis on other Operating Systems?

In regards X86 (due the inclusion of Intel into the discussion); does hate automatically equates to hatred of X86?

IF they hate the said product why don’t they buy/ switch completely to another alternative product? Nothing is stoping them in downloading/purchasing SUSE Linux, Lindows 4.0, Mandrake Linux, MorphOS, OS/2 Warp 4.5, BeOS, MacOS X, QNX, OpenBSD, NetBSD and ‘etc’.

Just don’t complain about the lack/little of vertical and horizontal market support, lack of driver support, lack/little of entrainment/application/development titles, incompatible hardware and 'etc'.

Changing the “real world” is a lot harder than changing the “reality distortion field”.

[Hammer]

China has not only rejected MS, but gone for a form of Linux and the PPC, Russia and Eastern Europe are almost Linux by default (server wise anyhow).

Such case is not abnormal since a couple of years ago government entities was with Unix. There’s still heavy piracy rates for Windows in the countries mentioned.

Now some statistics...

Refer
http://www.linuxdevices.com/news/NS7483572763.html

Far from missing the information revolution, they are in a position to reap much of the benefit without being caught up in the expensive MS merry-go-round (and it is very expensive, suit-driven and unproductive).

Open Source is a quick way to catch up with information oriented nations due to the free use of labour.

[Hammer]

As for the Taos solution, the plan, I believe is to develop AmigaOS up until not only is it all written in vanilla C but that it has precsiely the development tools, drivers etc that allow it to be verstile in the context of this next wave of technologicasl innovation ---- then I believe, the plan is to recompile the whole thing within Taos a processor agnostic system which has the virtue, because of level of abstraction, of keeping software permanently available regardless of processor technology or indeed OS development.

Such a solution will be competing with MONO**(OpenSourced dotNET) developments. **Available for Win32 and Linux.

Please get the big picture. For the next big move is not away from X86, but away from processor dependence. PPC offers a stepping stone no-more, and I would say the same for Linux.

For embedded processor statistics refer to
http://www.linuxdevices.com/articles/AT7301151332.html

"What we see here, is x86 in first place and ARM in second -- both growing, but with x86 maintaining a 9% lead over ARM. Incidentally, "ARM" includes Intel's highly popular StrongARM and XScale processors."

Rankings
1. X86, 1st position with 2nd growth rate.
2. ARM, 2nd position with 1st growth rate.
3. PPC, 3rd position with 3rd growth rate.

Current situation in X86 and ARM is closely linked to Intel’s market power**.

Just how do you read this as putting MS in a good long term position - I don't, hence I am happy to announce it a zombie (the walking dead).

Refer to http://www.linuxdevices.com/news/NS7483572763.html

That is; "VDC finds Microsoft, Wind River in "dead heat" as #1 embedded OS".  Windows CE is now an open source development…

Put all this together, the China move, the inroads Linux is making at the perifery, the growing prominance of PPC architecture (especially when linked to China),

China’s long term desire would be their MIPs based processor.

 the be-suited expensive and haphazard MS world in general, the squeezing out of small developers the lack of software innovation as a whole

MS has not squeezed out small developers i.e. refer to IE6’s third party licensees**…

Note that Apple has developed their own media player and web browser in their MacOS X. The only difference is that Apple’s OS not running the dominant HW platform.

**Help -> About Internet Explorer.

As for lack  innovation issue. Plenty of OS vendors would be guilty of this
1. MacOS X's kernal is BSD based.
2. MacOS's GUI is based on Xerox's WIMP concepts.
3. Amiga's Workbench is based on  Xerox's WIMP concepts.
4. Linux, *nix like OS for i386 - Paraphrase from Linus i.e. the purpose of Linux.
5. MorphOS, a clone of AOS.

Microsoft’s main innovation is opening up the HW PC market for clones.

"There is nothing it can do, no rabbits it can pull out of the hat, no matter how fast and small, that are not immediately bogged-down in the quagmire that is MS"

Refer to open sourced "Windows CE"...

[Hammer]

Besides which the cost factor is enormous, hot X86 chips in server farms means that the airconditioning costs alone far outwiegh the cost benefits on the chips themselves.(SNIP)

Tell that to IBM’s Power4. As for "thin and light" X86 refer to Pentium M blades....

[Hammer]

 don't know how to make this any clearer - China will not be going down any path that even gets them close to useing MS - they are not mad. If its X86 then it doesn't matter if it runs ice-cold and is fast as light - the Chinese will not have a bar of it.

Their action speaks for themselves in that regard e.g. AMD64 and MIPS and ‘etc’. I’m sure the Chinese can separate the HW and OS after all the Linux’s home HW is the X86.

It’s a matter of economics i.e. MS’s stuff is expensive but the hardware is not.  

Reduce the CPUs by this factor and then they either go for one of their own (I don't think this will be the case) or what?

Again, why did they invested on AMD64 based supercomputer?

But if your content to play selective quotes then go for it - say I contradict myself by also argueing that even though a X86 may be cooler it still is not in the running.

Refer to the statistics and trends.  

The Chinese have made it pretty clear that they do not want, nor can they afford, to go into MSes pocket. Using any form of X86, in short anything compatible to MS would in effect give their technological policies over to Billy Gates - and they ain't going to do it.

Again, I’m sure the Chinese can separate the HW and OS issues. The Chinese is not that stupid. Its wouldn’t take much for MS to port a Windows NT variant for PowerPC platform i.e. the Windows NT code base for PowerPC is already finished. A new HAL would be needed for Windows NT 4.0 to work on the current PowerPC based platforms. Another MS’s service e.g. WDM (part of NT5.X), DirectX and dotNET can be back ported to that HW platform. PowerPC wouldn’t be the insurance to keep MS at bay (refer to X-BOX 2).

Open sourced Windows CE for PowerPC would be another possibility.

Clues for the future PowerPC/X-BOX 2 and more Windows CE/PowerPC info refer to
http://www.microsoft.com/windows/Embedded/ce.NET/evaluation/hardware/processors.asp

A game bias Windows CE variant was running on Sega Dreamcast’s SH4 processor.

Sources for Windows CE and PowerPC tool kits
http://www.apogee.com/windows_ce_toolkits.html  

http://www.microsoft.com/catalog/display.asp?subid=22&site=763
; Microsoft® Windows® CE Toolkit for Visual Basic® 6.0 (PowerPC support is included).

[Hammer]

coldfish wrote:
On the issue of  Windows' and the x86 ISA's future?

 Well, looking at M$'s intentions to use IBM processor tech in Xbox2,  perhaps M$ will use Xbox2 to leverage a position in the PPC market?  More likely though, is that PPC fits better into console architecture.

In the near future, I wouldn't be too surprised if the home PC as we currently know it doesnt exist. Replaced by simpler "black box" hybrid console systems.  
Given the choice, the average person would opt for a very low cost game/communications/entertainment hub that just works, over current, more tempramental hardware solutions.

Just ask Sony and M$, they're both onto the idea...

Note that Windows CE is already running on low end PowerPC processors. We also know that Sega’s Dreamcast was powered by a game tweaked Windows CE (with PowerVR GPU/~DirectX6**) on SH4 platform.
**or equivalent to DirectX6.

Reference;
http://www.microsoft.com/windows/Embedded/ce.NET/evaluation/hardware/processors.asp#PowerPC

One can see the potential of the X-BOX 2 on the PowerPC. There’s nothing new in MS OS running PowerPC**.  **Not factoring Windows NT 4.0(SP3) PowerPC Edition.

[Hammer]

@GregS

Some statistics regarding Windows, Linux and X86 refer to   http://www.etmag.com/publication/magazine/2003-10/27-1.htm


"X86 server sales drive Western European server market"

"Both Windows and Linux continued to grow strongly - benefiting from their association with the x86 server market. Windows continues to command the lion's share of the unit market with 59.3%, while Linux has risen to 16.3% share".