[rescue] NeXTSTEP or OPENSTEP?

[Liam Proven]

On 9 August 2014 20:03, Lionel Peterson <lionel4287 at gmail.com> wrote:
> Releasing NT in SPARC would not have been a very good idea, it's success
would
> have relied on third-parties to cross-compile their applications to a new
CPU
> architecture for any real success, IMHO. Did Win NT on Alpha really amount
to
> much? Win NT/SPARC would have been a third platform for vendors to
support...

Oh, it probably would have bombed, yes, but the port was done, the
product was there -- just never publicly released.

It would not  have been the 3rd platform, by the way.

NT 3.1 (the 1st version) and 3.5 ran on x86, MIPS and DEC Alpha.

Clipper support was added -- but not released -- in the NT 3.5 timeframe:

https://groups.google.com/forum/#!msg/comp.sys.intergraph/Ym6j0XMrxoo/ncIxq2n
asi8J

NT 3.51 added PowerPC support.

The SPARC port was in the NT 3.51 time frame:
http://ftp.lanet.lv/ftp/sun-info/sunflash/1993/Jul/55.11-Sun-Intergraph:-SPAR
C-and-Windows-NT

So SPARC would have been the 5th CPU and 6th RISC platform for NT!

But I agree -- it would have been doomed; no RISC version of NT ever
thrived. The Alpha edition did best, partly because the Alpha was the
fastest of them, and partly because DEC implemented FX!32, meaning
x86-32 support. Base RISC NT only emulated 16-bit x86 for Win16 apps.


>> It required fancy high-end disks and
>> proprietary RAM and expansion cards for far too long, when it should
>> have been making PC-formfactor motherboards which could plug into all
>> the parts of a standard PC, so taking advantage of cheap PC
>> componentry.
>
> Well, the pedantist in me feels compelled to point out that. SCSI was
common
> on workstations (their intended market)

It was, yes, but IDE caught up sooner than most of the SCSI vendors
realised. It delivered more capacity and usable speed for a lot less
money. But then, all the RISC vendors continued to believe that people
would want $20K RISC Unix workstations even after $5K 32-bit PCs were
delivering decent performance.

> and us believe all the RAM Sun used
> was standard- based, they used obscure standards to be sure, but the
commodity
> SPARC boxes I'm aware of used standard form-factor DIMMs with fairly
standard
> technology on them, but no one else (AFAIK) ever adopted the same
combinations
> of standard DIMM form-factor and memory technology, effectively rendering
them
> proprietary, but not really.

Hmmm. OK. I didn't know that.

> Early Sun desktops used standard PC 30 and 72 pin SIMMs, but required parity
-
> an unusual, but not unheard of requirement at the time in the PC market.

Ahh, OK. I mainly have experience of very late-era ones, Ultras and so
on, when they were using modules utterly unlike anything in the PC
world.


--
Liam Proven b" Profile: http://lproven.livejournal.com/profile
Email: lproven at cix.co.uk b" GMail/G+/Twitter/Flickr/Facebook: lproven
MSN: lproven at hotmail.com b" Skype/AIM/Yahoo/LinkedIn: liamproven
Cell/Mobiles: +44 7939-087884 (UK) b" +420 702 829 053 (DR)