Tidbits 02/19/93
Dov Bai-MSI Visitor
>[InfoWorld; February 15, 1993; Page 1,83]
>WFW SELLING SLOWLY DESPITE MARKET BLITZ.
^^^^^^^^^^^^^
>Volume falls below all expectations.
>By Stuart J. Johnston And Ed Scannell
>
>Despite a massive marketing blitz and bundling deals with many
^^^^^^^^^^^^^^^^
>major PC OEMs, sales of Windows for Workgroups have fallen well
>below even conservative expectations, say major retailers and
>resellers.
I wonder what all those people who attributed MS hugh success to
marketing techniques have to say about that.
Dov
Peter Mayne
In article <1993Feb20....@fcom.cc.utah.edu>, br...@jensen.cs.utah.edu (Brian Sturgill) writes:
>There is no reason to change the byte ordering of the chip except
>to run NT, said Michael Slater, editor and publisher of
>_Microprocessor_Report_, in Sebastopool, Calif.
>
>"HP understands the need to have a mainstream operating system
>running on their microprocessor to proliferate it," said Ken
>Lowe, an analyst with Dataquest, in San Jose, Calif. "That's
>where the critical mass is."
>
>However, HP denied that it had firm plans to run NT on the chip.
>
>"We have designed the chip to support \[PC operating systems such
>as NT\] if people want to use those in the future," Logan said.
>"But we've made no commitment to do that."
>
>The 7100LC chip will not begin appearing in products until 1994,
>he said.
>
>[
>I wonder if this is the first time hardware has been designed to
>fit the OS?
Actually, it isn't.
VAX/VMS
OS/400 -> AS/400
SPARC for UNIX (and most RISC CPUs, for that matter) (but not Alpha 8->)
8086 (for some kind of CP/M compatibility)
The CRISP processor specifically designed for C (see a recent Byte)
>BTW, while NT isn't on an architecture with reverse byte order, it
>was said at the NT developer conference that this is not a
>requirement... however, for the sake of compatibility of data
>files, and DOS and 16-bit Windows compatibility boxes, the byte
>order change is probably essential.
"reverse byte order": now there's a little-endian point of view that
Sun and Apple might want ot argue with.
>]
>Brian
PJDM
--
Peter Mayne | My statements, not Digital's.
Digital Equipment Corporation |
Canberra, ACT, Australia | "AXP!": Bill the Cat
Steve Loughran
> In article <1993Feb20....@fcom.cc.utah.edu>, br...@jensen.cs.utah.edu (Brian Sturgill) writes:
> >
> >The 7100LC chip will not begin appearing in products until 1994,
> >he said.
> >
> >[
> >I wonder if this is the first time hardware has been designed to
> >fit the OS?
> Actually, it isn't.
> VAX/VMS
> OS/400 -> AS/400
> SPARC for UNIX (and most RISC CPUs, for that matter)
> 8086 (for some kind of CP/M compatibility)
> The CRISP processor specifically designed for C (see a recent Byte)
Peter is right in that RISC CPUs are normally optimised for the code which
is to be run on it, and given the amnout of CPU time the OS invariably
consumes that the OS must be used to determine which instructions
to provide[*] .Also things like memory protection models ought to match,
otherwise you get CPU designers coming up with bizarre segmented memory
architechtures when all us hackers want is a flat memory model ;-)
>(but not Alpha 8->)
You mean no-one had a look at the instruction usage of any unix before
designing the architecture? Surely they didnt just use VMS.
> The CRISP processor specifically designed for C (see a recent Byte)
Actually, according to some AT&T reps PenPoint was being rewritten to
work optimally with CRISP...the register window/stack approach means
that deep call nesting requires more memory access than shallow nesting.
In my experience Object Oriented programming and UI's tends towards
deep nesting of messages/methods/function calls...
What is interesting is that I am not aware of any CPUs which have been
optmised for NT yet. Ok, it's early days yet, but the performance of NT
and its apps may differ from that of Unix of the same platform merely
because they weren't designed for each other. Maybe Pentium will have
some suprises.
Currently video cards optimised for Windows 3.1 are sexy: does anyone
know how well S3 cards perform under OS2 or Unix/X11?. I'm curious because
if the trend for OS optimised hardware continues there may not be a choice
of which OS to run on your PC: only the OS which the manufacturer thought
you would use will work at a speed which is barely tolerable...
Oh yeah. MIPS processirs have had two endian modes for a long time. Something
to do with VaxStations I believe.
Steve
[*] cynics may argue that in fact most modern RISC CPUS are optimised purely
for impressive benchmark figures.
-------------------------------------------------------------------------
Just when you thought real mode was dead they invented palmtops...
-------------------------------------------------------------------------
Brian Sturgill
Unix groups unite against NT.
Users undeterred by high cost of Windows move.
HP redesigns their RISC chip for NT.
Windows NT to cost $99.
-------------
[InfoWorld; February 15, 1993; Page 1,83]
WFW SELLING SLOWLY DESPITE MARKET BLITZ.
Volume falls below all expectations.
By Stuart J. Johnston And Ed Scannell
Despite a massive marketing blitz and bundling deals with many
major PC OEMs, sales of Windows for Workgroups have fallen well
below even conservative expectations, say major retailers and
resellers.
At its October release, Microsoft Corp. officials confidently
predicted WFW would become by far its best-selling version of
Windows within a year.
...
Researchers interviewed last week pegged monthly sales of WFW at
40,000 units a month maximum, with most estimates ranging from
25,000 to 30,000 -- which by Microsoft product launch standards,
borders on failure. Those figures are a far cry from the 1.2
million per month run rate the company has claimed for its
popular Windows 3.1.
...
Nonetheless, retailers, ISVs, and even many users have not
written it off as dead -- especially given Microsoft's penchant
for sticking with slow sellers until they succeed.
...
Still, confusion with the company's offerings -- especially
Windows 3.1 and Windows NT -- may stifle sales, some users said.
...
-------------
[InfoWorld; February 15, 1993; Page 3]
UNIX GROUPS UNITE AGAINST NT
Unix International backs OSF's distributed processing
By Doug Barney And Ed Scannell
The nation's two leading Unix consortia, fearing that Microsoft
Corp.'s Windows NT freight train may barrel right over them, seem
poised to work together to block the tracks.
"[Microsoft] is a very powerful company with significant
marketing and ad budgets," said David Sandel, vice president of
marketing for Unix International. "We need to have a united
front, a set of interface specs and standards."
Unix International last week in New Orleans agreed to support
Open Software Foundation's Distributed Computing Environment,
which allows for OSF-compatible distributed processing.
UI has also revealed ambitions plans for multiprocessing,
transaction processing, and broader network support.
For its part, OSF said it will improve ease-of-use features with
"widget" technology from Lotus Development Corp., IBM, and
Digital Equipment Corp.
By the end of this year, UI will complete a Windows 3.1 emulation
system and provide Unix interoperability with AppleTalk, NetWare,
and LU6.2
Within two years, UI hopes to offer a scalable Unix release for
single- and multiprocessor systems with asynchronous I/O for
better system performance.
But some analysts said that if OSF and UI were serious about
fighting NT, they would simply merge.
...
-------------
[InfoWorld; February 15, 1993; Page 6]
USERS UNDETERRED BY HIGH COST OF WINDOWS MOVE
By Stuart J. Johnston And Shawn Willett
The cost of upgrading to Windows is a hefty $3,600 per user,
according to an industry survey to be released next month.
But corporate users believe the benefits of Windows still
outweigh the costs.
Vendors, meanwhile, are either phasing out DOS applications or
not upgrading them.
The survey, to be released next month by the Microcomputer
Managers Association, in New York, queried 400 of the country's
largest corporations and found the $3,6000 figure includes costs
for hardware, software, training, and support.
The figures do not reflect the cost of internal application
development, although they do include off-the-shelf applications.
The average cost of upgrading existing hardware to run Windows is
$1,183 and the cost of additional software is $921, the survey
said.
Sally Atkins, technology consultant for Boston-based John
Hancock Mutual Life Insurance Co., is one user who things Windows
is worth the high cost.
"With Windows, you are building a platform for the future with
graphical clients and client/server applications," Atkins said,
"Plus all the applications are heading for Windows."
Industry experts did not quarrel with the survey's cost estimates
for hardware and software. However, several said the training and
support costs -- pegged at $1,500 -- were overblown.
...
One reason users are flocking to Windows is that application
vendors are diverting resources from DOS products to Windows or
other platforms.
For example, Software Publishing Corp. president Irfan Salim said
last week that only maintenance releases would be provided for
its flagship product, Harvard Graphics for DOS.
Similarly, Lotus Development Corp. is offering only minor upgrades
to Freelance to ensure compatibility with new hardware, and DOS
versions of 1-2-3 will be retrofitted with features already on
the Widows version, Lotus said.
...
Even WordPerfect Corp., which prides itself on its commitment to
DOS users, said the company will devote increasing resources to
the Windows line.
"We'll move more toward a 70/30 \[Windows/DOS\] split in
development, marketing, and advertising," said a company
representative.
Analysts say vendors are motivated by unarguable economics.
"A lot of the people who have DOS have older machines, and many
don't buy new software," said Esther Dyson, editor of industry
newsletter _Release_1.-_, in New York. "Software vendors are
looking at returns on investments."
[
OS/2 fans should pay a lot of attention to this last paragraph.
Consider how many c.o.o.a people have bragged about the fact they
are running DOS 1.1 word processors and Windows 2.0 apps under
OS/2.
]
--Doug Barney and Jenette Borzo contributed to this article.
-------------
[InfoWorld; February 15, 1993; Page 8]
CHIP COULD GIVE HP MULTIMEDIA MASTERY.
7100LC runs video at 30 frames per second without added hardware
By Cate Corcoran And Jayne Wilson
Ina a move that could make it a leader in desktop multimedia,
Hewlett-Packard Co. is spinning of its workhorse 7100 RISC chip
and creating a low-cost version. That version will incorporate
parts of the MPEG and JPEG algorithms in the chip's instruction
set.
The new chip will be able to run video at 30 frames per second
without additional hardware -- something no PC chip can do --
said Dave Logan, marketing manager for HP's PA-RISC technology
group.
The 7100LC chip (LC stands for low cost) will also be able to
handle different kinds of byte ordering, making it able to run
ported PC operating systems such as Windows NT.
HP could use the chip in low-cost multimedia workstations or
high-end PCs running NT, potentially competing with Pentium
machines and DEC's Alpha PC as well as low-cost workstations from
Sun and IBM, analysts said.
There is no reason to change the byte ordering of the chip except
to run NT, said Michael Slater, editor and publisher of
_Microprocessor_Report_, in Sebastopool, Calif.
"HP understands the need to have a mainstream operating system
running on their microprocessor to proliferate it," said Ken
Lowe, an analyst with Dataquest, in San Jose, Calif. "That's
where the critical mass is."
However, HP denied that it had firm plans to run NT on the chip.
"We have designed the chip to support \[PC operating systems such
as NT\] if people want to use those in the future," Logan said.
"But we've made no commitment to do that."
The 7100LC chip will not begin appearing in products until 1994,
he said.
[
I wonder if this is the first time hardware has been designed to
fit the OS?
BTW, while NT isn't on an architecture with reverse byte order, it
was said at the NT developer conference that this is not a
requirement... however, for the sake of compatibility of data
files, and DOS and 16-bit Windows compatibility boxes, the byte
order change is probably essential.
]
-------------
[InfoWorld; February 15, 1993; Page 86]
[Cringely's gossip column.]
...
Still, holidays are better than nothing. They give us reasons to
do nice things for each other. The masses will probably
celebrate, for example, when Windows NT hits the market with
Double Space data compression and an introductory price of $99.
There won't be any celebrating at IBM, Novell, or Next.
That's the good news this week from Redmond. The bad news is that
its Access database can corrupt data because of problems locking
tables. Paradox users report they can write to supposedly locked
tables. Also, Access doesn't really use ODBC, Microsoft's vaunted
data access protocol, but is actually hard-coded to specific back
ends such as SQL Server.
News that can be good or bad, depending on who you are, is that
Microsoft is working on a 3270 connectivity product (bad for Wall
Data, which is approaching an IPO) and on a PIM code-named Ren
and Stimpy. Also, Excel 5.0 for Widows will be out in the fall,
minus Q+E form Pioneer, replaced by data access technology ripped
out of Access. The kids in Redmond hope to match Lotus Improv's
Dynamic Views with an enhanced outline feature but won't even try
Improv's English language formulas.
-------------
Brian
C. Brian Sturgill Windows, WfW, Windows NT and some OS/2 2.0
University of Utah information available via anonymous ftp to
Center for Software Science easy.cs.utah.edu. Also all my Tidbits posts.
br...@cs.utah.edu; CIS: 70363,1373 Windows NT SDK: $69; (800) 227-4679
Peter Mayne
In article <7001...@otter.hpl.hp.com>, s...@otter.hpl.hp.com (Steve Loughran) writes:
>> >I wonder if this is the first time hardware has been designed to
>> >fit the OS?
>
>> Actually, it isn't.
>
>> VAX/VMS
>> OS/400 -> AS/400
>> SPARC for UNIX (and most RISC CPUs, for that matter)
>> 8086 (for some kind of CP/M compatibility)
>> The CRISP processor specifically designed for C (see a recent Byte)
>
>Peter is right in that RISC CPUs are normally optimised for the code which
>is to be run on it, and given the amnout of CPU time the OS invariably
>consumes that the OS must be used to determine which instructions
>to provide[*] .Also things like memory protection models ought to match,
>otherwise you get CPU designers coming up with bizarre segmented memory
>architechtures when all us hackers want is a flat memory model ;-)
>
>>(but not Alpha 8->)
>
>You mean no-one had a look at the instruction usage of any unix before
>designing the architecture? Surely they didnt just use VMS.
The designers looked at application instruction usage. OS specific stuff
can be (and is) implemented in PALcode. Thus, OSF/1 isn't burdened by
OpenVMS features, Windows NT isn't burdened by OSF/1 features, etc, etc.
>Oh yeah. MIPS processirs have had two endian modes for a long time. Something
>to do with VaxStations I believe.
Only in the sense that DEC's MIPS chips ends are at the same end as VAXen ends.
> Steve
>[*] cynics may argue that in fact most modern RISC CPUS are optimised purely
>for impressive benchmark figures.
A tempting argument... ;-)
