Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Tidbits 3/7/93

0 views
Skip to first unread message

Brian Sturgill

unread,
Mar 7, 1993, 10:24:18 AM3/7/93
to
[
I'm a bit behind and there's lots of stuff, so I'll summarize
more than I usually do.

No index this time (too many articles... you can skip between
articles by searching for -----).
]

--------
[InfoWorld; February 22, 1993; Page 1,95]
[
NetWare 4.0 to ship March 31.
User's have some concerns about its speed.
]

--------
[InfoWorld; February 22, 1993; Page 1,95]
[
Network, mail, and a minor piece of disk compression have
been removed from DOS 6.0.
The disk compression piece was an off-line piece that would
give an extra 5%. Stac is expect to continue their suit.
The network and mail pieces were removed at the request of
corporate accounts that did not want their users "messing around
with a different network".
DOS 6.0 should ship very soon.
]

--------
[InfoWorld; February 22, 1993; Page 2]
Bug fixes delay release of OS/2 2.1
By Ed Scannell
Although beta testers of IBM's OS/2 2.1 generally favor the
product, sources last week said they don't expect the final version
to be available until May.

IBM has pushed its delivery goal of the first quarter back
another six to eight weeks to ensure fixes to a variety of minor
but persistent bugs.

Although it is doubtful the delay will seriously damage the
products chances against Microsoft Corp.'s Windows 3.1, it does
increase the company's pressures to release a version that is
rock solid.

"We have looked at OS/2, but my worry is they don't have a lot of
time left," said Brian Moura, assistant city manager for the city
of San Carlos, Calif. "They may need two or three more revs to
get it all right, but they may not have that time," he said.

With a May delivery, IBM may run up against Microsoft's mammoth
marketing of Windows NT, which is set to begin in May at Windows
World, sources said. However the final version of Windows NT is
not expected until June.

"IBM needs to make something happen with 2.1, or its' facing the
possibility of being a niche product for a very long time," said
one developer who requested his name be withheld.

The bulk of the OS/2 2.1 problems that are being addressed
revolve around making it work better with Windows applications
and improving the operating system's installation with lesser
known or generic clones.
...

--------
[InfoWorld; February 22, 1993; Page 18]
[
Borland's C++ for OS/2 will ship in March.
For the first 90 days it will cost $149.95... afterwards $495.
Looks like it is really nifty... is multi-thread... will do
compiles in a separate thread while you edit another, etc.
]

--------
[InfoWorld; February 22, 1993; Page 57-60]
[
IBM has a new ad campaign... sigh... I don't think they
understand anything about marketing.

It starts out with a page that has nothing but one phrase
in big TimesRoman type: "Now you don't have to operate within
the limits of the system."... it gets uglier and sillier from
there on. The pictures of the WPS Windows are gaudy beyond
words. The text and graphics layouts look like they were done by
children.
]

--------
[PC Week; February 15, 1993; Page 14]
[Begin Chart]
Pentium performance comparisons
The 66MHz Pentium doubles the performance of the 33/66MHz 486DX2 on Intel's
Icomp scale.

66MHz Pentium 567
60MHz Pentium 510
33/66MHz 486DX2 297
50 MHz 486DX 249
25/50MHz 486DX2 231
33MHz 486DX 166
25MHz 486SX 100

... but trails most RISC chips in SPECmark92 ratings

SPECint92 SPECfp92
66MHz Pentium 64.5 56.9
133MHz Alpha 65.3 111.0
40MHz SuperSPARC 52.6 64.7
99MHz HP PA-RISC 80.0 150.6
62.52MHz RS/6000 59.2 124.8
33/66MHz 486DX2 32.0 16.0

Intel's Icomp rating measures relative performance of Intel CPUs;
SPECint92 and SPEC fp92 measure CPU integer and floating-point
performance, respectively.

Source: Intel's Icomp rating: Intel sources. SPECmark92 ratings:
Pentium--Intel sources; other chips--individual manufacturers.
[End Chart]

--------
[PC Week; February 15, 1993; Page 27]
Sound features gaining market acceptance.
[
Two articles...

One talked about NCD X-terminals supporting sound.

The other talked about a large number of PC makes starting to
include sound... sometimes on the motherboard. Cheap prices
on sound chips and more usage in business applications are
driving the market.
The market is expected to triple by 1995.
]

--------
[PC Week; February 22, 1993; Page 1,20]
VISUAL C++: SUPERIOR CLASSES, COMPILER AND ENVIRONMENT
By Peter Coffee

...
With Visual C++ 1.0, to be released this week...
...
The Professional Edition priced at $499; upgrades from Microsoft
C/C++ 7.0 cost $139. The Standard Edition is priced at $199;
upgrades form Microsoft QuickC for Windows cost $79.

PERFORMANCE AND INTEGRATION OF TOOLS

At its lowest level, Visual C++ works with the same raw material
of source-code files, headers, resource scripts and the like that
Windows developers are already equipped to use. The compile,
likewise, can go to work with no addition overheads in parsing
the symbol stream.

Between Visual C++'s lower and upper levels are database engines
and other invisible mechanisms that provide, for example, dynamic
linkage between the Wizard utilities that generate code and the
raw source file that results. This linkage is flexible and
intelligently automatic: PC Week Labs' generate code was
annotated with literate comments.

Debugging was highly productive, with hypertext linkage of output
error messages and error-code explanations aided by color syntax
highlighting and other graphical aids. Within the constraints of
Windows' limited multitasking, compilation was able to proceed in
the background while we continued with other work.

PERFORMANCE AND SIZE OF APPLICATIONS
Refinements in the handling of p recompiled headers, and
convenient groupings of options into one-step settings for
"smallest size" and "highest speed," will further accelerate the
delivery of optimized code.

The enhanced linker in Visual C++ can now eliminate unused
library functions, while the compiler can be set to use the
faster, more concise instructions specific to 386 and later X86
chips.

SUPPORT FOR OBJECT-ORIENTED APPROACH
Visual C++ helps developers shift to an object approach with
semi-automated Wizards, accelerating and guiding the creation of
a new application and the development of new classes. Advanced
Windows features such as tool bars, Object Linking and Embedding
client support and print with preview became simple check-box
options under the new (but upward-compatible) Microsoft
Foundation Classes 2.0 model of Documents and Views.

Resources such as default toolbar bit maps were automatically
imported into the application's file set. We found it easy to
customize these with AppStudio tool.

We were impressed by the system's intelligent behavior in
recognizing when there was no logical connection with existing
classes, and in popping up the class construction aid
(ClassWizard) to structure the choice of Windows messages, member
function names and other necessary parameters.

VARIETY OF DELIVERY TARGETS
Visual C++ can deliver several types of executable files: Windows
and DOS applications, dynamic link libraries and VBX custom
controls, although OS/2 support is regrettably absent. DOS
support and the full Windows 3.1 Software Development Kit are
found only in the Professional addition.

DOCUMENTATION AND OTHER SUPPORT
Both the Professional and the Standard Edition benefit form
impressive aids to learning this large collection of tools. More
than 150,000 lines of sample code are included, along with
extensive hypertext-style on-line documentation and class library
sources.
...
[
The article doesn't really do VC++ justice... here are
some points that were left out:

Visual C++ code-editing environment is entirely Windows hosted.
App Studio uses drag and drop techniques for building resources.
It can use all those neat VBX controls available for Visual BASIC!
Has control automation facilities approaching that of Visual BASIC.
Multiple window editing environment.
Class Wizard is fancier than they mentioned.

Borland had better start worrying... Microsoft appears to have
finally passed them.
(Alas Visual C++ still does not have templates :-(.)
]

----------
[PC Week; March 1, 1993; Page 113, 114]
IBM ADDS TO LIST OF POSSIBLE NEW CEOs, DIVULGES FINANCIAL DETAILS.
By Neal Boudette

...
Worldwide revenue for personal systems -- which encompasses PCs,
RS/6000 workstations, displays and terminals, as well as AIX and
OS/2 -- fell 7 percent to $7.9 billion, down from $8.5 billion in
1991, the annual report says.

The revenue drop hit even as IBM experienced solid growth in unit
shipments of PCs and RS/6000 workstations. According to market
researcher International Data Corp. (IDC), in Framingham, Mass.,
IBM's worldwide PC shipments jumped 12 percent in 1992 to 3.2
million units, up from 2.85 million in 1991.
...

----------
[PC Week; March 1, 1993; Page 126]
[Spenser the Katt's gossip column]
...
Always under consideration: bring IBM Ambra PCs across the pond
to the United States. Unlikely right now, confided IBM PC chief
Jim Cannavino. Cannavino and his high-powered colleagues are
still doing their jobs -- jobs they may not have when a successor
to John Akers is named in the next 2 to 3 weeks.

One long-shot rumor is that packages have been offered to Mike
Spindler and John Sculley to come over as a team. That makes
sense, the Power Meower concurs, given the speed with which one
CEO surrounded by 25-year IBMers could quickly become isolated.

Morale at IBM is lower than low, since pink slips could come at
any moment. Layoffs, particularly in White Plains, Armonk and the
mid-Hudson Valley, are now a certainty.

Akers has all but checked out, going through motions and leaving
day-to-day operations -- "asking the hard questions" to Vice
Chairman and Chief Financial Officer Paul Rizzo. That has irked
Vice Chairman Jack Kuehler, but it all may not matter after a new
CEO takes over.
...

----------
[Open Systems Today; March 1, 1993; Page 1,69]
NT BEGIN SHORED UP WITH SERVER VERSION
By Mitch Wagner

Redmond, Wash. -- Microsoft said it plans to support
enterprisewide networking services and systems administration
features in a server version of Windows NT, to be released in June
or July, soon after the client version.

Microsoft is bringing out an Advanced Server Version of Windows
NT to address criticisms that NT might be suitable for the desktop
and low-end server but will lack the robustness and functionality
to make it a high-end platform. The server software will include
data integrity features to support disk arrays of any size;
enterprisewide naming and directory services; gateway software
for IBM, Apple and Unix; and advanced dial-in features, said
David Thacher, group product manager for Windows NT.

NT is scheduled to be introduced May 24 at Windows World in
Atlanta. Ga. The client version will ship in the last week of May
or in June. The client will be priced at less than $500 and the
server at more than $1,000, Thacher said.

The final, fully functional beta version, Beta 2, will ship
sometime this month Thacher said.

REQUIRES HIGH-END PC
The software requires an extremely high-end PC to run. The
minimum memory configuration on the client will be about 8
Mbytes, with bout 50 Mbytes disk storage needed for NT code and
swap space. On the server, a minimum 12 Mbytes memory and about
50 Mbytes storage will be needed. The storage space does not take
into account space needed for applications, Thacher said.

...

The client software will include the kernel and commands and
libraries to run applications, as well as a GUI. It will have
client networking software so it can be plugged unmodified into
the network. These include support for TCP/IP, Windows sockets,
IBM Data Link Control (DLC) and Remote Procedure Calls (RPCs),
compatible with the Open Software Foundation's Distributed
Computing Environment (DCE).

ADDRESSING INDUSTRY CONCERNS
The server features in Windows NT address precisely the kinds of
networking and large-scale server functionality that some NT
adversaries -- most notably those in the Unix community -- have
said NT will lack, a drawback that would keep it from becoming a
truly scalable server. These are functions which allow systems
administrators to manage large numbers of users and network
nodes, and also store gigabytes of data with reasonable
integrity.

Microsoft's Thacher said the server version of NT will add
network-management, network services and fault-tolerant features
to the client functionality. The server version replaces a
product that had been planned for later this year: LAN Manager
for NT.

The server version will support disk mirroring and striping for
RAID 5 functionality, to allow users to construct large, reliable
disk arrays that appear to NT to be a single volume. "We can
support terabytes of disk storage," Thacher said. "It's limited
by the hardware at this point."

On the networking side, the server OS will include a global name
space with domains functionality. This features is designed to
streamline systems administration and maximize user access across
an enterprisewide network. A user can maintain a single account
over the entire network, rather than having to maintain duplicate
accounts on every server he has access to, Thacher said. The
naming feature appears to be comparable to StreetTalk from Banyan
Vines and to the directory service that will be part of Novell's
forthcoming Novell NetWare Version 4.0.

APPLETALK SUPPORT
AppleTalk support will be built into the server version, through
a program called Services for Macintosh, will filesharing, print
sharing and full client-server support over X.25 and ISDN
networks, Thacher said.

Dial-in support comes form Remote Access Services, which makes
the client dialing in look like a part of the network, able to
receive and E-mail transparently and share file and print
services, no differently than over any other kind of network
connection, Thacher said.

The server version will include systems management features for
icon-driven setup and maintenance of desktops, multiple servers,
setup of profiles for users, and other system administration
functions, Thacher said.

Among the high-end platforms on which NT will run is Sequent. In
addition, Digital Equipment Corp. has said it will port NT to its
entire Alpha line of workstations and servers.

International language support will include German, French,
Spanish and Italian immediately, Swedish within 30 days, and
Japanese by the fourth quarter, Thacher said.

SNA connectivity will be accomplished by merging in the existing
Microsoft Comm Server product about 90 days from the initial
shipment of NT. The Comm Server is designed as an intelligent LAN
gateway to SNA hosts, to give access for DOS, Windows and
Macintosh clients.

----------
[PC Week; March 1, 1993; Page 1, 16]
SUN EMULATOR TO LINK UNIX, WINDOWS APPS
By Amy Cortese

Sun Microsystems Inc. will challenge Microsoft Corp.'s ownership
of Windows next month as it unveils emulation software that allows
shrink-wrapped Windows applications to run on Unix systems.

Sun's SunSelect division is developing a Windows application
binary interface (ABI) that it hopes to license to software
developers; SunSoft Inc., Hewlett-Packard Co., The Santa Cruz
Operation Inc. and Novell Inc. are among those considering
licensing it, sources said. With broad acceptance, Sun would
promote the ABI as the basis of a standard, SunSelect officials
said.

"There is a bunch of applications developers that would like to
see a standard set of documented Windows calls," said David
Roselund, market development manager for SunSelect in Chelmsford,
Mass.

Two factors could blunt Sun's push, however: Microsoft's ability
to quickly change the Windows API (application programming
interface), on which the ABI is based; and Microsoft's own plans
to provide ways for developers to create Windows applications
that run on Unix.

Unlike most Windows emulation technologies, Sun's Windows ABI
does not require that Windows or DOS be installed on the system.

The as-yet-unnamed software provides a translation layer between
Microsoft's published 16-bit Windows API and the Unix-based X
Window System. As a result, off-the-shelf Windows applications
should run on Intel- or RISC-based Unix system running the ABI as
long as the applications adhere to documented calls in the
Windows API, Roselund said.

The technology, which Sun obtained when it acquired Praxsys
Technologies Inc. last September, would give Unix users the same
ability to run Windows applications that IBM provides with OS/2
2.0. IBM however, has a licensing agreement with Microsoft.

Microsoft Chairman Bill Gates has said that the Praxsys
technology, without a license from Microsoft, would be a
violation of Microsoft's intellectual property rights. But
SunSelect officials maintain there are no legal obstacles to
offering the technology.

"We are certainly aware that Microsoft had made statements that
we could be infringing," said Cliff Utstein, SunSelect's senior
product manager for emulation products. "We have done everything
to satisfy out lawyers that we are not." he added that the
company has not precluded working with Microsoft.

Microsoft officials in Redmond, Wash., said they had no been
contacted about working with Sun on the technology.
...
Meanwhile, Microsoft is exploring its own ways to make the
Windows API available on Unix. It has showcased Win/U software
from Bristol Technology Inc., which allows ISVs to recompile
their Windows applications to run on Unix. Microsoft is also
working with two developers on a Windows-to-Unix project that
could be announced this spring, sources said.

----------
[PC Week; March 1, 1993; Page 1, 16]
WFW upgrade enters beta.
[
Features ... 32-bit I/O system (this means that DOS is almost not
used at all).
VFAT (appears to be a 32-bit version of code to handle the FAT filesystem).
FastMail
Support for NDIS 3.0 (the ethernet card driver spec for NT).
Microsoft-developed version of NetBios.
Support for Novell's ODI drivers and support to run over IPX.
Auditing support.
Mail -> fax support.
8-user chat.
]

----------
[PC Week; March 1, 1993; Page 3]
IBM announces corporate layoffs.
[
Mainly in mainframe areas. Exact numbers of how many layoffs (as
opposed to other options) not given number likely to be less than
10,000 though.
]
In other news IBM has begun warning users that its 386SLC based
ValuePoint 235T PC can experience memory parity errors when
running applications in a DOS-only environment. Users can receive
a fix by calling IBM at (800) 845-4263.
...
IBM plans to ship in the next two weeks a new beta release of
OS/2 2.1.
...

----------
[PC Week; March 1, 1993; Page 6]
[
Interesting side issue in article about slimmed down HP PA-RISC chip
(possibly for NT) that I reported on last time
]
Eichhorn said the forthcoming PA71000LC-based systems are aimed
at what he calls the "professional" workstation market. The
professional user, he said, wants a powerful desktop machine with
a multitasking operating system and built-in networking to tap
into client/server applications.

The professional market makes up about $1.3 billion of the
current $9 billion workstation market and could grow as much as
50 percent annually over the next four years, Eichhorn estimated.
[And doesn't that market have NT written all over it!]
...

----------
[PC Week; March 1, 1993; Page 8]
[Zeos reports a $28.2M loss on sales of $206.1 million.]

----------
[InfoWorld; March 1, 1993; Page 3]
...
IBM said it will probably exceed its original estimate of 25,000 job
cuts in 1993 but has yet to determine the exact number.
-- Ed Scannell

----------
[Open Systems Today; February 1, 1993; Page 4]
UNIX TO SUPPORT MORE OBJECTS

Unix will soon offer built-in features enabling it to run
applications and transaction-processing software that are
object-oriented and distributed over a network. But only if Unix
System Laboratories follows recommendations of the latest System
V Roadmap, set to be released at the next meeting of Unix
International in New Orleans on Feb. 11 and 12. According to a
copy of the unpublished roadmap, USL is being asked to have
distributed object management capabilities built into Unix by
next year. These capabilities would be based on object management
specifications set by the Object Management Group. The roadmap
calls for development of APIs to AppleTalk, IBM System Network
Architecture and NetWare protocols by next year, said Peter
Cunningham, UI chief executive. It also asks that, beginning this
year, Unix include a distributed file-system, with caching
abilities to allow users to share files across a wide-area
network independent of underlying network protocols. The OS would
have file locking, so two users would not be able to write to the
same file at the same time and confuse updates. It also would
include a distributed naming system, as well as a Windows 3.1
emulation package, with NT emulation expected in 1995.
Fault-tolerant software also will be built in. Novell, set to
purchase USL by the end of this quarter, said it will abide by
the roadmap.

----------
[Open Systems Today; February 1, 1993; Page 66]
[Chart associated with an article on SCO's current state.]
Contenders To the Throne
Vendors Vying For Business In The Multiuser, Multitasking PC Market.
SCO Microsoft Sun Novell
...
[
Note the absence of Portable OS/2 and NextStep.
]

----------
[PC Week; March 1, 1993; Page 13]
...
IBM TO SUPPORT VESA
IBM's new ValuePoint PCs will include a new local-bus mother-board
sporting two VESA-compatible slots and a P24T upgrade socket,
said sources.
...

----------
[PC Week; March 1, 1993; Page 53, 62]
NT GAINS VERTICAL MOMENTUM
Microsoft targets finance, health care, pharmaceuticals
By Amy Cortese

Microsoft Corp. has collected its vertical marketing activities
under one group to increase its efforts to create
vertical-market-specific standards for the forthcoming Windows
NT.

Company officials are counting on NT to give them a foot in the
door at many businesses that have traditionally relied on Unix or
proprietary systems to run their operations.

The Redmond, Wash., firm has begun working with vendors in a
dozen targeted industries, including finance, pharmaceuticals and
health care, to create vertical-market-specific software
standards.

Microsoft's most visible work to date has been in the area of
financial services, where the company has defined Windows Open
Services Architecture (WOSA) extensions and other specifications
for retail banking and securities customers and vendors.

"Microsoft has been very willing to understand out marketplace,"
said Kelsey Biggers, managing director of information delivery
services at Banker's Trust in New York. The bank, which has
standardized on Windows, is looking at moving to NT for its
mission-critical applications, Biggers said.

Microsoft has also quietly been lining up partners in other
vertical markets, such as pharmaceuticals, manufacturing, health
care and insurance, as well as professional disciplines
including accounting, engineering and legal.

The company is stepping up efforts in the education and
government markets as well.

"We are realigning a lot of marketing around verticals," said
Doug Martin, public sector marketing manager at Microsoft. "The
timing is there to exploit NT.
...

----------
[PC Week; March 1, 1993; Page 122]
SMALL BUSINESSES SHOP RETAIL, STUDY SHOWS

The majority of small businesses buy through retailers as their
chief distribution channel, according to a recent survey by
Business Research Group, a market-research firm in Newton, Mass.
The survey of almost 700 small businesses (those with fewer than
100 employees) showed that 60 percent shop retail outlets for
PCs, peripherals, software and networking equipment. Telephone and
mail-order sales were made by 16 percent of the respondents.

----------
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

Sean J. Orourke

unread,
Mar 7, 1993, 2:12:21 PM3/7/93
to
In article <C3Iys...@ms.uky.edu> br...@easy.cs.utah.edu (Brian Sturgill) writes:
>[
>I'm a bit behind and there's lots of stuff, so I'll summarize
>more than I usually do.
>
When you have the time, please run your Tidbits posts and the stuff at your
FTP site through a spell-checker.

[some deleted]


>[InfoWorld; February 22, 1993; Page 1,95]
>[
>NetWare 4.0 to ship March 31.
>User's have some concerns about its speed.

Users, you mean?

[more deleted]


>The disk compression piece was an off-line piece that would
>give an extra 5%. Stac is expect to continue their suit.

I assume Stac is expected to continue their suit, or is it that Stac is
expecting to continue their suit? Please clarify.

>The bulk of the OS/2 2.1 problems that are being addressed
>revolve around making it work better with Windows applications
>and improving the operating system's installation with lesser
>known or generic clones.

I hope they're also working on the Pallette Manager... It's the only thing
I've ever had trouble with on my KLH 386/40 Spirit 61, a lesser known or
generic clone. :)

>--------
>[InfoWorld; February 22, 1993; Page 18]
>[
>Borland's C++ for OS/2 will ship in March.
>For the first 90 days it will cost $149.95... afterwards $495.
>Looks like it is really nifty... is multi-thread... will do
>compiles in a separate thread while you edit another, etc.
>]

$150 vs $500? Looks like I'm buying my next C++ product from Borland
instead of Microsoft. :)

>--------
>[InfoWorld; February 22, 1993; Page 57-60]
>[
>IBM has a new ad campaign... sigh... I don't think they
>understand anything about marketing.
>
>It starts out with a page that has nothing but one phrase
>in big TimesRoman type: "Now you don't have to operate within
>the limits of the system."... it gets uglier and sillier from
>there on. The pictures of the WPS Windows are gaudy beyond
>words. The text and graphics layouts look like they were done by
>children.
>]

Will someone who gets InfoWorld post the full text of this ad? Brian, please
stop with the "IBM's new ad really sucks: The first sentence is blah-blah-blah,
and it gets worse from there."
>
[Intel's new chips get good Icomp ratings and Visual C++ is god deleted]


>
>Borland had better start worrying... Microsoft appears to have
>finally passed them.
>(Alas Visual C++ still does not have templates :-(.)
>]

Ah, and I bet there's no Visual C++ for OS/2 and they've broken it for working
under WIN-OS/2. No thanks.


>
>----------
>[PC Week; March 1, 1993; Page 113, 114]
>IBM ADDS TO LIST OF POSSIBLE NEW CEOs, DIVULGES FINANCIAL DETAILS.
>By Neal Boudette
>
>...
>Worldwide revenue for personal systems -- which encompasses PCs,
>RS/6000 workstations, displays and terminals, as well as AIX and
>OS/2 -- fell 7 percent to $7.9 billion, down from $8.5 billion in
>1991, the annual report says.
>

That's it! Even the simplest math shows that OS/2 lost $600 million in
1992. :)
>
[some deleted]


>----------
>[PC Week; March 1, 1993; Page 126]
>[Spenser the Katt's gossip column]
>...
>Always under consideration: bring IBM Ambra PCs across the pond
>to the United States. Unlikely right now, confided IBM PC chief

--------


>Jim Cannavino. Cannavino and his high-powered colleagues are
>still doing their jobs -- jobs they may not have when a successor
>to John Akers is named in the next 2 to 3 weeks.
>

I never believe anything that was supposedly "confided" to the writer.
In my dictionary, to confide in is to entrust a secret to... And I'm supposed
to trust the person divulging the secret?

[everyone at IBM is about to lose their job deleted]


>
>----------
>[Open Systems Today; March 1, 1993; Page 1,69]
>NT BEGIN SHORED UP WITH SERVER VERSION

I assume NT is BEING shored up with a Server Version.


> By Mitch Wagner
>
>Redmond, Wash. -- Microsoft said it plans to support
>enterprisewide networking services and systems administration
>features in a server version of Windows NT, to be released in June
>or July, soon after the client version.
>
>Microsoft is bringing out an Advanced Server Version of Windows
>NT to address criticisms that NT might be suitable for the desktop
>and low-end server but will lack the robustness and functionality
>to make it a high-end platform. The server software will include
>data integrity features to support disk arrays of any size;
>enterprisewide naming and directory services; gateway software
>for IBM, Apple and Unix; and advanced dial-in features, said
>David Thacher, group product manager for Windows NT.
>
>NT is scheduled to be introduced May 24 at Windows World in
>Atlanta. Ga. The client version will ship in the last week of May
>or in June. The client will be priced at less than $500 and the
>server at more than $1,000, Thacher said.
>
>The final, fully functional beta version, Beta 2, will ship
>sometime this month Thacher said.
>

I think it's been pointed out that according to most definitions of Alpha
and Beta releases, this is really the first Beta release since it's the first
that's fully functional. But then, since using the Undelete functions in OS/2
2.1b is a bad idea, maybe that's just an Alpha too. :)

>REQUIRES HIGH-END PC
>The software requires an extremely high-end PC to run. The
>minimum memory configuration on the client will be about 8
>Mbytes, with bout 50 Mbytes disk storage needed for NT code and
>swap space. On the server, a minimum 12 Mbytes memory and about
>50 Mbytes storage will be needed. The storage space does not take
>into account space needed for applications, Thacher said.
>

I'd hate to speculate how many people won't be going for NT because of this.
Probably fewer than I'd think, though. :)
>
[lots deleted]


>----------
>[PC Week; March 1, 1993; Page 1, 16]
>WFW upgrade enters beta.
>[
>Features ... 32-bit I/O system (this means that DOS is almost not
>used at all).
>VFAT (appears to be a 32-bit version of code to handle the FAT filesystem).
>FastMail
>Support for NDIS 3.0 (the ethernet card driver spec for NT).
>Microsoft-developed version of NetBios.
>Support for Novell's ODI drivers and support to run over IPX.
>Auditing support.
>Mail -> fax support.
>8-user chat.
>]

Sound like it'd make a nice OS/2 app.
>
[more deleted]

Can anyone who has as many magazine subscriptions as Brian apparently has
say whether or not most of his stuff comes from rumour pages? I noticed that
quite a few of the page numbers are rather small; back when I subscribed to
Byte, the first couple of pages were the rumour area. I'm curious as to
whether or not this is true of other magazines.

Sean

Deryk Barker

unread,
Mar 7, 1993, 12:51:53 PM3/7/93
to
br...@easy.cs.utah.edu (Brian Sturgill) writes:
: [PC Week; February 22, 1993; Page 1,20]

: VISUAL C++: SUPERIOR CLASSES, COMPILER AND ENVIRONMENT
: By Peter Coffee
[....]
: The enhanced linker in Visual C++ can now eliminate unused
: library functions

Ah progress! I seem to recall the mainframe linker I worked with 19
years ago did this - and nobody called it enhanced either.

--
Real: Deryk Barker, Computer Science Dept., Camosun College, Victoria B.C.
Email: (dba...@spang.camosun.bc.ca)
Phone: +1 604 370 4452

C.R.S. Schanck

unread,
Mar 7, 1993, 6:39:37 PM3/7/93
to
In article <1993Mar7.1...@ringer.cs.utsa.edu> soro...@lonestar.utsa.edu (Sean J. Orourke) writes:
>When you have the time, please run your Tidbits posts and the stuff at your
>FTP site through a spell-checker.

I think it is a bit much to complain in this fashion -- it has to be a
fair amount of work for Brian to type in this stuff -- I think spelling is
a bit much to pick on. Let's stay a little decent here?

Chris
--
======================================================================
csch...@access.digex.com "OS/2 -- The Operating
C.R.S. Schanck System for You and Me!"
====== My Opinions are *mine* -- not that I am proud of them... ======

David White

unread,
Mar 7, 1993, 7:26:08 PM3/7/93
to

>In article <C3Iys...@ms.uky.edu> br...@easy.cs.utah.edu (Brian Sturgill) writes:
>>[
>>I'm a bit behind and there's lots of stuff, so I'll summarize
>>more than I usually do.
>>
>When you have the time, please run your Tidbits posts and the stuff at your
>FTP site through a spell-checker.

>[some deleted]
>>[InfoWorld; February 22, 1993; Page 1,95]
>>[
>>NetWare 4.0 to ship March 31.
>>User's have some concerns about its speed.
>Users, you mean?

>[more deleted]
>>The disk compression piece was an off-line piece that would
>>give an extra 5%. Stac is expect to continue their suit.
>I assume Stac is expected to continue their suit, or is it that Stac is
>expecting to continue their suit? Please clarify.

Sean,

It's nice to have all one's spelling and grammar correct. But if a person
does not have much time, I will excuse them, provided their writing is still
intelligible to me, and interesting.

I for one am grateful for the information that Brian passes on. It is not
worth subscribing from here to the weekly trade publications quoted due to
airmail costs. So I actually look forward to his postings.

And yes, it is obvious that Brian has a certain bias in his OS views. But I
know what it is and can take that into account. Most writers for computer
publications would have preconceptions about what is and is not good
software/hardware.

I suggest that disagreements be confined to technical matters. Otherwise,
those for whoom English (or American! :)) is not their first language will
be hesitant to contribute to the discussions.

(spot the mistake above!) david

--
david white (engineer, Goon fan & son of my Dad)
Internet dav...@auck.irl.cri.nz Fax +64 9 443-4737

Charles Kincy

unread,
Mar 8, 1993, 4:15:27 PM3/8/93
to
In article <C3Iys...@ms.uky.edu> br...@easy.cs.utah.edu (Brian Sturgill) writes:
[..8<..]

> SPECint92 SPECfp92
> 66MHz Pentium 64.5 56.9
> 133MHz Alpha 65.3 111.0

^^^^

Excuse me for a second.
*falls out of chair laughing*
*gets back into chair*
*falls out of chair laughing again*
*pulls keyboard to the floor*

So, this is it?! A 1% performance improvement over Pentium?
With a chip that is twice as fast in MHz? Excuse me whilst I
bust a gut again....

[Note that a 133MHz Pentium would probably deliver as many SPECfps
as an Alpha....]

Sure, if you want floating point, prefer an Alpha over a Pentium.
But *I* would suggest that you skip the Alpha and go straight to
the PA-RISC, because if you are interested in number-crunching,
you should forsake all else in search of that goal.

In other words, DEC is hosed.

[..8<..]

cpk
--
"The WHITE ZONE is for loading and unloading only. If you gotta load or un-
load, go to the WHITE ZONE. You'll love it. It's a way of life!" --Zappa

Technology always seems to keep one step ahead of human wisdom.

Matt Kennel

unread,
Mar 8, 1993, 5:26:32 PM3/8/93
to
cki...@cs.umr.edu (Charles Kincy) writes:

: In article <C3Iys...@ms.uky.edu> br...@easy.cs.utah.edu (Brian Sturgill) writes:
: [..8<..]
:
: > SPECint92 SPECfp92
: > 66MHz Pentium 64.5 56.9
: > 133MHz Alpha 65.3 111.0
: ^^^^
:
: Excuse me for a second.
: *falls out of chair laughing*
: *gets back into chair*
: *falls out of chair laughing again*
: *pulls keyboard to the floor*
:
: So, this is it?! A 1% performance improvement over Pentium?
: With a chip that is twice as fast in MHz? Excuse me whilst I
: bust a gut again....

Can you think of many truly CPU-intensive programs that don't
use floating point?

Also note that SPECint also counts I/O as 'time' in some benchmarks.

: [Note that a 133MHz Pentium would probably deliver as many SPECfps
: as an Alpha....]

Maybe there's a *reason* why the Pentium's clockspeed can't go as
fast as Alpha's?

Is the Alpha chip cheaper? (I have no idea here)

Also, Intel compilers are mature, and Alpha compilers aren't. Yes
that makes a *big* difference, for chips like Alpha.

Isn't the 133Mhz part the *bottom* of the Alpha line?

: Sure, if you want floating point, prefer an Alpha over a Pentium.


: But *I* would suggest that you skip the Alpha and go straight to
: the PA-RISC,

Yes,

: because if you are interested in number-crunching,


: you should forsake all else in search of that goal.

No.

: --

: "The WHITE ZONE is for loading and unloading only. If you gotta load or un-
: load, go to the WHITE ZONE. You'll love it. It's a way of life!" --Zappa
:
: Technology always seems to keep one step ahead of human wisdom.

--
-Matt Kennel m...@inls1.ucsd.edu
-Institute for Nonlinear Science, University of California, San Diego
-*** AD: Archive for nonlinear dynamics papers & programs: FTP to
-*** lyapunov.ucsd.edu, username "anonymous".

Marc Fraioli

unread,
Mar 8, 1993, 7:05:27 PM3/8/93
to
In article <1993Mar7.1...@ringer.cs.utsa.edu> soro...@lonestar.utsa.edu (Sean J. Orourke) writes:
>In article <C3Iys...@ms.uky.edu> br...@easy.cs.utah.edu (Brian Sturgill) writes:
>>[
>>I'm a bit behind and there's lots of stuff, so I'll summarize
>>more than I usually do.
>>
>When you have the time, please run your Tidbits posts and the stuff at your
>FTP site through a spell-checker.
>
This criticism was uncalled for. Why don't you try typing all that
stuff in some time? He's doing it since other people might like to read
it. No need to be so nasty.

>
>Can anyone who has as many magazine subscriptions as Brian apparently has
>say whether or not most of his stuff comes from rumour pages? I noticed that
>quite a few of the page numbers are rather small; back when I subscribed to
>Byte, the first couple of pages were the rumour area. I'm curious as to
>whether or not this is true of other magazines.
>

It's not. The trade rags tend to put news in front, reviews and opinion
columns in the middle, and rumor in the back.

>Sean
>


--
Marc Fraioli
mfra...@grebyn.com (So I'm a minimalist...)

forrest d whitcher

unread,
Mar 8, 1993, 11:19:16 PM3/8/93
to
In article <1993Mar8.2...@umr.edu> cki...@cs.umr.edu (Charles Kincy) writes:
>In article <C3Iys...@ms.uky.edu> br...@easy.cs.utah.edu (Brian Sturgill) writes:
>[..8<..]
>
>> SPECint92 SPECfp92
>> 66MHz Pentium 64.5 56.9
>> 133MHz Alpha 65.3 111.0
> ^^^^
>
>Excuse me for a second.
>*falls out of chair laughing*
>*gets back into chair*
>*falls out of chair laughing again*
>*pulls keyboard to the floor*
>
>So, this is it?! A 1% performance improvement over Pentium?
>With a chip that is twice as fast in MHz? Excuse me whilst I
>bust a gut again....

Yes this is what all the hoopla is about.

>[Note that a 133MHz Pentium would probably deliver as many SPECfps
>as an Alpha....]
>
>Sure, if you want floating point, prefer an Alpha over a Pentium.
>But *I* would suggest that you skip the Alpha and go straight to
>the PA-RISC, because if you are interested in number-crunching,
>you should forsake all else in search of that goal.
>

and I would commend you look at ibm's RIOS as well, depends on the flavor
of your NC program, for _most_ large matix operations, the super-scalar is
worth a 1.5, sometimes 2 x clock, so expect a 60 mhz rs/6000 to perform on
a bit better than a 90 mhz pa/risc platform in lets say finite element
analysis. Of course the HP chip set really cranks on integer stuff and some
types of number crunching.

And of course you will soon (sic) be able to run os/2 on rios.

>In other words, DEC is hosed.
>

As nice a thought as that is :-) the NT folk seem to rather like the DEC
platform, and of course VMS users need something to hang their codes on.

forrest whitcher f...@world.std.com


thom...@kuhub.cc.ukans.edu

unread,
Mar 9, 1993, 4:06:36 AM3/9/93
to
In article <1ngh6o...@network.ucsd.edu>, m...@lyapunov.ucsd.edu (Matt Kennel) writes:
> cki...@cs.umr.edu (Charles Kincy) writes:
> : In article <C3Iys...@ms.uky.edu> br...@easy.cs.utah.edu (Brian Sturgill) writes:
>
> Can you think of many truly CPU-intensive programs that don't
> use floating point?
>
Windows, OS/2?

Almost any database program?

> --
> -Matt Kennel m...@inls1.ucsd.edu
> -Institute for Nonlinear Science, University of California, San Diego
> -*** AD: Archive for nonlinear dynamics papers & programs: FTP to
> -*** lyapunov.ucsd.edu, username "anonymous".

--
| Charles A. Thompson | thom...@ukanvm.cc.ukans.edu |
| University of Kansas | thom...@ukanvax.cc.ukans.edu |
| > They call me SkiBum < | 32 bit OS/2 2.0 is for you! |
|----------------------------------------------------------------|
| Listen to Rush Limbaugh from 12-3 EST on the radio EVERYWHERE! |
------------------------------------------------------------------

Torben Friis

unread,
Mar 9, 1993, 7:25:46 AM3/9/93
to
thom...@kuhub.cc.ukans.edu writes:

>In article <1ngh6o...@network.ucsd.edu>, m...@lyapunov.ucsd.edu (Matt Kennel) writes:
>> cki...@cs.umr.edu (Charles Kincy) writes:
>> : In article <C3Iys...@ms.uky.edu> br...@easy.cs.utah.edu (Brian Sturgill) writes:
>>
>> Can you think of many truly CPU-intensive programs that don't
>> use floating point?
>>
>Windows, OS/2?

>Almost any database program?

Compilers, filesystems (server)?

>> --
>> -Matt Kennel m...@inls1.ucsd.edu
>> -Institute for Nonlinear Science, University of California, San Diego
>> -*** AD: Archive for nonlinear dynamics papers & programs: FTP to
>> -*** lyapunov.ucsd.edu, username "anonymous".

>--
>| Charles A. Thompson | thom...@ukanvm.cc.ukans.edu |
>| University of Kansas | thom...@ukanvax.cc.ukans.edu |
>| > They call me SkiBum < | 32 bit OS/2 2.0 is for you! |
>|----------------------------------------------------------------|
>| Listen to Rush Limbaugh from 12-3 EST on the radio EVERYWHERE! |
>------------------------------------------------------------------

--

Kind regards,
Torben Friis
----------------------------------------------------------------------
The Institute of Applied Computer Science (IFAD) | Something
Forskerparken 10, DK-5230 Odense M, Denmark | smart
Phone: +45 65 932 300, Fax: +45 65 932 999 | should
E-mail: tfr...@ifad.dk or tfr...@imada.ou.dk | be here
----------------------------------------------------------------------

Nectar the Unnamed

unread,
Mar 9, 1993, 11:34:02 AM3/9/93
to
f...@world.std.com (forrest d whitcher) writes:
>And of course you will soon (sic) be able to run os/2 on rios.

Does anybody have any information to support this? Has there been
any sight of OS/2 on Mach or on any chip besides Intel?

--
Nectar --><--
nec...@world.std.com

cna...@vnet.ibm.com

unread,
Mar 9, 1993, 12:27:14 PM3/9/93
to
In <C3MrC...@world.std.com> nec...@world.std.com (Nectar the Unnamed) writes:
>f...@world.std.com (forrest d whitcher) writes:
>>And of course you will soon (sic) be able to run os/2 on rios.
>
>Does anybody have any information to support this? Has there been
>any sight of OS/2 on Mach or on any chip besides Intel?
>

IBM has demonstrated the OS/2 MK/AIX running on a RS/6000 (rios) at a few trade
shows and for some press folks.

Cliff Nadler
IBM Federal Systems Company

Note: This post represents the views of the poster, not that of IBM or IBM
Federal Systems Company

Nectar the Unnamed

unread,
Mar 9, 1993, 4:09:51 PM3/9/93
to
cna...@vnet.ibm.com writes:

>In <C3MrC...@world.std.com> nec...@world.std.com (Nectar the Unnamed) writes:
>>f...@world.std.com (forrest d whitcher) writes:
>>>And of course you will soon (sic) be able to run os/2 on rios.
>>
>>Does anybody have any information to support this? Has there been
>>any sight of OS/2 on Mach or on any chip besides Intel?
>>

>IBM has demonstrated the OS/2 MK/AIX running on a RS/6000 (rios) at a few trade
>shows and for some press folks.


Details? Which trade shows? I remember that it was supposed to be
shown at Networld, but was cancelled at the last moment. I hadn't
heard anything since then, and I've been watching. I'm going to Uniforum
next week. Seems like a good place to show such a product. Has anyone
heard if it'll be there?

Kamal Shaker.

unread,
Mar 9, 1993, 11:40:32 PM3/9/93
to
f...@world.std.com (forrest d whitcher) writes:
:
: As nice a thought as that is :-) the NT folk seem to rather like the DEC

: platform, and of course VMS users need something to hang their codes on.
:
: forrest whitcher f...@world.std.com

Oh now we're not, I've been using VMS for 4 years now and OS/2 2.x for
1 now, I love both and there's no way I'll be running NT. In fact I'm
a bit dissapointed in DEC for pushing NT on thier alpha chips. You think
they'd want to show something fast to show it off. What I can't for in
portable OS/2 on an alpha, then I'd be truly happy.....


--
------------------------------------------------------------------------------
Kamal Shaker, | cs...@luxury.latrobe.edu.au or sha...@latcs1.lat.oz.au
Student Vax Cluster, | %SYSTEM-F-EXQUOTA, Alcohol quota exceeded,
La Trobe Uni, | stomach dumped.
------------------------------------------------------------------------------
There's enough water on earth to drown the human population 87.6 million times

Charles Kincy

unread,
Mar 10, 1993, 10:32:08 AM3/10/93
to
In article <C3LtC...@world.std.com> f...@world.std.com (forrest d whitcher) writes:
[...]

>and I would commend you look at ibm's RIOS as well, depends on the flavor
>of your NC program, for _most_ large matix operations, the super-scalar is
>worth a 1.5, sometimes 2 x clock, so expect a 60 mhz rs/6000 to perform on
>a bit better than a 90 mhz pa/risc platform in lets say finite element
>analysis. Of course the HP chip set really cranks on integer stuff and some
>types of number crunching.

Of course, you're right. What I'm really waiting for is IBM to leapfrog
HP again right about the time that everyone feels the itch to upgrade
their workstations. :-)

Besides, I'll prefer PM over VUE any old day...NeXTStep even..which
leads me to.....

>And of course you will soon (sic) be able to run os/2 on rios.

Of course...that means all this PM work I'm doing won't lock me
into an Intel platform. *whew* :-)

Ya think they might port it to PA-RISC? Nah.... :)

>>In other words, DEC is hosed.
>>
>
>As nice a thought as that is :-) the NT folk seem to rather like the DEC
>platform, and of course VMS users need something to hang their codes on.

Well, that's all well and good, but DEC is going to have a difficult
time expanding its market, given Intel's marketing prowess and HP/IBM's
raw CPU power.

At any rate, I *snork* heartily at whoever said that "the PC war may
be over before it really begins". He may be right, but I don't think
the outcome will be what he had hoped. :-)

Charles Kincy

unread,
Mar 10, 1993, 10:38:05 AM3/10/93
to
In article <1993Mar10....@lugb.latrobe.edu.au> cs...@luxury.latrobe.edu.au (Kamal Shaker.) writes:
>f...@world.std.com (forrest d whitcher) writes:
>:
>: As nice a thought as that is :-) the NT folk seem to rather like the DEC
>: platform, and of course VMS users need something to hang their codes on.
>:
>: forrest whitcher f...@world.std.com
>
>Oh now we're not, I've been using VMS for 4 years now and OS/2 2.x for
>1 now, I love both and there's no way I'll be running NT. In fact I'm
>a bit dissapointed in DEC for pushing NT on thier alpha chips. You think
>they'd want to show something fast to show it off. What I can't for in
>portable OS/2 on an alpha, then I'd be truly happy.....

Remember what the project leader of NT (Cutler) said? His driving
force behind developing NT? His reason for living?

"Screw Digital!" [If you don't believe me, read UnixWorld a few
months back...they did a nice little article on the guy.]

With that in mind, I think that DEC is going to push NT on
their chips only half-heartedly, as it is the baby of the guy they screwed
four years ago....who, in turn, is out to return the favor to Digital.

Expect to see DEC push VMS a lot harder than it pushes NT.

fred j mccall 575-3539

unread,
Mar 10, 1993, 3:50:37 PM3/10/93
to

>In article <C3Iys...@ms.uky.edu> br...@easy.cs.utah.edu (Brian Sturgill) writes:
>>[
>>I'm a bit behind and there's lots of stuff, so I'll summarize
>>more than I usually do.
>>
>When you have the time, please run your Tidbits posts and the stuff at your
>FTP site through a spell-checker.

Only the Clueless flame over spelling and grammar. Some of us don't
have English as a first language, some of us who do just never learned
to spell, and some of us are lousy typists. If you want perfect
spelling and grammar on what he posts, I guess you could send Brian a
check. Then, since you'd be paying for the service, you'd have some
right to complain about how he performs it.

--
"Insisting on perfect safety is for people who don't have the balls to live
in the real world." -- Mary Shafer, NASA Ames Dryden
------------------------------------------------------------------------------
Fred....@dseg.ti.com - I don't speak for others and they don't speak for me.

Paul S Secinaro

unread,
Mar 9, 1993, 7:50:14 AM3/9/93
to
In article <1ngh6o...@network.ucsd.edu> m...@lyapunov.ucsd.edu (Matt Kennel) writes:
>: So, this is it?! A 1% performance improvement over Pentium?
>: With a chip that is twice as fast in MHz? Excuse me whilst I
>: bust a gut again....
>
>Can you think of many truly CPU-intensive programs that don't
>use floating point?

Any system that doesn't have a graphics coprocessor will hammer the
CPU with integer math for line drawing, blitting, etc. Also, many
standard DSP'ish algorithms are written using fixed-point math, since
many DSP chips don't do floating point. For example, I believe the
free JPEG image compression code being circulated around the net is
written so that no floating point math is used.

>Also note that SPECint also counts I/O as 'time' in some benchmarks.

Also, SPEC is a compiler benchmark as much as it is a CPU benchmark.
I suspect bad Alpha compilers could be the culprit here. The chip
just shouldn't be that slow.

--
Paul Secinaro
ps...@kepler.unh.edu
Synthetic Vision and Pattern Analysis Laboratory
UNH Dept. of Electrical and Computer Engineering

Glen Anthony Graham

unread,
Mar 11, 1993, 2:28:26 AM3/11/93
to
In article <1ni3q6$g...@mozz.unh.edu> ps...@kepler.unh.edu (Paul S Secinaro) writes:
>Also, SPEC is a compiler benchmark as much as it is a CPU benchmark.
>I suspect bad Alpha compilers could be the culprit here. The chip
>just shouldn't be that slow.

Yes, and as someone pointed out, the Alpha AXP is a new chip so the compilers
are very immature...

However, the _Pentium_ is also a new processor. I understand that it is quite
different from the 486 with respect to pipelines etc. So, couldn't the same
thing be said about the Pentium - that the compiler is inefficient and can only
get a lot better?

--
sl...@ccwf.cc.utexas.edu // gl...@cs.utexas.edu **** Glen A. Graham ****
-------------------------------------------------------------------------------
I don't want to achieve immortality through my work. I want | Fools wander,
to achieve immortality through not dying. --Woody Allen | Wise men Travel.

scott eric cantor

unread,
Mar 12, 1993, 12:13:46 PM3/12/93
to
In article <1993Mar7.1...@ringer.cs.utsa.edu> soro...@lonestar.utsa.edu (Sean J. Orourke) writes:
>[lots of valid and some petty criticisms of Brian's Tidbits post deleted]

>
>Can anyone who has as many magazine subscriptions as Brian apparently has
>say whether or not most of his stuff comes from rumour pages? I noticed that
>quite a few of the page numbers are rather small; back when I subscribed to
>Byte, the first couple of pages were the rumour area. I'm curious as to
>whether or not this is true of other magazines.

Sorry, but InfoWorld and PC Week do not put rumors on the early pages.
They're in the back. His data is almost all from normal press coverage.

>
>Sean
>

[-------------------------------------------------------------------------]
[ Scott Cantor UNIX is a small, fast, efficient ]
[ Ohio State University operating system upon which every ]
[ Columbus, Ohio graduate student in the universe has ]
[ played pin the tail on the donkey. ]
[ can...@cis.ohio-state.edu ---- a UNIX guru ]
[-------------------------------------------------------------------------]

My name is...

unread,
Mar 13, 1993, 4:05:04 PM3/13/93
to

In article <1993Mar9.1...@imada.ou.dk>, tfr...@imada.ou.dk (Torben Friis) writes...

Luca Polo

unread,
Mar 15, 1993, 4:59:45 AM3/15/93
to
In article <1993Mar9.1...@imada.ou.dk> tfr...@imada.ou.dk (Torben Friis) writes:
>thom...@kuhub.cc.ukans.edu writes:
>
>>In article <1ngh6o...@network.ucsd.edu>, m...@lyapunov.ucsd.edu (Matt Kennel) writes:
>>> cki...@cs.umr.edu (Charles Kincy) writes:
>>> : In article <C3Iys...@ms.uky.edu> br...@easy.cs.utah.edu (Brian Sturgill) writes:
>>>
>>> Can you think of many truly CPU-intensive programs that don't
>>> use floating point?
>>>
>>Windows, OS/2?
>
>>Almost any database program?
>
>Compilers, filesystems (server)?
>
Fractint ? :-) :-)

Luca Polo.
--
+------------------------+ "Or mirate la vostra semenza:
|Luca Polo | fatti non foste, a viver come bruti,
|ja...@paola.dei.unipd.it | ma per seguir virtute e canoscenza."
+------------------------+ (Dante Alighieri, "Divina Commedia")

0 new messages