[comp.sys.next] SPARCstation 2 --> workstation wars

daugher@cs.tamu.edu (Dr. Walter C. Daugherity) (10/18/90)

According to the October 1st issue of Computer Reseller News, "Sun Microsystems
Inc. is expected to unveil in the next six weeks its next-generation SPARC-
station 2 computer...27 MIPS...6 to 8 MFLOPS...priced about the same as Sun's
current 1+ machines."

What with the 68040 NeXT's, DEC's 5000, and IBM's RS/6000 (do I hear 7000?),
we're getting a new price/performance leader every month!  Come on, NeXT,
how about announcing a 50 MHz 68050???

Seriously, it is clear that the workstation wars will continue for the 
forseeable future, escalating at an ever-increasing rate and forcing a
shakeout (e.g., HP/Apollo).  Consequently, knowledgable users will put less
emphasis on horsepower (as long as it's adequate) and more emphasis on
compatibility, upgradability, etc.

Thus, I'd like to see NeXT make a commitment to providing an upgrade path for
the NeXTstation and NeXTstation color, even if it's just to say "We intend to
offer a board swap when the NeXT 3 arrives."  Otherwise people may hesitate to
buy a non-upgradable computer--anybody want my Sun 3/50 real cheap? :-)


-------------------------------------------------------------------------------
Walter C. Daugherity			Internet, NeXTmail: daugher@cs.tamu.edu
Knowledge Systems Research Center	uucp: uunet!cs.tamu.edu!daugher
Texas A & M University			BITNET: DAUGHER@TAMVENUS
College Station, TX 77843-3112		CSNET: daugher%cs.tamu.edu@RELAY.CS.NET
	---Not an official document of Texas A&M---

melling@cs.psu.edu (Michael D Mellinger) (10/18/90)

   Seriously, it is clear that the workstation wars will continue for the 
   forseeable future, escalating at an ever-increasing rate and forcing a
   shakeout (e.g., HP/Apollo).  Consequently, knowledgable users will put less
   emphasis on horsepower (as long as it's adequate) and more emphasis on
   compatibility, upgradability, etc.

   Thus, I'd like to see NeXT make a commitment to providing an
upgrade path for the NeXTstation and NeXTstation color, even if it's
just to say "We intend to offer a board swap when the NeXT 3 arrives."
Otherwise people may hesitate to buy a non-upgradable
computer--anybody want my Sun 3/50 real cheap? :-)


   -------------------------------------------------------------------------------
   Walter C. Daugherity			Internet, NeXTmail: daugher@cs.tamu.edu
   Knowledge Systems Research Center	uucp: uunet!cs.tamu.edu!daugher
   Texas A & M University			BITNET: DAUGHER@TAMVENUS
   College Station, TX 77843-3112		CSNET: daugher%cs.tamu.edu@RELAY.CS.NET
	   ---Not an official document of Texas A&M---


This brings up an interesting question.  Should we buy cubes or slabs?
The cube will be more upgradeable.  You should just be able to drop in
another CPU board and upgrade to the next generation CPU.  But it
costs $2000($3000??) more than the slab.  So, do we pay $2300 for the
slab now, and buy another slab in 2 years because there isn't an
upgrade path, or do we spend the extra money for a Cube now(the
difference will almost by a new slab in two years)?

NeXT shouldn't have too much trouble being competitive.  Motorola
announced that the next generation of the 88000(its RISC) chip will
perform somewhere between 60-100mips when running at 25Mhz.  And I'm
sure NeXT has been listening to my advice on using a RISC in their
NeXT generation machine.:-) NeXT doesn't have to make faster machines
than Sun as long as they have better system software.  If speed was
all that mattered, then Sun would have crushed Apple a long time ago.

-Mike

BTW. I think the reason that a Sun 3/50 can't be upgraded is a
political decision brought to you by Sparc Inc.(aka Sun)

scott@nic.gac.edu (10/18/90)

   Seriously, it is clear that the workstation wars will continue for the
   forseeable future, escalating at an ever-increasing rate and forcing a
   shakeout (e.g., HP/Apollo).  Consequently, knowledgable users will put less
   emphasis on horsepower (as long as it's adequate) and more emphasis on
   compatibility, upgradability, etc.

This is why I'd like to see a less hardware-dependant NextStep.  NeXT could
move from hardware/software to more of a dependance on software - which is
where the money is.  NeXT could really clean up if they were running
NeXTStep on Suns and {Dec,VAX}Stations, IMHO.  I don't know that this
must be to the exclusion of hardware - the new NeXTStations are pretty
awesome, if you just compare hardware-wise (without discussion of cost
of software, and how much memory it really needs).  If Apple came out
with a comparable machine (a Mac with 105M, 8M, '040, etc) and offered
it for $3000 min to educational institutions, they would sell like
hotcakes.

Of course, Apple is not doing that.

I've heard that NeXT will work with anyone who wants to license NextStep
and DPS to get it ported to new machines.  This is great, this is good.
In nearly the same paragraph, I basically heard that they expected
schools to do it (it wasn't an outright statement, but that is what
it sounded like to me).  This is not the way to go.  I think that
NeXTStep would at least port fairly easily to Sun-[34], Mac IIs, and
VAXStations.  Sparcs and DECStations (how about Personal Iris!) would
be another story, but quite doable.

I think NeXT should do this.  We are already seeing that NeXT is being
forced to run a thin line between the PC world and the Workstation
world.  Why not, instead, concentrate on blurring the distinctions
by porting NextStep to high-end PCs and all workstations?  Besides,
that'd give me a greater market . . . :-)

scott hess
scott@gac.edu
Independent NeXT Developer	(Stuart)
NeXT Campus Consultant		(Not much, really)
GAC Undergrad			(Horrid.  Simply Horrid.  I mean the work!)
<I still speak for nobody>

sysmgr@KING.ENG.UMD.EDU (Doug Mohney) (10/18/90)

In article <Fvo.hft2@cs.psu.edu>, melling@cs.psu.edu (Michael D Mellinger) writes:

>than Sun as long as they have better system software.  If speed was
>all that mattered, then Sun would have crushed Apple a long time ago.

Sun and Apple are not in direct competition. Sun has a "workstation" niche
and has never, ever, tried to market a "home" machine for "The rest of us."
Besides, the pre-1990 prices on Sun vs Apple aren't even close to comparable.
You're comparing...well...Apples and oranges.

>BTW. I think the reason that a Sun 3/50 can't be upgraded is a
>political decision brought to you by Sparc Inc.(aka Sun)

Well, this is a myth. I've heard that Sun will give you a $1500 trade-in
credit for your old machine towardsa a Sun 4 *AND* you get to keep your
Sun 3/50. 

Now, what is NeXT going to do with all those old '030 boards?

 

fozzard@alumni.colorado.edu (Richard Fozzard) (10/19/90)

In article <9010180146.AA09967@mcs-server.gac.edu> scott@nic.gac.edu writes:
>
>This is why I'd like to see a less hardware-dependant NextStep.  NeXT could
>move from hardware/software to more of a dependance on software - which is
>where the money is.  NeXT could really clean up if they were running
>NeXTStep on Suns and {Dec,VAX}Stations, IMHO.  

3 cheers! I have said this ever since NeXTStep appeared - the fact that
apps I develop on a NeXT won't run on all the Macs and Suns we have around
here is what has kept me from being able to convince anyone here to
approve buying any machines. Just like UNIX, X-windows is taking over
everywhere (against the better judgement of right-thinking folks :-)

>If Apple came out
>with a comparable machine (a Mac with 105M, 8M, '040, etc) and offered
>it for $3000 min to educational institutions, they would sell like
>hotcakes.
>
>Of course, Apple is not doing that.
>
Now, now, let's be realistic. As anyone who works with both NeXTs and Macs
knows, a "comparable machine" is not one with the same CPU, RAM, and HD.
A 25Mhz '030 cube with a 300MB HD and 8MB RAM launches and runs apps only
slightly faster than an 8Mhz '000 Mac SE (and far slower than a 16Mhz
'020 Mac II), and 4MB RAM + 100MB HD on a Mac will hold
as many simultaneously running apps and store as much as on the cube. Of
course, the cube will have UNIX; but wasn't the whole point of NeXTStep
to *hide* us from UNIX?

I don't want to get into a flame war of Mac vs NeXT (I like both, for
different reasons), but remember that Apple *is* offering two (arguably)
comparable machines, the '020 LC and the '030 IIsi (about $2000 and 
$3000 to education, respectively, with 3rd party large monochrome 
monitors). And they are already (unlike the slab, they are not waiting
for Motorola to release the '040) selling like hotcakes. Both of these
have an expansion slot and provide color capability for just the cost of 
a monitor - two strikes against the slab in educational markets.

It remains to be seen whether the slab will sell like hotcakes - such
were the predictions about the original cube, also.

>I've heard that NeXT will work with anyone who wants to license NextStep
>and DPS to get it ported to new machines.  This is great, this is good.
>In nearly the same paragraph, I basically heard that they expected
>schools to do it (it wasn't an outright statement, but that is what
>it sounded like to me).  This is not the way to go.  I think that
>NeXTStep would at least port fairly easily to Sun-[34], Mac IIs, and
>VAXStations.  Sparcs and DECStations (how about Personal Iris!) would
>be another story, but quite doable.

You seem to be a little confused here - why would it be easy to port
to a Sun-4, but not a "Sparc"? Sun-4s *are* "Sparc"s!

Aside from these minor disagreements, I echo your sentiments - I would
really like to see NeXTStep compatibility on something other than 
vaporware IBM AIX systems. Then maybe I could finally buy a NeXT 
or two for around here!
-- 
========================================================================
Richard Fozzard					"Serendipity empowers"
Univ of Colorado/CIRES/NOAA	R/E/FS  325 Broadway, Boulder, CO 80303
fozzard@boulder.colorado.edu                   (303)497-6011 or 444-3168

chris@island.uu.net (Chris King) (10/19/90)

In article <9010180146.AA09967@mcs-server.gac.edu> scott@nic.gac.edu writes:
>
>   Seriously, it is clear that the workstation wars will continue for the
>   forseeable future, escalating at an ever-increasing rate and forcing a
>   shakeout (e.g., HP/Apollo).  Consequently, knowledgable users will put less
>   emphasis on horsepower (as long as it's adequate) and more emphasis on
>   compatibility, upgradability, etc.
>
>This is why I'd like to see a less hardware-dependant NextStep.  NeXT could
>move from hardware/software to more of a dependance on software - which is
>where the money is.  NeXT could really clean up if they were running
>NeXTStep on Suns and {Dec,VAX}Stations, IMHO. 
	[ stuff about porting NeXTStep deleted ]
>
>I think NeXT should do this.  We are already seeing that NeXT is being
>forced to run a thin line between the PC world and the Workstation
>world.  Why not, instead, concentrate on blurring the distinctions
>by porting NextStep to high-end PCs and all workstations?  Besides,
>that'd give me a greater market . . . :-)


I disagree with this statement quite strongly. I think that engineers
like the idea of mix and match software and hardware, however the
general public does not. One very important lesson that I think that
the success of the Macintosh should have taught the computer industry
is that the mainstream of computer buyers don't want to buy a chunk of
hardware and a bunch of miscellaneous software. They want to buy a
appliance. Early Mac buyers just wanted a word processing appliance.
They didn't give a damn what hardware it ran on. But they wanted to go
to the store, bring it home (or into their office), have it set up in
a hour or two and have it doing useful stuff for them in less than a
day. Exaclty the same way you would expect a buyer to treat a
microwave, a VCR or a answering machine.

This is the market that NeXT is ultimatly after. A large percentange of
computer buyers have this attitude. It is very difficult to build a
product for this market unless you have a good degree of control over
both the hardware and the software and supply a well intergrated
complete prooduct. 

					Chris King
					Island Graphics Corp.
					San Rafael Ca.
					{sun,ucbcad,uunet}!island!chris

scott@NIC.GAC.EDU (10/19/90)

fozzard@alumni.colorado.edu (Richard Fozzard) writes:
>>If Apple came out
>>with a comparable machine (a Mac with 105M, 8M, '040, etc) and offered
>>it for $3000 min to educational institutions, they would sell like
>>hotcakes.
>>
>>Of course, Apple is not doing that.
>>
>Now, now, let's be realistic. As anyone who works with both NeXTs and Macs
>knows, a "comparable machine" is not one with the same CPU, RAM, and HD.
>A 25Mhz '030 cube with a 300MB HD and 8MB RAM launches and runs apps only
>slightly faster than an 8Mhz '000 Mac SE (and far slower than a 16Mhz
>'020 Mac II), and 4MB RAM + 100MB HD on a Mac will hold
>as many simultaneously running apps and store as much as on the cube. Of
>course, the cube will have UNIX; but wasn't the whole point of NeXTStep
>to *hide* us from UNIX?

You missed my point.  If Apple came out with an '040 Mac, with a 105M
hard drive, 8M of RAM, and an 1120x832 display, all for $3000 at max
discount, they would sell like hotcakes.  No?

   >I've heard that NeXT will work with anyone who wants to license NextStep
   >and DPS to get it ported to new machines.  This is great, this is good.
   >In nearly the same paragraph, I basically heard that they expected
   >schools to do it (it wasn't an outright statement, but that is what
   >it sounded like to me).  This is not the way to go.  I think that
   >NeXTStep would at least port fairly easily to Sun-[34], Mac IIs, and
   >VAXStations.  Sparcs and DECStations (how about Personal Iris!) would
   >be another story, but quite doable.

   You seem to be a little confused here - why would it be easy to port
   to a Sun-4, but not a "Sparc"? Sun-4s *are* "Sparc"s!

I am confused.  Must have got my sun-n's mixed with sun-n-1's.  Forget
I said that, too.  I've only ever played with sun-3, and didn't
particularily enjoy it.

>Aside from these minor disagreements, I echo your sentiments - I would
>really like to see NeXTStep compatibility on something other than 
>vaporware IBM AIX systems. Then maybe I could finally buy a NeXT 
>or two for around here!

Acutally, I was sort of wondering how an i486 system with superVGA and
a Weitek (or 34010 board) would do running NextStep.  Considering that
those beasts are gradually creeping out into the market . . . OK, compared
to NeXTs, those beasts are _stampeding_ into the market!

scott hess
scott@gac.edu
Independent NeXT Developer	(Stuart)
NeXT Campus Consultant		(Not much, really)
GAC Undergrad			(Horrid.  Simply Horrid.  I mean the work!)
<I still speak for nobody>

cnh5730@calvin.tamu.edu (Chuck Herrick) (10/19/90)

In article <0093E5EF.ABEDCFA0@KING.ENG.UMD.EDU> sysmgr@KING.ENG.UMD.EDU (Doug Mohney) writes:
>Now, what is NeXT going to do with all those old '030 boards?

Well, they could ship them to John Scummey (or is the spelling
Sculley?) over at apple. They could use a good lesson in quality.


-- 
	Chuck Herrick				cnh5730@calvin.tamu.edu

barry@pico.math.ucla.edu (Barry Merriman) (10/19/90)

In article <Fvo.hft2@cs.psu.edu> melling@cs.psu.edu (Michael D Mellinger) writes:

>NeXT doesn't have to make faster machines
>than Sun as long as they have better system software. 

Speaking of that---we just got the X11-NeWS for our Suns.
Its a major improvement in look and feel over X11 and suntools.
In fact, visually its a Mac-NeXTSteP hybrid (Mac icons, NeXTSteP
windows, buttons, sliders). Since its PostScript based, it looks
much nicer than X.

Functionally, it comes with some nice utilities---e.g. graphical
directory tree browser, graphical tools for setting preferences
and defaults, printing, etc. An ``Unknown''-like feature that automatically
does the right thing when you click on a file, a very nice datebook/calendar,
and some demos. Plus, you can transparently run X apps (modulo bugs).
You can drag & drop files, etc.

Its not as nice as NeXTSteP, though. It seems a bit slower, and
its not as friendly, slick, or bug free. On a scale of 1--10,
I give NeXTStep a 9, NeWS a 7, X a 5, and suntools a 4.

Still, my point is: NeWS and NeXTStep are the same order of magnitude
of goodness. Superficially, there is not that much practical difference.
(Of course, NeXT is vastly superior internally). So, will a NeXT
running NeXTStep really look much different than a Sparc running
NeWS to the masses? And NeWS gives you X compatibility.

I think NeXT should build X into NeXTStep the same way X is 
built into NeWS. It wouldn't hurt anything, and it would 
broaden their appeal somewhat.


--
Barry Merriman
UCLA Dept. of Math
UCLA Inst. for Fusion and Plasma Research
barry@math.ucla.edu (Internet)

scott@NIC.GAC.EDU (10/19/90)

cnh5730@calvin.tamu.edu (Chuck Herrick) writes:
>In article <0093E5EF.ABEDCFA0@KING.ENG.UMD.EDU> sysmgr@KING.ENG.UMD.EDU (Doug Mohney) writes:
>>Now, what is NeXT going to do with all those old '030 boards?
>
>Well, they could ship them to John Scummey (or is the spelling
>Sculley?) over at apple. They could use a good lesson in quality.

Huh?  I thought everyone already knew:  Apple's going to ship a bunch
of black Macs during April on '91, as a test.  Then, since NeXT is
currently going to sell a billion of the new machines, when the NeXT
NeXT comes out, they'll send all the newly old boards to Apple again.
Since by that time NeXT will be so far ahead of Apple (say, 1st quarter
'92) that this will easily take care of Apple's needs.

Hand-me downs, they're called.

Of course, I've also heard rumors that the people at Apple are going
to keep them all for their own offices . . .

[ Sprinkle smileys liberally throughout this post.  Here, I include
a handy supply for you :-) :-) :-) :-) :-) :-) :-) :-) :-) :-) :-) :-) ]

scott hess
scott@gac.edu
Independent NeXT Developer	(Stuart)
NeXT Campus Consultant		(Not much, really)
GAC Undergrad			(Horrid.  Simply Horrid.  I mean the work!)
<I still speak for nobody>

skankman@ux1.lbl.gov (David Schurig) (10/19/90)

In article <28337@boulder.Colorado.EDU> fozzard@alumni.colorado.edu (Richard Fozzard) writes:

>Now, now, let's be realistic. As anyone who works with both NeXTs and Macs
>knows, a "comparable machine" is not one with the same CPU, RAM, and HD.
>A 25Mhz '030 cube with a 300MB HD and 8MB RAM launches and runs apps only
>slightly faster than an 8Mhz '000 Mac SE (and far slower than a 16Mhz
>'020 Mac II), and 4MB RAM + 100MB HD on a Mac will hold
>as many simultaneously running apps and store as much as on the cube.

What the devil!!!!!!!!
Why is this the case? I have not used a NeXT yet, but I assumed from the
numbers that an 040 NeXT would blow away a IIfx ( at less than half the
cost. )  I was all set to buy a NeXTStation.  An 040 NeXT is going to be
slower than the IIci I am using now?  Is Unix such a burden for the blinding,
15MIPS 040, or is it that NeXT applications are so inefficient. What is the
deal?  I am going to require psychotherapy if the new NeXTs are comparable
to an SE.

Dav

smithw@hamblin.math.byu.edu (Wild Willy) (10/19/90)

skankman@ux1.lbl.gov (David Schurig) writes:
>Why is this the case? I have not used a NeXT yet, but I assumed from the
>numbers that an 040 NeXT would blow away a IIfx ( at less than half the
>cost. )  I was all set to buy a NeXTStation.  An 040 NeXT is going to be
>slower than the IIci I am using now?  Is Unix such a burden for the blinding,
>15MIPS 040, or is it that NeXT applications are so inefficient. What is the
>deal?  I am going to require psychotherapy if the new NeXTs are comparable
>to an SE.

It is not the case. I have used both machines, running big Mathematica
problems for instance.  There simply is no comparison in terms of
capability.  The IIfx and NeXTstation are not the same class of machine,
on any level I know about.  When it comes to that, I'll take my 030
cube over a IIfx for any purpose I can imagine.  Try doing a real
Mathematica compute sometime on the fx and watch is croak.  Sure,
windows scroll faster on the fx than the 030 cube but who cares??
So far I haven't found any other way the fx is superior to the old
cube. Of course if the main use you are going to put your machine to
is playing Dark Castle at home on wintry evenings, or filling out 
your 1040 go ahead and buy an SE.  There is no comparison in terms of
real computing capability between the NeXTstation and *any* Mac.
Just my opinion, of course.....

-Bill  
-- 
        Jan L. Peterson	-- note new e-mail address below.
EMail:  jlp@hamblin.math.byu.edu  or  uunet!hamblin.math.byu.edu!jlp
Mail:   Math Dept. -- 292 TMCB; BYU; Provo, UT 84602 (USA)
Phone:  +1 801 378 2183		FAX:  +1 801 378 2800

verket@venice.SEDD.TRW.COM (Paul Verket) (10/19/90)

From article <0093E5EF.ABEDCFA0@KING.ENG.UMD.EDU>, by sysmgr@KING.ENG.UMD.EDU (Doug Mohney):
> Well, this is a myth. I've heard that Sun will give you a $1500 trade-in
> credit for your old machine towardsa a Sun 4 *AND* you get to keep your
> Sun 3/50. 

Not quite, it's $1500 LIST trade in (gets discounted with everything else) and
you HAVE TO send back the 3/50. It gets sent back in the sparcstation carton,
gets an RMA number, instructions on how to modify the carton's foam to protect
the 3/50, etc.
> 
> Now, what is NeXT going to do with all those old '030 boards?

I too am curious what they (or sun with the 3/50's) do with them.
			Paul Verket

smithw@hamblin.math.byu.edu (Wild Willy) (10/20/90)

l
--
             ____________________ Bill Smith____________________.
EMail:  smithw@hamblin.math.byu.edu  or  uunet!hamblin.math.byu.edu!smithw
Mail:          Math Dept. -- 314 TMCB; BYU; Provo, UT 84602 (USA)
Phone:            +1 801 378 2126         FAX:  +1 801 378 2800
-- 
        Jan L. Peterson	-- note new e-mail address below.
EMail:  jlp@hamblin.math.byu.edu  or  uunet!hamblin.math.byu.edu!jlp
Mail:   Math Dept. -- 292 TMCB; BYU; Provo, UT 84602 (USA)
Phone:  +1 801 378 2183		FAX:  +1 801 378 2800

brtmac@maverick.ksu.ksu.edu (Brett McCoy) (10/20/90)

In <7669@dog.ee.lbl.gov> skankman@ux1.lbl.gov (David Schurig) writes:

>In article <28337@boulder.Colorado.EDU> fozzard@alumni.colorado.edu (Richard Fozzard) writes:

>>Now, now, let's be realistic. As anyone who works with both NeXTs and Macs
>>knows, a "comparable machine" is not one with the same CPU, RAM, and HD.
>>A 25Mhz '030 cube with a 300MB HD and 8MB RAM launches and runs apps only
>>slightly faster than an 8Mhz '000 Mac SE (and far slower than a 16Mhz
>>'020 Mac II), and 4MB RAM + 100MB HD on a Mac will hold
>>as many simultaneously running apps and store as much as on the cube.

>What the devil!!!!!!!!
>Why is this the case? I have not used a NeXT yet, but I assumed from the
>numbers that an 040 NeXT would blow away a IIfx ( at less than half the
>cost. )  I was all set to buy a NeXTStation.  An 040 NeXT is going to be
>slower than the IIci I am using now?  Is Unix such a burden for the blinding,
>15MIPS 040, or is it that NeXT applications are so inefficient. What is the
>deal?  I am going to require psychotherapy if the new NeXTs are comparable
>to an SE.

I've never used, touched, or even personally seen a NeXT so I can't say
how it might compare with a Mac, but I have used Macs and I use SPARCstation's
extensively and I can tell you that a SPARCstation completely and utterly
blows a Mac away.  Trying to compare an 8 Mhz 68000 Mac to a 20 Mhz SPARC
is one of the most rediculous things I've ever heard, and even a IIci, which
is the fastest Mac I've ever used, seems incredibly slow compared to my
SPARC.  Also, NeXT computers have virtual memory, something that a Mac
user has probably never heard of.  So it's applications are larger, only
the part of the program that is actually executing needs to be in memory,
unlike a Mac where the entire application has to be loaded before it will
run.  The amount of disk space you devote to swap space is the only limit.
Physical memory (RAM) doesn't control how many programs may be resident and
running at once, only how quickly they can run.  The Mac will never be able
to compare to machines like Sun workstations and NeXT's until they come out
with an OS that supports true multitasking, virtual memory and paging.

--
Every day you scream at me to turn the music low.
But if you keep on screaming, you'll make me deaf, ya know. -- Judas Priest

Brett McCoy	brtmac@maverick.ksu.ksu.edu	NetNews Dictator

sysmgr@KING.ENG.UMD.EDU (Doug Mohney) (10/20/90)

In article <878@venice.SEDD.TRW.COM>, verket@venice.SEDD.TRW.COM (Paul Verket) writes:

>Not quite, it's $1500 LIST trade in (gets discounted with everything else) and
>you HAVE TO send back the 3/50. It gets sent back in the sparcstation carton,
>gets an RMA number, instructions on how to modify the carton's foam to protect
>the 3/50, etc.

Hmmm...welllllll....I am not one to speak to how commercial accounts are
handled......

n8443916@unicorn.wwu.edu (John Gossman) (10/20/90)

In article <878@venice.SEDD.TRW.COM> verket@venice.SEDD.TRW.COM (Paul Verket) writes:
>From article <0093E5EF.ABEDCFA0@KING.ENG.UMD.EDU>, by sysmgr@KING.ENG.UMD.EDU (Doug Mohney):
>> Well, this is a myth. I've heard that Sun will give you a $1500 trade-in
>> credit for your old machine towardsa a Sun 4 *AND* you get to keep your
>> Sun 3/50. 
>
>Not quite, it's $1500 LIST trade in (gets discounted with everything else) and
>you HAVE TO send back the 3/50. It gets sent back in the sparcstation carton,
>gets an RMA number, instructions on how to modify the carton's foam to protect
>the 3/50, etc.
>> 
>> Now, what is NeXT going to do with all those old '030 boards?
>
>I too am curious what they (or sun with the 3/50's) do with them.
                               ^^^^^^^^^^^^^^^^^^^^^
Sun offices are all tied in to big networks, so everybody has a machine.
Obviously, you don't want your secretaries to be Word Processing on
a SparcSTATION 2, but neither do you want them to be using a PC Clone,
not when you manufacture the competition for PCs.  So they take SUN 3s,
refurbish them, update the software, and hide them away with secretaries
and loading clerks.  That's what I heard anyway.


--John Gossman
SoftSource.

Lovstrand@EuroPARC.Xerox.COM (Lennart Lovstrand) (10/20/90)

In article <2090@island.uu.net>, chris@island.uu.net (Chris King) writes:
> In article <9010180146.AA09967@mcs-server.gac.edu> scott@nic.gac.edu writes:
     [...]
> > This is why I'd like to see a less hardware-dependant NextStep.  NeXT could
> > move from hardware/software to more of a dependance on software - which is
> > where the money is.  NeXT could really clean up if they were running
> > NeXTStep on Suns and {Dec,VAX}Stations, IMHO. 
> 	[ stuff about porting NeXTStep deleted ]
> >
> > I think NeXT should do this.  We are already seeing that NeXT is being
> > forced to run a thin line between the PC world and the Workstation
> > world.  Why not, instead, concentrate on blurring the distinctions
> > by porting NextStep to high-end PCs and all workstations?  Besides,
> > that'd give me a greater market . . . :-)
> 
> I disagree with this statement quite strongly. I think that engineers
> like the idea of mix and match software and hardware, however the
> general public does not.

Alas, there are Customers out there (with a Capital C) that are either
bound by company policy or simply too conservative to buy computers of
any other brand than the "company approved" one.  Sun has already become
a bestseller at most universities that I know and they sure aren't going
to junk their investment just because there is a snazzier box on the
market.  However, these guys usually *don't* feel any strong obligations
to buy any particular software, so a portable NeXTStep environment would
seem to me to potentially be a very big product.  I mean, just look at
the level of the competition!  "X.11" as a *user* interface environment?
What a joke!  [This isn't even funny.]

> One very important lesson that I think that
> the success of the Macintosh should have taught the computer industry
> is that the mainstream of computer buyers don't want to buy a chunk of
> hardware and a bunch of miscellaneous software. They want to buy a
> appliance. Early Mac buyers just wanted a word processing appliance.
> They didn't give a damn what hardware it ran on. But they wanted to go
> to the store, bring it home (or into their office), have it set up in
> a hour or two and have it doing useful stuff for them in less than a
> day. Exaclty the same way you would expect a buyer to treat a
> microwave, a VCR or a answering machine.

I couldn't agree less with your general sentiment!  I don't know about
NeXTs specificially, but in my experience, installing your generic UNIX box
usually requires an MSc in Computer Science to get everything right.  This
includes everything from acquiring and setting up an IP subnet to loading
up a DNS server, getting the latest version of sendmail, ftpd, etc from the
Net and installing them, etc, etc.  This is just ridiculous in comparison
to the Mac (and Xerox XNS) world.

On the other hand, if you compare with the HiFi market you can easily find
customers for complete "systems" as well as separate units (tuner, CD player,
amplifier, etc).  I don't think the ability to mix-n-match should be ruled
out to quickly.

-- 
--Lennart <Lovstrand@EuroPARC.Xerox.COM>		R   _A _  N_   K
Rank Xerox EuroPARC, 61 Regent St, Cambridge, UK	\/ |_ |_) | | \/
Zany, Sun-4/60 at EuroPARC, SunOS Release 4(0.3c)-3	/\ |_ | \ |_| /\
TOPS-20 Command processor 7(86)-4 [alpha]		E u r o  P A R C

ric@cs.arizona.edu (Ric Anderson) (10/20/90)

In article <7669@dog.ee.lbl.gov>, skankman@ux1.lbl.gov (David Schurig) writes:
[stuff deleted]
>>Now, now, let's be realistic. As anyone who works with both NeXTs and Macs
>>knows, a "comparable machine" is not one with the same CPU, RAM, and HD.
>>A 25Mhz '030 cube with a 300MB HD and 8MB RAM launches and runs apps only
>>slightly faster than an 8Mhz '000 Mac SE (and far slower than a 16Mhz
>>'020 Mac II), and 4MB RAM + 100MB HD on a Mac will hold
>>as many simultaneously running apps and store as much as on the cube.
>
> What the devil!!!!!!!!
> Why is this the case? I have not used a NeXT yet, but I assumed from the
> numbers that an 040 NeXT would blow away a IIfx ( at less than half the
> cost. )  I was all set to buy a NeXTStation.  An 040 NeXT is going to be
> slower than the IIci I am using now?  Is Unix such a burden for the blinding,
> 15MIPS 040, or is it that NeXT applications are so inefficient. What is the
> deal?  I am going to require psychotherapy if the new NeXTs are comparable
> to an SE.
> Dav

I also have a IIci, and a 16MB NeXT (68030 cube) with the NeXT
300MB hard drive.  For getting things started, the Mac (with
an internal Quantum ProDrive-80S) really does blow the NeXT
away.  Most of this seems to be the relative disk performance,
rather than the CPU, per se.

I find the NeXT to be "slow" when starting an application, and
doing disk activity, but quite acceptable for cpu tasks.  I
don't think speed is a factor in a NeXT/Mac choice.  The systems
do different things, and the environment and application
requirements (e.g., does one need a word processor or a
spreadsheet or whatever, that is only on the Mac) should
drive the selection of hardware.

Compared to a Sun Sparcstation, The NeXT (again the 68030 original
cube) is a real performance dog.  But, the NeXT has the NeXT Step
interface, and the Sun doesn't.  That might be important in
some appications.

So, I guess for a Unix box, I personally would lean toward
the Sun Sparcstation - disk access (Quantum Prodrive-105S
as shipped by Sun) is good, and cpu speed is also good.

For non-Unix box, I'd take the Mac over the NeXT (which is
why I have a IIcx), since it "feels" faster, and, more importantly,
has the tools I use every day (word processor, etc.), and (via
an ethernet card and NCSA telnet) lets me run a multiple window
environment to keep up the the Vax/Sun/Sequent systems I have
system-management responsibilities on.

Application software to do your job should always be considered
ahead of raw speed.   If you are going to write applications to 
do something special for yourself, then disk access, and overall
performance, are much more important than raw cpu speed, since
compiles (which one does a lot of during the debugging phase)
are more likely to be disk intensive than CPU bound.

Just my 2 cents worth,
Ric

Ric Anderson                    Bitnet: Ric@Arizrvax
Member of the Technical Staff   Internet: ric@cs.arizona.edu
University of Arizona           UUCP: uunet!arizona!ric
Department of Computer Science  AT&T: (602) 621-4048
Gould-Simpson Room 721
Tucson, Arizona 85721

jmann@angmar.sw.stratus.com (Jim Mann) (10/20/90)

In article <1990Oct19.181008.24196@maverick.ksu.ksu.edu>,
brtmac@maverick.ksu.ksu.edu (Brett McCoy) writes:
|>
|>I've never used, touched, or even personally seen a NeXT so I can't say
|>how it might compare with a Mac, but I have used Macs and I use
SPARCstation's
|>extensively and I can tell you that a SPARCstation completely and utterly
|>blows a Mac away.  Trying to compare an 8 Mhz 68000 Mac to a 20 Mhz SPARC
|>is one of the most rediculous things I've ever heard, and even a IIci, which
|>is the fastest Mac I've ever used, seems incredibly slow compared to my
|>SPARC.  Also, NeXT computers have virtual memory, something that a Mac
|>user has probably never heard of.  So it's applications are larger, only
|>the part of the program that is actually executing needs to be in memory,
|>unlike a Mac where the entire application has to be loaded before it will
|>run.  The amount of disk space you devote to swap space is the only limit.
|>Physical memory (RAM) doesn't control how many programs may be resident and
|>running at once, only how quickly they can run.  The Mac will never be able
|>to compare to machines like Sun workstations and NeXT's until they come out
|>with an OS that supports true multitasking, virtual memory and paging.
|>
Define "fast."  Sure, a SPARC-station is faster than a Mac or 68030 NeXT
if what you mean by fast is "how long does it take to compile program X" or
whatever.  If by "fast" you mean "how long does it take user X to do A,
then go over and do B for a while, then learn new application C, then
go back and paste some stuff from B to A" the Mac and the NeXT are both
faster than the Sun.  And if by "fast" you mean "how long does it take
to develop an application that does Z" well, the NeXT blows both machines
away.                                          

Jim Mann
Stratus Computer
jmann@es.stratus.com

phd_ivo@gsbacd.uchicago.edu (10/20/90)

Yes, there are points in favor of NeXT and points in favor of Mac. Yes,
starting a small spreadsheet or wordprocessor, a Mac feels faster than a Unix
'030 machine. Yes, a Unix '030 does give you something for the performance
penalty.

We have had all these arguments a thousand times on this net already. I don't
think it'll help a lot to repeat it again.

/ivo welch	ivo@next.agsm.ucla.edu

asd@mace.cc.purdue.edu (Kareth) (10/20/90)

In <165@cheltenham.cs.arizona.edu> ric@cs.arizona.edu (Ric Anderson) writes:

>> 15MIPS 040, or is it that NeXT applications are so inefficient. What is the
>> deal?  I am going to require psychotherapy if the new NeXTs are comparable
>> to an SE.
>> Dav

>I also have a IIci, and a 16MB NeXT (68030 cube) with the NeXT
>300MB hard drive.  For getting things started, the Mac (with
>an internal Quantum ProDrive-80S) really does blow the NeXT
>away.  Most of this seems to be the relative disk performance,
>rather than the CPU, per se.

Well, from what I learned at our Interface Builder course, this is not
the case.  Due to the object oriented nature of NeXTstep programming,
a LOT of setup is left for run time.  The NeXT dude was saying that
was why the NeXT was/is a dog at loading programs was because much of
initialization and things were left to runtime.

Of course there is 2.0.  I've seen/worked on a NeXT with beta versions
of 2.0, on an 030 Cube.  Response time, as far as starting programs,
looked to me to be twice as fast if not better.

>Application software to do your job should always be considered
>ahead of raw speed.   If you are going to write applications to 
>do something special for yourself, then disk access, and overall
>performance, are much more important than raw cpu speed, since
>compiles (which one does a lot of during the debugging phase)
>are more likely to be disk intensive than CPU bound.

The through-put factor.  Now if only NeXT would put a graphics
coprocessor on the main board to handle the tons of graphics info the
NeXT sloshes around.

brtmac@maverick.ksu.ksu.edu (Brett McCoy) (10/20/90)

In <579@roo.UUCP> Lovstrand@EuroPARC.Xerox.COM (Lennart Lovstrand) writes:

>I couldn't agree less with your general sentiment!  I don't know about
>NeXTs specificially, but in my experience, installing your generic UNIX box
>usually requires an MSc in Computer Science to get everything right.  This
>includes everything from acquiring and setting up an IP subnet to loading
>up a DNS server, getting the latest version of sendmail, ftpd, etc from the
>Net and installing them, etc, etc.  This is just ridiculous in comparison
>to the Mac (and Xerox XNS) world.

While I'll agree that setting up a UNIX box takes more brains than your
average Mac user has, it isn't impossible by any means.  Even I, a lowly
ugrad, can set up and install a system from the ground up.  It does take
more work and investment of time than most people are willing to impart
though.  But, all of the things you complain about taking so much time
and effor (i.e. sendmail, ftpd, DNS, etc.) aren't even available for your
average Mac, so why compare them.  As far as setting up a single user,
standalone Sun goes, just stick the installation tape in the drive, sit
down with your installation manual and turn it on.  It takes a few hours
to get it up and running and doing everything you want it to do, but it
isn't impossible.  Most systems you buy today don't even have to have that
done to them.  If you buy a SS1 with an internal drive you will get the
OS preinstalled for you.  All you have to do is go through and change the
site specific stuff.  It may take longer to set up, but the difference
between what an average Sun can do and what an average Mac can do is
ridiculous in comparison, especially considering the price comparison.

--
Every day you scream at me to turn the music low.
But if you keep on screaming, you'll make me deaf, ya know. -- Judas Priest

Brett McCoy	brtmac@maverick.ksu.ksu.edu	NetNews Dictator

brtmac@maverick.ksu.ksu.edu (Brett McCoy) (10/20/90)

In <2793@lectroid.sw.stratus.com> jmann@angmar.sw.stratus.com (Jim Mann) writes:

>Define "fast."  Sure, a SPARC-station is faster than a Mac or 68030 NeXT
>if what you mean by fast is "how long does it take to compile program X" or
>whatever.  If by "fast" you mean "how long does it take user X to do A,
>then go over and do B for a while, then learn new application C, then
>go back and paste some stuff from B to A" the Mac and the NeXT are both
>faster than the Sun.  And if by "fast" you mean "how long does it take
>to develop an application that does Z" well, the NeXT blows both machines
>away.                                          

"Fast" is in the eye of the beholder.  I would never even consider buying
a Mac since it doesn't have a real OS (by that I mean virtual memory,
multitasking, paging), but as for comparing a SPARC against a NeXT when
it comes to developing software I don't have any comparison since I've
never used a NeXT.  I do know that most of the things I want to do somebody
out on the net has usually already done it or done something close enough
that I can just modify it to do what I want it to do.  Nearly every program
that exists on the anonymous ftp archives will compile and run on a Sun
without to much trouble.  Can that be said of a NeXT?  And if you want to
use the complaint that the normal mortal can't use all of what is on the
net then I'll say that the normal mortal isn't going to be developing
software, they are going to be using what exists, and they are going to
be looking at how fast machine X does job A, not at how long it takes to
develop application A, and when it comes to actually performing the task,
a SPARC is going to blow away a Mac or '30 NeXT, and will probably outperform
a '40 NeXT.

--
Every day you scream at me to turn the music low.
But if you keep on screaming, you'll make me deaf, ya know. -- Judas Priest

Brett McCoy	brtmac@maverick.ksu.ksu.edu	NetNews Dictator

gft_robert@gsbacd.uchicago.edu (10/20/90)

----- 
In article <1990Oct20.004404.28918@maverick.ksu.ksu.edu>, brtmac@maverick.ksu.ksu.edu (Brett McCoy) writes...
[...]
>"Fast" is in the eye of the beholder.  I would never even consider buying
>a Mac since it doesn't have a real OS (by that I mean virtual memory,
>multitasking, paging), but as for comparing a SPARC against a NeXT when

Just to get it straight: the Mac will have virtual memory in System 7. 
(Actually, you can get it now.  I'm running VM on a lowly Mac II w/PMMU right
now with the help of Virtual.  But that's not built into the OS, so that's
not really addressing the OS point above).

The Mac also has multi-tasking.  It is cooperative multitasking, not
pre-emptive.  I know which one you prefer -- and there are good reasons for
that --  but they ARE both multitasking.  If you disagree with that, please
direct followups to comp.sys.vm.here.we.go.again. 

One other thing the Mac OS doesn't have: protected memory.  That seems to me to
be more important than pre-emptive multitasking.

Of course you can always run A/UX and get all of the above goodies, even w/o
waiting for 7.0.  But of course you wouldn't, 'cause Apple is Evil and Steve is
God, right? :->

Robert

	


============================================================================
= gft_robert@gsbacd.uchicago.edu * generic disclaimer: * "It's more fun to =
=            		         * all my opinions are *  compute"         =
=                                * mine                *  -Kraftwerk       =
============================================================================

eps@toaster.SFSU.EDU (Eric P. Scott) (10/20/90)

In article <1990Oct20.004404.28918@maverick.ksu.ksu.edu>
	brtmac@maverick.ksu.ksu.edu (Brett McCoy) writes:
>                                                       Nearly every program
>that exists on the anonymous ftp archives will compile and run on a Sun
>without to much trouble.  Can that be said of a NeXT?

Yes!  And in my experience it's been easier to compile random
net sources on NeXTs than Sun-4s... and that was before 4.1
"broke" lots of things that worked under 4.0.3.

					-=EPS=-

gft_robert@gsbacd.uchicago.edu (10/20/90)

-------
In article <1990Oct20.050207.1910@maverick.ksu.ksu.edu>, brtmac@maverick.ksu.ksu.edu (Brett McCoy) writes...
[...]
>There is the problem of the price of A/UX, especially when you add it to the
>cost of a IIci or IIfx.  You can go out an buy a SPARC loaded to the hilt
>for what a basic A/UX system costs. 

I don't think so: A/UX will run on an SE/30, and possibly an LC (there are
already '030 cards for the LC, I've heard).  But yes, A/UX is an additional
operating system for the Mac, and is thus an additional cost.


> Besides, do all of the Mac programs
>un under A/UX, and does the Mac retain all of its GUI functions when running
>A/UX?  (This is really a question, I don't know the answer)

I'm not an expert (check out comp.unix.aux for definitive answers).  From using
A/UX a lot at a client's, all Mac applications seem to work fine.  Certainly
the commercial one's I've used.  Of course they should be "32 bit clean", but
you can run A/UX in 24-bit mode for iffy apps.  The only problems I had were
with a scanner.

Yes, the Mac retains all GUI functions under A/UX.  In addition you can have
lots of command shell windows floating around for unix CLI stuff.  It really
looks very similar to the Next, except w/o the 3D effects.  You can also boot
A/UX to run simply in tty mode, but that's really only useful for debugging
(i.e, why would you want to?).

One neat thing IMHO is the seamless melding of the unix and Mac environments. 
You can copy files back and forth between unix and Mac formatted disks just as
if you were running under the Mac OS.  But anyway, check out comp.unix.aux if
you're interested.

Robert

============================================================================
= gft_robert@gsbacd.uchicago.edu * generic disclaimer: * "It's more fun to =
=            		         * all my opinions are *  compute"         =
=                                * mine                *  -Kraftwerk       =
============================================================================

brtmac@maverick.ksu.ksu.edu (Brett McCoy) (10/20/90)

In <1990Oct20.025628.11602@midway.uchicago.edu> gft_robert@gsbacd.uchicago.edu writes:

>The Mac also has multi-tasking.  It is cooperative multitasking, not
>pre-emptive.  I know which one you prefer -- and there are good reasons for
>that --  but they ARE both multitasking.  If you disagree with that, please
>direct followups to comp.sys.vm.here.we.go.again. 

I know that the multifinder exists and I've played with it a little bit
and for the most part it seems to work.  I haven't done much more than
just play so I can't pass judgement on it.

>One other thing the Mac OS doesn't have: protected memory.  That seems to me
>to be more important than pre-emptive multitasking.

Protected memory is a must for a good pre-emtive multitasking system to
work, especially if people writing the code aren't worrying about following
all of the guidlines.

>Of course you can always run A/UX and get all of the above goodies, even w/o
>waiting for 7.0.  But of course you wouldn't, 'cause Apple is Evil and Steve
>is God, right? :->

There is the problem of the price of A/UX, especially when you add it to the
cost of a IIci or IIfx.  You can go out an buy a SPARC loaded to the hilt
for what a basic A/UX system costs.  Besides, do all of the Mac programs
un under A/UX, and does the Mac retain all of its GUI functions when running
A/UX?  (This is really a question, I don't know the answer)
--
Every day you scream at me to turn the music low.
But if you keep on screaming, you'll make me deaf, ya know. -- Judas Priest

Brett McCoy	brtmac@maverick.ksu.ksu.edu	NetNews Dictator

cnh5730@calvin.tamu.edu (Chuck Herrick) (10/20/90)

In article <2090@island.uu.net> chris@island.uu.net (Chris King) writes:
>.... I think that engineers
>like the idea of mix and match software and hardware, however the
>general public does not...
>...
>This is the market that NeXT is ultimatly after...

     No, this is the market that you THINK that NeXT is after. Big diff
here. I THINK that what NeXT is all about is "having your cake and
eating it too." 
     Please, let us avoid forcing NeXT onto one side or
the other of some seemingly dichotomous issue. Look for the solution
that avoids choice.

-- 
	Chuck Herrick				cnh5730@calvin.tamu.edu

yoo@well.sf.ca.us (Young-Kyu Yoo) (10/21/90)

Why all this talk about about Mac/NeXT/Sun comparisons when NeXT 2.0 has
been announced and should be three to four times faster than the old
NeXT?  Anyone worried about the old NeXT's performance will upgrade to
2.0.  Yes, the old NeXT was slow, but the new NeXTs are speed demons.
I've had a chance to play with them.  

Here are some facts.  The Mac IIfx runs at 7-8 Mips.  The Sun SLC at 12-13
Mips.  The Sparc+ at 15-16 Mips.  The 040 NeXTs are at 15 Mips (or 20 if
you believe Motorola).  The 040 NeXTs do 2.5-3 MFLOPS.  A Mathematica
benchmark showed that the NeXTs were faster in this area than the Suns and
several times faster than any of the Macs.

In terms of program loading, the 2.0 software alone (without the 040 chip)
improves performance by a factor of 2 or 3. In terms of program execution,
because of the vastly improved floating point performance, postscript is
fast and, thus, drawing on screen is fast.  Some PostScript routines also
have been optimized.  xyshow, for instance, is an order of magnitude faster
on the new NeXTs.  

My general impression is that the new NeXTs are significantly faster than
any Mac.  They are comparable to the Sun.  And in the one true benchmark that
counts (i.e., how much work can the user do in a given amount of time), the
NeXT wins hands down.


 

mikel@Apple.COM (Mikel Evins) (10/22/90)

In article <1990Oct20.050207.1910@maverick.ksu.ksu.edu> brtmac@maverick.ksu.ksu.edu (Brett McCoy) writes:
>There is the problem of the price of A/UX, especially when you add it to the
>cost of a IIci or IIfx.  You can go out an buy a SPARC loaded to the hilt
>for what a basic A/UX system costs.  Besides, do all of the Mac programs
>un under A/UX, and does the Mac retain all of its GUI functions when running
>A/UX?  (This is really a question, I don't know the answer)

All Mac applications do not run under A/UX. Lots do. A/UX provides
a "compatability" mode to make life easier for those who are running
ill-behaved software. Most of the problem arises from the fact that,
historically, Macs used a 24-bit address space. Since the 68000 family
uses 32-bit addresses, this left eight bits unaccounted for. Many
Mac applications used the unused eight bits for their own purposes.
Naturally, such apps will behave strangely in a 32-bit operating system.
The "compatability" mode (it isn't really called that; I forget the
official terminology) uses 24-bit addressing to improve the behavior
of apps that are not "32-bit clean". 32-bit cleanliness is a major
issue in the Mac world right now, because the forthcoming Mac OS
version 7.0 demands it, so more and more commercial apps are becoming
32-bit clean.

A/UX 2.0 has a completely mac-like interface. As a matter of fact, a
naive user probably would not be able to tell from looking at a Mac
whether it is running Mac OS or A/UX. For knowlegeable users, there are
some clues, however. For example, the A/UX Finder has a menu item
that opens a command shell window from which you can execute commands
with your favorite shell. (A/UX includes Korn, Bourne, and C). One
thing that I particularly like about A/UX as compared to other
Unix systems is Commando. Commando is a tool that was swiped from 
Apple's MPW development system. It presents a command-line command
as a dialog box. The dialog shows all the commands flag options
as buttons, menus, checkboxes, sliders, radio buttons, and so on so
that you can build a useful command line without knowing the flag
options in advance. The command line appears in a text box at the
bottom of the dialog, and you can transfer it to your shell window
on execution, so that it appears in your history, should you want
to use it again sometime. I think this is really handy for those
commands that you don't use all the time. Oh, and, no you don't
have to use Commando; you can just type the commands that you do
know well.

The issue of price is still a problem, of course. Also, a Mac IIfx
is not going to perform like a SPARCStation or a NeXTStation. Always
tradeoffs.

melling@cs.psu.edu (Michael D Mellinger) (10/23/90)

In article <0093E5EF.ABEDCFA0@KING.ENG.UMD.EDU> sysmgr@KING.ENG.UMD.EDU (Doug Mohney) writes:

   In article <Fvo.hft2@cs.psu.edu>, melling@cs.psu.edu (Michael D Mellinger) writes:

   >than Sun as long as they have better system software.  If speed was
   >all that mattered, then Sun would have crushed Apple a long time ago.

   Sun and Apple are not in direct competition. Sun has a "workstation" niche
   and has never, ever, tried to market a "home" machine for "The rest of us."
   Besides, the pre-1990 prices on Sun vs Apple aren't even close to comparable.
   You're comparing...well...Apples and oranges.

I didn't realize Apple sold all(or even most) of their computers to
the home market.  Until Oct. 15 Macs were priced out of the average
consumer's budget.  I was talking about the business market.  Sun got
Lotus and Ashton-Tate(dBase) to port their wares to their computers so
that they could get part of the business market; that's where they see
their future growth coming from.  Sun does sell a Sparc Station for
$5000, and I would say that you get considerably more performance out
of it than a $5000 Mac.  Are we talking Apples and Apples yet?

-Mike

glang@Autodesk.COM (Gary Lang) (10/23/90)

>apps I develop on a NeXT won't run on all the Macs and Suns we have around
>here is what has kept me from being able to convince anyone here to
>approve buying any machines.

So your Sun software runs on your Macs, eh? Sure it does.

Has anyone had a chance to look at the new Btree stuff in 2.0? Is there a
.nib object interface to it? COuld you email me the protocols for it so that
I can plan on using it?

Thanks,
g

Lovstrand@EuroPARC.Xerox.COM (Lennart Lovstrand) (10/23/90)

In article <1990Oct20.003611.28732@maverick.ksu.ksu.edu>, brtmac@maverick.ksu.ksu.edu (Brett McCoy) writes:
> In <579@roo.UUCP>  Lovstrand@EuroPARC.Xerox.COM (Lennart Lovstrand) writes:
> > I couldn't agree less with your general sentiment!  I don't know about
[Hrrm, that should have been "more" of course.  But then again, "less" is
 "more", right? (Or so the doc says at least. ;-)]
> > NeXTs specifically, but in my experience, installing your generic UNIX box
> > usually requires an MSc in Computer Science to get everything right.  This
> > includes everything from acquiring and setting up an IP subnet to loading
> > up a DNS server, getting the latest version of sendmail, ftpd, etc from the
> > Net and installing them, etc, etc.  This is just ridiculous in comparison
> > to the Mac (and Xerox XNS) world.
>
> While I'll agree that setting up a UNIX box takes more brains than your
> average Mac user has, it isn't impossible by any means.  Even I, a lowly
> ugrad, can set up and install a system from the ground up.  It does take
> more work and investment of time than most people are willing to impart
> though.  But, all of the things you complain about taking so much time
> and effor (i.e. sendmail, ftpd, DNS, etc.) aren't even available for your
> average Mac, so why compare them.

Hmm, is that really true?  There are name lookup services on Appletalk
(NBP), as well as print services, file services, and mail services.  And
if you want to claim that they are limited in scope and function, I won't
argue with you but rather point you to the Xerox Network Services which
support all of this and more over very large internetworks (10,000+ nodes)
but with comparatively very little administration.  [It's by no means
perfect, but it serves well as an example.]

To get a new Mac or Xerox Workstation -- or server -- set up on a net,
all you typically need to do is plug it in and maybe give it a name.
It acquires the net and node numbers from the network directly and starts
advertising its service automatically.  Not so with any UNIX boxes I know.

Look, I'm not saying that you *have* to have a degree to do this stuff and
I'm sure you're doing a great job at it. But try giving it to Joe Random
Luser and watch him sweat (and cry!) over setting up UUCP links and/or
sendmail.cf for example.  All I'm arguing is that there is no reason why
it should be as hard as it is and require as much knowledge about
networking etc as it currently does to set up a networked computer.

> As far as setting up a single user,
> standalone Sun goes, just stick the installation tape in the drive, sit
> down with your installation manual and turn it on.

Sure, if it's a standalone machine...  But then plug it in on a network
and watch hell break lose when duplicate addresses and broadcast storms
start flourishing.  OK, I don't know -- maybe all Suns don't come with the
same IP address and use all-zeroes for broadcasts any more.  It doesn't
matter, I'm sure there's something else broken instead ;-).

> It may take longer to set up, but the difference
> between what an average Sun can do and what an average Mac can do is
> ridiculous in comparison, especially considering the price comparison.

Yup, CPU/OS-wise I agree with you, but how about comparing what your
average Sun and Mac *user* can do instead?  The machine and the user
aren't necessarily the same, you know.

I guess another way of putting it is that there is a great number of
non-hacker/computer science users out there who could benefit from high
powered workstations -- if they just were packaged the right way.

[My opinion is not necessarily the same as Xerox' etc etc]
-- 
--Lennart <Lovstrand@EuroPARC.Xerox.COM>		R   _A _  N_   K
Rank Xerox EuroPARC, 61 Regent St, Cambridge, UK	\/ |_ |_) | | \/
Zany, Sun-4/60 at EuroPARC, SunOS Release 4(0.3c)-3	/\ |_ | \ |_| /\
TOPS-20 Command processor 7(86)-4 [alpha]		E u r o  P A R C

mikel@Apple.COM (Mikel Evins) (10/24/90)

In article <Fsb*zjx2@cs.psu.edu> melling@cs.psu.edu (Michael D Mellinger) writes:
>Sun does sell a Sparc Station for
>$5000, and I would say that you get considerably more performance out
>of it than a $5000 Mac.  Are we talking Apples and Apples yet?

Closer, but not quite. A totally mindless drone can open a 
boxed Mac system and have it up and running and on a network
in about fifteen minutes. I'm a pretty big fan of Sun's SPARCStations,
but I think you still kind of need to have a System Adminstrator
around when you get a new one. I have personally seen more than
one situation in which a complete novice was given a Mac with
no instructions at all, and they still managed to get useful work
done more or less immediately. Now, I think Open Look is a pretty
good user interface design, but I haven't seen a lot of applications
that use it well.
  I do not mean to put Suns down; I like them. I also like NeXT
very well. I also happen to have agreed for a long time with
critics who said that Macs were priced too high. However, Apples
and Suns are still not quite the same kinds of machines, and so
I don't expect them to sell quite the same way.

rca@cs.brown.edu (Ronald C.F. Antony) (10/25/90)

In article <1990Oct20.025628.11602@midway.uchicago.edu>, gft_robert@gsbacd.uchicago.edu writes:
|> Of course you can always run A/UX and get all of the above goodies, even w/o
|> waiting for 7.0.  But of course you wouldn't, 'cause Apple is Evil and Steve is
|> God, right? :->

The only thing in this respect that has to do with either hell or heaven are
Apples prices compared to NeXT's. Apple is a hell of a lot more expensive.

-- Ronald
------------------------------------------------------------------------------
"The reasonable man adapts himself to the world; the unreasonable one persists
in trying to adapt the world to himself. Therefore all progress depends on the
unreasonable man."  Bernhard Shaw | rca@cs.brown.edu or antony@browncog.bitnet
------------------------------------------------------------------------------