Traffic on c.u.sco.*
Peter
it looks like the traffic on the sco-ngs has dramatic decreased?!
It' only my feeling?
Cheers Peter
Tony Lawrence
>
> Hi,
> it looks like the traffic on the sco-ngs has dramatic decreased?!
> It' only my feeling?
That's my impression, too. I'd estimate comp.os.linux.misc
is running easily 10 times the trafiic of this; maybe more-
of course the newbie ratio is incredible, and the business
(as opposed to home use) is extremely low, so I dunno what
conclusions you can draw from that.
I think we have to reluctantly accept that OSR5, while not
dead, is now damaged goods. Those of us who earn our living
supporting it no doubt will have plenty of customers for at
least a few years, but unless Caldera somehow breathes new
life into it and brings it up to date (> 2GB fs, etc.) I
think that more and more of that business will be what I
call "Dr. Kevorkian gigs" - taking down the OSR5 and
replacing it with Linux and rarely) with Unixware. And, of
course, some part of it will go NT and Win2k- probably more
than we'd like, especially for those of us who don't want to
work in that environment unless and until it's the only way
we can survive. Actually, if that day comes, I'm going to go
do something else- there's no point in doing work you hate.
Speaking of all that, I had a curious thought the other
day: SCO and all the would be Intel Unix vendors had their
sales model upside down. What they should have done was
make the base single user OS, including devsys, close to
free, and charged for user licenses and business addons like
better fs's, extended memory support, smp and so on. Had
they done that, Linux never would have existed, Unix could
have competed on the desktop, and the infiltration into
server rooms would have been much higher.
Too late now, of course- or maybe it isn't.
--
Tony Lawrence (to...@aplawrence.com)
SCO/Linux articles, help, book reviews, tests,
job listings and more : http://www.pcunix.com
ligure.sis
"Tony Lawrence" <to...@aplawrence.com> ha scritto nel messaggio
news:39C769F0...@aplawrence.com...
Hi am Nico.
I perfectly agree wth you.
I was thinking this problem was big just in Italy or a lot bigger than
everywhere else.
just in my wee world Microsoft is getting the whole market and Linux doesn't
really esist yet.
In the new year , for the millennium bug, we have sold about 50 new server
for tp replace old server and old O.S.
more than half were unix server and become Microsoft.
The customars want that, so the software house had started to don't
implement the unix version of their programs anymore.
about 10 years ago i supported about 100 unix/xenix servers. Now are about
25.
we have not big customer but the government.
our normal installation is a server and 6 client.
maybe the big companies fell the problem a lot less, but i think that is the
touch in Italy.
I just hope some big software house start and write somethingh for Linux.
Nico
Bill Vermillion
>> Hi,
>> it looks like the traffic on the sco-ngs has dramatic decreased?!
>> It' only my feeling?
>That's my impression, too. I'd estimate comp.os.linux.misc
>is running easily 10 times the trafiic of this; maybe more-
>of course the newbie ratio is incredible, and the business
>(as opposed to home use) is extremely low, so I dunno what
>conclusions you can draw from that.
Well here that storage based on a 7 day expire shows that
the Linux groups account for 7.73 times more space.
Newbie ration on comp.unix.sco.misc went up after the
free OSR5 became available.
>I think we have to reluctantly accept that OSR5, while not
>dead, is now damaged goods. Those of us who earn our living
>supporting it no doubt will have plenty of customers for at
>least a few years, but unless Caldera somehow breathes new
>life into it and brings it up to date (> 2GB fs, etc.)
Well OSR5 is Sys V.3 technology based - and were you here
when the Dellevangelist Larry Snyder was always here decrying the
naming of SCO's Unix - and he said they were trying to deceive
people into thinking it was Sys V.4. He became invisible
after Dell stopped selling their Unix V.4 platform. There were
more choice of iNTEL based Unix then. Esix sufferered in the
V.3 to V.4 transition becoming a James River product.
I lost the lineage of the the iNTEL Unix - but was it no
iNTEL > Kodak > Interactive > Sun. Microport became an OEM and
porting house as did HHCS. [ I think the name was HHCS - based in
Dallas - with the first iNTEL port of V.4 - and showed first
on the Olivetti machine which as I recall - the main processor was
not iNTEL but all the math processing was done by the iNTEL i860].
That was when everyone just KNEW that Unix was the future. Windows
1.x and 2.x were pigs and only Windows 3 showed any promise.
You have to move fast in this business.
>I think that more and more of that business will be what I
>call "Dr. Kevorkian gigs" - taking down the OSR5 and
>replacing it with Linux and rarely) with Unixware. And, of
>course, some part of it will go NT and Win2k- probably more
>than we'd like, especially for those of us who don't want to
>work in that environment unless and until it's the only way
>we can survive. Actually, if that day comes, I'm going to go
>do something else- there's no point in doing work you hate.
I'm moving more to networking - but I agree about the NT and Win2K
environment. I find I miss so many tools that I've grown
accustomed to that I don't use those either. I have found in
making the MS based systems work in networks that I know more about
the MS product than some of those using it - more than I'd probably
like to know - but if you could sum up Sun Tzu in just three
words it would be "Know your enemy".
>Speaking of all that, I had a curious thought the other
>day: SCO and all the would be Intel Unix vendors had their
>sales model upside down. What they should have done was
>make the base single user OS, including devsys, close to
>free, and charged for user licenses and business addons like
>better fs's, extended memory support, smp and so on. Had
>they done that, Linux never would have existed, Unix could
>have competed on the desktop, and the infiltration into
>server rooms would have been much higher.
Part of Linux's existance would have to be blamed on the licensing
imposed by AT&T. It's hard to make something free when you have
to charger per user, charge for development system, charge again
for text processing tools. The latter was the first to go.
The question arises however about Unix on the desktop. It's going
to be hard to make something that is optimized for both sides.
One of the few that did a very good job was SGI. Narrow target
however. Nice workstations. Impressive servers on the big end.
But awfully expensive.
As to server infiltration - Unix still rules in internet side of
things. In the small/medium business environment I think NT
is used because of perception of NT in that marketplace.
It has a lot to do with popular press and also to do with small
offices with 1 or 2 machines getting a few more and then more until
it grew beyond all original expectations. However watching
a large network [where I was the sole Unix person making some apps
run for 3 or 4 small departments - maybe 40 users total] - seeing
this network grow and grow until there were rack and racks of
Netware - later MS - servers, I notice that the support personel
to user ratio was much much higher there than the Unix side.
But by the time you get to that level and see the problem you often
can't turn back. Reminds me of the oft repeated story about
putting a frog in hot water and he'll jump out, but if you put him
in cold water and slowly heat it he'll die and be cooked before he
realizes whats happening.
Didn't mean to get this carried on - but I agree on not wanting to
work in the MS environment. I notice that MS's ME is targetted
more toward the home while and I see something missing where it
won't be a nice in a business environment as the 95/98 stations.
It also may be that I haven't found where everything is. But it
looks like a step to force some changes. I also saw that one
reason MS isn't pushing upgrades is that of the almost 90M copies
of Win98 shipped almost all were on brand new machines and not
many were 95 upgrades. A least they know where the dolalrs are.
--
Bill Vermillion - bv @ wjv . com
Peter
<to...@aplawrence.com> wrote:
>Peter wrote:
>>
>> Hi,
>> it looks like the traffic on the sco-ngs has dramatic decreased?!
>> It' only my feeling?
>
>
>That's my impression, too. I'd estimate comp.os.linux.misc
>is running easily 10 times the trafiic of this; maybe more-
>of course the newbie ratio is incredible, and the business
>(as opposed to home use) is extremely low, so I dunno what
>conclusions you can draw from that.
Oh, then unixware must be a lot better than OS5, the traffic there is
about 1/10 of this one here.
You know what i mean and all the others too, but noone will really
believe it. I too!
The time is over and especially SCO with all it's parts of licenses
and such mucks is one of the important firm bringing unix down.
I'm now working 20 years with xenix/unix, was comming via TANIX
(Triumph-Adler Germany, later Ollvetti), SINIX (Siemens), Generix (a
Europe Version of SVR3.2) to SVR4 (of USL), Unixware (of Novell) to
Unixware (of SCO).
What will be the way, what could be a way?
Unixware from Novell had no lisence problematic, 2 user and unlimited,
all other needed things included to a very good price; SCO divided it
in its bits, just for licensing, not for service!
Next week here is the next QBB. 2 years ago we had 4 QBBs per year,
last year 3 and this year the first QBB was called the 2.QBB and this
now (the 2.) is called the 4.QBB. ;-) My goodness, what a company,
the new counting or area of SCO/Caldera.
The SCO country managers will tell us a lot, before two years they
said: only 3 unix will be in the future - ONE of this is SCO. Today he
will say ?
Here in Germany Caldera is a nowbody, in Germany (and i thing in
Europe) SUSE is the Linux Mainplayer.
Hard job today! Cheers Peter
Ben Rosenthal
> Peter wrote:
> >
> > Hi,
> > it looks like the traffic on the sco-ngs has dramatic decreased?!
> > It' only my feeling?
>
> That's my impression, too. I'd estimate comp.os.linux.misc
> is running easily 10 times the trafiic of this; maybe more-
> of course the newbie ratio is incredible, and the business
> (as opposed to home use) is extremely low, so I dunno what
> conclusions you can draw from that.
>
> I think we have to reluctantly accept that OSR5, while not
> dead, is now damaged goods. Those of us who earn our living
> supporting it no doubt will have plenty of customers for at
> least a few years, but unless Caldera somehow breathes new
> life into it and brings it up to date (> 2GB fs, etc.) I
> think that more and more of that business will be what I
> call "Dr. Kevorkian gigs" - taking down the OSR5 and
> replacing it with Linux and rarely) with Unixware. And, of
> course, some part of it will go NT and Win2k- probably more
> than we'd like, especially for those of us who don't want to
> work in that environment unless and until it's the only way
> we can survive. Actually, if that day comes, I'm going to go
> do something else- there's no point in doing work you hate.
>
> Speaking of all that, I had a curious thought the other
> day: SCO and all the would be Intel Unix vendors had their
> sales model upside down. What they should have done was
> make the base single user OS, including devsys, close to
> free, and charged for user licenses and business addons like
> better fs's, extended memory support, smp and so on. Had
> they done that, Linux never would have existed, Unix could
> have competed on the desktop, and the infiltration into
> server rooms would have been much higher.
Of course this is a better model. It is always better to have exposure,
because that leads to sales. If people don't know the product then they
are unlikely to buy it (or invest time in it). SCO is the best kept
secret in the world. The runner up is Tarantella.
This industry has always been driven by "make yourself indispensable."
Many view M$ as indispensable. M$ Word is the defacto standard for
email attachments. Owning a copy of M$ office has become a requirement,
just for compatibility.
>
>
> Too late now, of course- or maybe it isn't.
There will always be a place for a "commercial" version of UNIX on
intel. Many corporations still want "someone" to be responsible for the
product.
They can still save Tarantella by making it a household word. Ask 100
IT managers about Ctrix and 90% will nod their head, ask about
Tarantella and maybe 1 (not percent) know what it is.
>
>
> --
> Tony Lawrence (to...@aplawrence.com)
> SCO/Linux articles, help, book reviews, tests,
> job listings and more : http://www.pcunix.com
Ben Rosenthal
Bill Vermillion
Ben Rosenthal <bcr...@attglobal.net> wrote:
>Tony Lawrence wrote:
>> Speaking of all that, I had a curious thought the other
>> day: SCO and all the would be Intel Unix vendors had their
>> sales model upside down. What they should have done was
>> make the base single user OS, including devsys, close to
>> free, and charged for user licenses and business addons like
>> better fs's, extended memory support, smp and so on. Had
>> they done that, Linux never would have existed, Unix could
>> have competed on the desktop, and the infiltration into
>> server rooms would have been much higher.
>This industry has always been driven by "make yourself
>indispensable." Many view M$ as indispensable. M$ Word is the
>defacto standard for email attachments. Owning a copy of M$ office
>has become a requirement, just for compatibility.
Only rarely do I have to fire of M$ Word - on an attached machine -
to read files. I tend to live in a different world - still a lot
of command line only interfaces in such things as routers, etc.
but my MS emails get filtered through 'catdoc' and works most of
the time. The only time I always have to use M$ word is when a
place I work for sends me 'text' files. Those are always in
rtf format generated from M$ Word in a Mac OS/9 and OS/X
environment. Those are enough different that I have to use the
M$ Word.
>> Too late now, of course- or maybe it isn't.
>There will always be a place for a "commercial" version of UNIX on
>intel. Many corporations still want "someone" to be responsible for
>the product.
Guess who else is now has iNTEL based hardware - but at the moment
not Unix but Linux. Sun just acquired one of the larger of the
'internet rack mount' companies, Cobalt. The original Cobalt Raq
systems were running on MIPS chipsets - same as the SGI systems -
but they moved to iNTEL chips about a year ago. It might be
interesting to watch.
>They can still save Tarantella by making it a household word. Ask 100
>IT managers about Ctrix and 90% will nod their head, ask about
>Tarantella and maybe 1 (not percent) know what it is.
1 out of 100 is still 1% :-) GOTCHA! 1 out of a larger number
than 100 is highly probable IMO.
Geoff Johnson
> Peter wrote:
> >
> > Hi,
> > it looks like the traffic on the sco-ngs has dramatic decreased?!
> > It' only my feeling?
>
> That's my impression, too. I'd estimate comp.os.linux.misc
> is running easily 10 times the trafiic of this; maybe more-
> of course the newbie ratio is incredible, and the business
> (as opposed to home use) is extremely low, so I dunno what
> conclusions you can draw from that.
>
> I think we have to reluctantly accept that OSR5, while not
> dead, is now damaged goods. Those of us who earn our living
> supporting it no doubt will have plenty of customers for at
> least a few years, but unless Caldera somehow breathes new
> life into it and brings it up to date (> 2GB fs, etc.) I
> think that more and more of that business will be what I
> call "Dr. Kevorkian gigs" - taking down the OSR5 and
> replacing it with Linux and rarely) with Unixware. And, of
> course, some part of it will go NT and Win2k- probably more
> than we'd like, especially for those of us who don't want to
> work in that environment unless and until it's the only way
> we can survive. Actually, if that day comes, I'm going to go
> do something else- there's no point in doing work you hate.
>
> Speaking of all that, I had a curious thought the other
> day: SCO and all the would be Intel Unix vendors had their
> sales model upside down. What they should have done was
> make the base single user OS, including devsys, close to
> free, and charged for user licenses and business addons like
> better fs's, extended memory support, smp and so on. Had
> they done that, Linux never would have existed, Unix could
> have competed on the desktop, and the infiltration into
> server rooms would have been much higher.
>
Does nobody remember Novell
2 users very cheap.
Unlimited users very affordable.
Downside:
Paid far too much for unixware
Not enough balls or money to fight it out.
Open Look / Motif war (no wonderWindoze won)
X terminals far too expensive
PC X terminal software too slow.
Bulk async cabling is a night mare.
US Government says Operatings Systems to be POSIX compliant.
and accepts NT!
And now for the final act
SUSE/Caldera/Red Hat ... starting to fight a "FEATURE" war.
I can hear the gurgles already.
We are going down the same Univel/OSF/SUN OS/BSD path again.
M$ plays the lock in game to perfection.
Unless Linux can not only keep up but lead the features race as a
UNIFIED product. It will always be a marginal product.
M$ is still out there bastardizing standards, forcing the other players
to be bug compatible. (See Ted Lemon and DHCP).
This results in endless wasted development time for the open
source comunity.
If the OS movement can keep in front, then this M$ game will loose
its impact and force M$ to play the compatibilty game from a very poor
development base.
If you want to know why OSR5 is dying its because you were too stodgy
in your thinking to see a good product when you saw it and jump
en-mass to UW7.
The problem is not with OSR5's demise but having no viable platform to
progress to. You don't cry for XENIX bcause the natural progresion was
better.
OSR5 is just a tarted up SVR3 with an crude veneer of SVR4
UW7 could have been infinitely better but for the lack of dedication to
its
development.
Think of the money IBM pissed up the wall on OS2.
5% of that would have saved UW.
A bulk movemnet of OSR5 users to UW would have saved both operating
systems.
SCO would have had the money to develope UW and it would have kept up
support
for OSR5, just has it has its other legacy systems.
Just how many OSR5 and UW proffessionals are there out there.?
Did we have a $1000 dollars to spare?
> Too late now, of course- or maybe it isn't.
>
It is if we let linux fragment the way Sys V unix did.
Tony Lawrence
> I can hear the gurgles already.
Unfortunately, yes.
> If you want to know why OSR5 is dying its because you were too stodgy
> in your thinking to see a good product when you saw it and jump
> en-mass to UW7.
It's too big and too damn expensive.
>
> The problem is not with OSR5's demise but having no viable platform to
> progress to. You don't cry for XENIX bcause the natural progresion was
> better.
> OSR5 is just a tarted up SVR3 with an crude veneer of SVR4
> UW7 could have been infinitely better but for the lack of dedication to
> its
> development.
OK - could have been. Could have been if they had just
scrapped OSR5 and devoted their efforts to making the switch
seamless and unnoticeable to the average user. But it
wasn't, and though it's better, it still isn't. And it
still costs too damn much because there's too much fluff.
Bill Vermillion
Geoff Johnson <gp...@optushome.com.au> wrote:
>Tony Lawrence wrote:
>> I think we have to reluctantly accept that OSR5, while not
>> dead, is now damaged goods. ..
>Does nobody remember Novell
> 2 users very cheap.
> Unlimited users very affordable.
>Downside:
> Paid far too much for unixware
> Not enough balls or money to fight it out.
Part of the problem was the CEO wanted to become everything to
everyone. They bought WordPerfect, bought Unix, and did they not
also purchase Borland.
The old guard at Novell fought the integration mightily while the
new wave felt that putting a Unix based file system into Novell
would make it a much better OS. There were a lot of problems
internally with Novell over this.
When Novell sold the Unix code to SCO - the staunch Unix supports
who didn't move with the product are the ones who started Caldera.
You have noticed that Novell and Caldera are in the same city.
Some interesting things have come from people who left Novell.
>And now for the final act
> SUSE/Caldera/Red Hat ... starting to fight a "FEATURE" war.
You left out the fourth MAJOR Linux player - TurboLinux. It
practically owns Asia, while SuSE is preffered in Europe.
>We are going down the same Univel/OSF/SUN OS/BSD path again.
Sun just bought Cobalt which puts them squarely in the iNTEL
Linux based server game.
>Unless Linux can not only keep up but lead the features race as a
>UNIFIED product. It will always be a marginal product.
And each month sees at least one or two new Linux variants released
into the marketplace.
>If you want to know why OSR5 is dying its because you were too stodgy
>in your thinking to see a good product when you saw it and jump
>en-mass to UW7.
A lot has to do with the users of the product. I picked up one
customer earlier this year - Y2K changeover time - who is firmly
wedded to the SCO's 3.2.4.x on their NCR. "I don't want to change
anything because it works so well" is almost the exact quote.
Many moved from Xenix this past year because of Y2K. One didn't
move because they found they could update the Xenix to be Y2K
compatible. So they paid their vendor $500 for the patches.
Guess what the patch was. The Xenix Y2K patches freely
downloadable from the SCO site. Vendors still exists who take
advantage of customers trying to lock the customers to the vendor.
I've helped convert into MS a few sites that were locked by the
vendor. One had a standard brand serial board - but I could find
NO getty's enabled. SW vendor wrote their own routines and wanted
outrageous sums to upgrade the SW/OS. Scratch more more SCO site.
Another had what looked like Wyse terminals - but they had their
own proprietary ROMS in them and their SW was locked to that.
Paying almost as much for a new terminal as a PC cost - a few
hundred over list for Wyse - made them jump ship. One more SCO
site gone.
>A bulk movemnet of OSR5 users to UW would have saved both operating
>systems. SCO would have had the money to develope UW and it would
>have kept up support for OSR5, just has it has its other legacy
>systems.
So how do you convince the users.
>Just how many OSR5 and UW proffessionals are there out there.?
>Did we have a $1000 dollars to spare?
I'm doing more networking all the time.
>> Too late now, of course- or maybe it isn't.
>It is if we let linux fragment the way Sys V unix did.
One problem with some of the Linux advocates is they have learned
what to do and not why they do it. They learned the details and not
the system.
A short example. Client has a co-located server at our site. It
has a problem. I power cycle - nothing. I connect up a terminal
to the serial port. The system won't even come up into single user
mode - no shell. I pulled the unit and drove it the 60 miles to
the client. They could not re-install the OS on the drive. Drive
swap time.
Install Solaris on a new drive. The old drive did have the system
installed but was not bootable so it wouldn't load the new OS on
top of itself - that was the impression I had. Client installed
the OS, and I took it back and put it in the rack.
Support person leave and now that's another machine I have to take
care of. I get the passwords and login and I found why the system
would not boot before. The former sysadmin had learned on Linux -
had become a strong Bash advocate - the holiest of shells to hear
him speak - and that was the heart of the problem.
The bash was dynamically linked, when the system came up corrupt it
needed to go to single user mode. The libraries were not available
hence no shell.
There are some great Linux people out there - but since it is so
readily affordable there are a great many who have taught
themselves and haven't learned the consequences of seemingly
inoccuous actions - eg moving to a different shell for the root
account - they fall into holes like this. Inoperative systems - no
matter what the OS - never make clients happy.
There seems to be somewhat of a polarization between the young
Linux camps and the older traditonal Unix way. We need more
communications going and create a synergy so that we all grow and
are able to survive.
Tony Lawrence
> Support person leave and now that's another machine I have to take
> care of. I get the passwords and login and I found why the system
> would not boot before. The former sysadmin had learned on Linux -
> had become a strong Bash advocate - the holiest of shells to hear
> him speak - and that was the heart of the problem.
Ayup. We'll see more and more of that, I'll bet. I used to
get that a lot with Sun morons who would change a SCO system
to csh. Dopes.
I have a customer switching their main app software. New
app vendor insists that they must have Unitrends Ctar.
Nothing wrong with that, except the customer already has
Microlite Edge- but they insisted they buy Unitrends
nonetheless. Perfectly OK- but a waste of money.
> There are some great Linux people out there - but since it is so
> readily affordable there are a great many who have taught
> themselves and haven't learned the consequences of seemingly
> inoccuous actions - eg moving to a different shell for the root
> account - they fall into holes like this. Inoperative systems - no
> matter what the OS - never make clients happy.
A recent article in one of the trade mags had some dope
saying that this issue is why Linux is susceptible to
security problems- because inexperienced people install it.
Yeah, like Windows installs are all done by security
conscious professionals- sheesh!
>
> There seems to be somewhat of a polarization between the young
> Linux camps and the older traditonal Unix way. We need more
> communications going and create a synergy so that we all grow and
> are able to survive.
I dunno. It could be just 'cause I'm having a bad day in
general (couple of nasty installs that are not going well),
but if I could find something that paid even 3/4's as well,
I'd get out of the business now. Linux has killed SCO and
it will probably kill itself- dammit. I hope I'm just
stupidly depressed and completely wrong.
Bill Vermillion
Tony Lawrence <to...@aplawrence.com> wrote:
>Bill Vermillion wrote:
>> There are some great Linux people out there - but since it is so
>> readily affordable there are a great many who have taught
>> themselves and haven't learned the consequences of seemingly
>> inoccuous actions - eg moving to a different shell for the root
>> account - they fall into holes like this. Inoperative systems - no
>> matter what the OS - never make clients happy.
>A recent article in one of the trade mags had some dope
>saying that this issue is why Linux is susceptible to
>security problems- because inexperienced people install it.
>Yeah, like Windows installs are all done by security
>conscious professionals- sheesh!
But Windows won't let you do things the Unix systems will permit
you to do, and with Windows and it's goofey - woops gooey -
interface you can't screw up quite so badly. What you see is what
you get. It may not be what you wanted however.
>> There seems to be somewhat of a polarization between the young
>> Linux camps and the older traditonal Unix way. We need more
>> communications going and create a synergy so that we all grow and
>> are able to survive.
>I dunno. It could be just 'cause I'm having a bad day in
>general (couple of nasty installs that are not going well),
>but if I could find something that paid even 3/4's as well,
>I'd get out of the business now. Linux has killed SCO and
>it will probably kill itself- dammit. I hope I'm just
>stupidly depressed and completely wrong.
I don't know if I'd say that Linux killed SCO. Ten or more years
ago there were lots of branded Unix choices for the iNTEL
marketplace. The only one to hang on in the SW side alone was SCO,
and Sun's iNTEL Solaris has never been what'd you'd call a great
seller.
Small Unix systems seem to be readily replaced by MS based systems.
People may grumble and complain about how it's not as fast, or as
good, but they have plenty of company. SCO has always had the
image of the Unix for small businesses - while the likes of
HP, SUN, IBM, et al, have been considered for bigger systems - each
having their niche - one is strong in scientific, another in
networks, etc.
As more vendors to small business in the Unix world start dropping
the Unix support and move their products to NT for example, the
sales of small Unix systems fades. One client moved to NT two
years ago because the salesman he talked to said they weren't
supporting SCO's Unix anymore - so he contacted the HW people, and
got new SW, and I was only called in after it was all said and done
[after supporting him for five year] to move the data.
True - the vendor was not supporting SCO Unix - but they were
supporting OSR5. But the person in charge - who thought he knew
more about computers than he really did - made the move. So
instead of a $5K or so upgrade it became a $40K replacement.
I'm sure none of the salespeople involved in this were in the least
bit unhappy. This started back in the days when people would
say "What's Linux?".
While it is easy to say that "Linux has killed SCO" I think the
dwindling sales of small Unix based business systems started long
ago and Linux is just the easiest one to blame.
The eternal fighting of the Unix vendors over the years more than
helped bring it's downfall. Unix International scared some thinking
AT&T wouild control it all so the alternate [I forget it's name]
was started. Pitting Unix vendors against one another instead of
working together helped the disarray.
No offered alternatives to MS's software, and the ones that did
were either forced out of business by unscrupulous business
practivces [Read the story about how MS put Go out of business -
it's similar to things brought up in the anti-trust suit last
year], or they were bought by MS.
That was one thing MS had going for it - all the SW was from one
company or written for a compatible OS. But so much Unix software
had to be written/adapted for each platform.
Even on the iNTEL based Unix system the stock serial and printer
ports always had different interface names so there was the
installation packages had to be written with a lot of if/defs
or separate instructions for each system. Market place confusion
has killed more than one technology while the consumers just
sat on the sidelines and waited for the winner. For most people
MS was the winner. Only in the last year - particularly after
the trial has there been much anti-MS sentiment.
You also said:
>but if I could find something that paid even 3/4's as well,
>I'd get out of the business now.
I've had days like that too.
Bill
crma...@127.0.0.1
>
> Ayup. We'll see more and more of that, I'll bet. I used to
> get that a lot with Sun morons who would change a SCO system
> to csh. Dopes.
>
his 5.0.5 root account to ksh? Maybe he's lucky he hasn't rebooted in 3
months! Seriously, do I need to change it back? I thought ksh was a
superset of the Bourne shell? Chastised I am (dope now too it seems).
CRM
--
Clayton R. Malaker, M.D. crmalaker@xta.(substitute-com-to-unmunge)
824 Van Buren Street Belvidere, IL 61008-2252
Jean-Pierre Radley
| Tony Lawrence wrote:
| >
| > Ayup. We'll see more and more of that, I'll bet. I used to
| > get that a lot with Sun morons who would change a SCO system
| > to csh. Dopes.
| >
| Holy Korn shell Batman! Did the "Boy"-wonder screw up when he changed
| his 5.0.5 root account to ksh? Maybe he's lucky he hasn't rebooted in 3
| months! Seriously, do I need to change it back? I thought ksh was a
| superset of the Bourne shell? Chastised I am (dope now too it seems).
It probably won't be a problem.
--
JP
Bill Vermillion
>Tony Lawrence wrote:
>> Ayup. We'll see more and more of that, I'll bet. I used to
>> get that a lot with Sun morons who would change a SCO system
>> to csh. Dopes.
>Holy Korn shell Batman! Did the "Boy"-wonder screw up when he
>changed his 5.0.5 root account to ksh? Maybe he's lucky he hasn't
>rebooted in 3 months! Seriously, do I need to change it back? I
>thought ksh was a superset of the Bourne shell? Chastised I am
>(dope now too it seems).
A BIG difference between csh and sh, and ksh is supposed to be 100%
sh compatible. I've set up the ksh for root and have had no
problems. For problems with csh see the legendary "CSH programming
considered harmful" post which has been regularly reposted to the
'net for the last 10 years or so. Author is Tom Chritiansen.
Make sure any shell you use for root is in a directory accessible in
single user mode and is compiled staticly. Normal methods are
for all /bin and /sbin to be static and that every tool you might
ever need in single user mode should also be static.
Geoff Johnson
> Tony Lawrence wrote:
> >
> > Ayup. We'll see more and more of that, I'll bet. I used to
> > get that a lot with Sun morons who would change a SCO system
> > to csh. Dopes.
> >
> Holy Korn shell Batman! Did the "Boy"-wonder screw up when he changed
> his 5.0.5 root account to ksh? Maybe he's lucky he hasn't rebooted in 3
> months! Seriously, do I need to change it back? I thought ksh was a
> superset of the Bourne shell? Chastised I am (dope now too it seems).
> CRM
> --
> Clayton R. Malaker, M.D. crmalaker@xta.(substitute-com-to-unmunge)
> 824 Van Buren Street Belvidere, IL 61008-2252
I have run ksh on UW 2.1.3 and uw7.1 for years with no bad effects.
The problem is that you cannot get a statically linked version.
So if /usr is a seperate file system populate /usr/lib with the needed
libraries before mounting /usr.
That way single user mode still works if the /usr filesyhstem is damaged.
Tony Lawrence
> Holy Korn shell Batman! Did the "Boy"-wonder screw up when he changed
> his 5.0.5 root account to ksh? Maybe he's lucky he hasn't rebooted in 3
> months! Seriously, do I need to change it back? I thought ksh was a
> superset of the Bourne shell? Chastised I am (dope now too it seems).
> CRM
First, csh is an entirely different animal and startup scripts
break horribly if root's shell is csh. Ksh runs these scripts,
with one caveat: ksh has quite a number of built in functions: if
a startup script happened to define its own function that clashed
with ksh's name space, guess who'd win (or lose, more
accurately)?
Then there's the static link issue- most everything nowadays is
compiled with shared libraries. If you do "ldd /bin/sh" and "ldd
/bin/ksh" you'll see the libraries that these shells depend on.
Without those libraries, the shells won't work. If you look in
/sbin (on OSR5.0.5 and up- get tls704 otherwise), you'll finded
statically linked versions of sh and some other things, but not
ksh. These do NOT get used automagically and aren't even in your
PATH by default, so I'm not sure this is really germane to this
discussion, but it's good to know they are there should you
accidentally lose the symlink to the libraries.
Anyway, running ksh as root's shell is probably OK, but I still
wouldn't do it. I set up my .kshrc as usual, and export ENV
pointing to it in .profile, but leave it as /bin/sh in passwd.
When I login, if I want ksh (and I usually do), I just type ksh-
one extra step, 4 keystrokes, no big deal and no concern
whatsoever about any startup conflicts (unlikely as they may be).
Bill Vermillion
>> Tony Lawrence wrote:
>> > Ayup. We'll see more and more of that, I'll bet. I used to
>> > get that a lot with Sun morons who would change a SCO system
>> > to csh. Dopes.
>> Holy Korn shell Batman! Did the "Boy"-wonder screw up when he
>> changed his 5.0.5 root account to ksh? Maybe he's lucky he hasn't
>> rebooted in 3 months! Seriously, do I need to change it back? I
>> thought ksh was a superset of the Bourne shell? Chastised I am
>> (dope now too it seems).
>I have run ksh on UW 2.1.3 and uw7.1 for years with no bad effects.
>The problem is that you cannot get a statically linked version.
It depends on how much you want it. David Korn has ksh93
download for Unixware at the att site.
http://www.att.research.com/tools/reuse
>So if /usr is a seperate file system populate /usr/lib with the
>needed libraries before mounting /usr. That way single user mode
>still works if the /usr filesystem is damaged.
Good idea.
Bill Vermillion
Tony Lawrence <to...@aplawrence.com> wrote:
>crmalaker@[127.0.0.1] wrote:
>> Holy Korn shell Batman! Did the "Boy"-wonder screw up when he
>> changed his 5.0.5 root account to ksh? Maybe he's lucky he hasn't
>> rebooted in 3 months! Seriously, do I need to change it back? I
>> thought ksh was a superset of the Bourne shell? Chastised I am
>> (dope now too it seems).
>Then there's the static link issue- most everything nowadays is
>compiled with shared libraries. If you do "ldd /bin/sh" and "ldd
>/bin/ksh" you'll see the libraries that these shells depend on.
>Without those libraries, the shells won't work. If you look in
>/sbin (on OSR5.0.5 and up- get tls704 otherwise), you'll finded
>statically linked versions of sh and some other things, but not
>ksh.
You'll also note that /sbin/sh is a COFF file while the version
in /bin are ELF.
What I've also noticed is that other OSes have everything in
/sbin as static. Sun makes all of /bin dynamic and /sbin static.
The BSD's - iNTEL based - make everything in both /bin and /sbin
static while all the dyamics live under the /usr/bin /usr/sbin.
Rhapsody appears NOT to have any dynamics in the system anywhere.
>These do NOT get used automagically and aren't even in your
>PATH by default, so I'm not sure this is really germane to this
>discussion, but it's good to know they are there should you
>accidentally lose the symlink to the libraries.
I'd just as soon see all the programs that are available in single
user mode be statically linked. That means that if something,
upgrade, accident, errant software, hit's the libraries you will
at least have all your needed recovery files avaiable. I guess
you'd call this the belt and suspendors approach or perhaps
fail-safe. I don't think any program should have any dependancies
when you are in a single-user - often disaster recovery - mode.
Your other points are noted.
Tony Lawrence
>
> Tony Lawrence wrote:
> >
> > Ayup. We'll see more and more of that, I'll bet. I used to
> > get that a lot with Sun morons who would change a SCO system
> > to csh. Dopes.
> >
> his 5.0.5 root account to ksh? Maybe he's lucky he hasn't rebooted in 3
> months! Seriously, do I need to change it back? I thought ksh was a
> superset of the Bourne shell? Chastised I am (dope now too it seems).
Actually the reason I call 'em dopes is because it usually went
like this:
Customer hires some Sun "guru" to work on their system. The Sun
guru immediately changes root's shell to csh 'cause he's too dumb
to know that csh is crap. He then does whatever the customer
needs (probably screws that up too, but maybe not) and leaves.
At some later time, customer reboots, all hell breaks loose
because of root's csh, they call the Great Guru who looks at in,
is dumbfounded, tells them that SCO is crap, that he has no idea
what's wrong but he knows it is junk and he can't fix it.
I've been involved (after the fact) in scenarios just like this
at least three times. It's one thing to break something from
ignorance, it's quite another not to realize you broke it when
the evidence is plain as day- csh gives very predictable output
when it tries to run sh scripts, and I call them "dopes" more for
not being able to diagnose it than for their initial foolishness.
Jeff Liebermann
>There seems to be somewhat of a polarization between the young
>Linux camps and the older traditonal Unix way.
Somewhat of a polarization? Try snobbery bordering on illogical hatred.
Search Slashdot for SCO and see what the Linux crowd thinks of SCO. At
best, SCO is the evil predatory corporate giant, out to extort obscene
fees, for what by should, by all rights, be free to the unwashed masses.
Religious arguements between various Linux incantations only slightly
less acrimonious.
>We need more
>communications going and create a synergy so that we all grow and
>are able to survive.
By the same logic, all the religions of the world, should have organized
into some kind of OPEC-like cartel or standards organization, and unified
the structure of religion. The last time someone tried that was
Constantine in 400AD. Balkanization and entrenchment was the general
trend after that.
Commerical Unix is like the religions of Mithraism, where the high
priests of the religion maintained a large collection of "secrets", which
were gradually dispensed to the GUM (great unwashed masses). It was
quite popular in its time, but lost favour when just about everyone knew
these "secrets". Instead of doing R&D, and coming up with a new supply
of secrets, the priests of Mithra simply sat on their collection until it
was obsolete and depreciated in value.
Unix was similar, where the AT&T source code tree, was the great secret.
The chosen few were allowed to look upon the source. The GUM only got
compiled interpretations. AT&T had the same problems as the priests of
Mithra, and also mis-managed their collection of secrets. SCO did no
better. Rather than risk collapse by slowly revealing their secrets,
they simply constricted the supply.
The GUM were suitably dismayed at the divergence of the Unix religion,
where each incantation and proprietary enhancement, was followed by its
followers making war on its ancenstors. So, the GUM created their own
secrets, and passed them out freely, and were dismayed when their secrets
were deemed popular, but of little commercial value. Same problem as the
priests of Mithra. You can't sell something that everyone already knows
and has. The question of the hour seems to be how can one make money
with Linux. The stock market prices of RHAT and CALD should give some
indication of what is thought of the prospects.
Yeah, the various Unix vendors should have gotten together and
standardized everything in sight. Proprietary enhancements should have
been distributed seperately from the one true Unix. Shrink wrap and
portable Unix apps should have been available 10 years ago. Source code
should have been slowly leaked to forstall the inevitable.
I don't know what will happen with commerical Unix. I do know what
happened to Mithraism. It's followers found a new paradigm, with
mysteries and secrets that nobody could possibly understand. This
insured a never ending need for interpretors and consultants, and
prevented the inevitable dbasement of these mysteries and secrets by
demanding that one had to die before achieving complete understanding.
Kinda sounds like the Windoze of its day.
I'm not predicting the demise of commercial Unix. However, it is
interesting to note that none of the ancient religions in the form of
Mithraism, could generate sufficient capital to finance a sufficient army
in its defense. Commercial Unix has the same problem today. How does
one compete with a monopoly on one side, and a free product on the other?
--
Jeff Liebermann je...@comix.santa-cruz.ca.us
150 Felker St #D Santa Cruz CA 95060
831-421-6491 pager 831-429-1240 fax
http://www.cruzio.com/~jeffl/sco/ SCO stuff
Tony Lawrence
> >Linux camps and the older traditonal Unix way.
>
> Search Slashdot for SCO and see what the Linux crowd thinks of SCO.
Not even the Linux community gives much credit to anything said
on Slashdot..
There are always zealots. Zealotry is the most accurate marker
known for ignorance and bull-headed stupidity.
Jeff Liebermann
wrote:
>Not even the Linux community gives much credit to anything said
>on Slashdot..
>
>There are always zealots. Zealotry is the most accurate marker
>known for ignorance and bull-headed stupidity.
Hey. That's my ancestors you're talking about. The Zealots subbornly
held out for 3 years at Masada against Titus and his corporate might.
When they saw they were losing, they all committed suicide. Hmmm....
Yeah, I guess the Slashdot crowd qualifies as Zealots.
J. L. Schilling
: >We need more
: >communications going and create a synergy so that we all grow and
: >are able to survive.
: By the same logic, all the religions of the world, should have organized
: into some kind of OPEC-like cartel or standards organization, and unified
: the structure of religion. The last time someone tried that was
: Constantine in 400AD. Balkanization and entrenchment was the general
: trend after that.
Hmm, do you mean Constantine I and the First Council of Nicaea in 325,
or Constantine IV, who tried to unify the Eastern and Western churches
in 680, or Constantine V and the synod of 754, or Constantine VI and
the Second Council of Nicaea in 787 (reversing the icon veneration
decision of 754!), or Constantine XI and the Eastern/Western unification
proclamation of 1452 ...
Indeed, the track record isn't promising. And VI ended up being deposed
and blinded by his mother (!). I don't know what the UNIX parallel to that
would be, but it can't be too good
--
Jonathan Schilling SCO, Inc. j...@sco.com
Jim Richardson
> Tony Lawrence (to...@aplawrence.com)
> SCO/Linux articles, help, book reviews, tests,
> job listings and more : http://www.pcunix.com
>
One final point on sh .v ksh. Despite warnings, many folks make /usr a
separate file system. On UW7, as an example, /bin is a link to /usr/bin
while /sbin is a directory. In /sbin is the much discussed static sh,
but ksh is in /bin (aka /usr/bin). If /usr is not available for some
reason, ksh will not be available. My advice has been to leave root as
sh. You can always change to any shell you want once all is working.
JMTCW-YMMV
--
Jim Richardson
I like NT because it constantly reminds me of my daughter.
"Honest Daddy, I wasn't doing anything and it just broke."
Sent via Deja.com http://www.deja.com/
Before you buy.
Evan Hunt
Don't blame me, I voted for j...@sco.com.
>Hmm, do you mean Constantine I and the First Council of Nicaea in 325,
>or Constantine IV, who tried to unify the Eastern and Western churches
>in 680, or Constantine V and the synod of 754, or Constantine VI and
>the Second Council of Nicaea in 787 (reversing the icon veneration
>decision of 754!), or Constantine XI and the Eastern/Western unification
>proclamation of 1452 ...
>
>Indeed, the track record isn't promising. And VI ended up being deposed
>and blinded by his mother (!). I don't know what the UNIX parallel to that
>would be, but it can't be too good
EMACS, I'm guessing.
--
Evan Hunt - evanh at sco dot com
"I have always wished my computer would be as easy to use as my telephone.
My wish has come true. I no longer know how to use my telephone."
- Bjarne Stroustrup
Bill Vermillion
Jeff Liebermann <je...@comix.santa-cruz.ca.us> wrote:
>On Thu, 21 Sep 2000 14:37:03 GMT, bi...@wjv.com (Bill Vermillion) wrote:
>>There seems to be somewhat of a polarization between the young
>>Linux camps and the older traditonal Unix way.
>Somewhat of a polarization? Try snobbery bordering on illogical
>hatred. Search Slashdot for SCO and see what the Linux crowd thinks
>of SCO. At best, SCO is the evil predatory corporate giant, out to
>extort obscene fees, for what by should, by all rights, be free to
>the unwashed masses. Religious arguements between various Linux
>incantations only slightly less acrimonious.
I was in my be kind to animals mode :-) ^32
I've been used to having SCO bashed from the time I moved from the
worlds largest selling Unix based platform [Radio Shack 16 - sold
more units by about 50% than the second place Unix vendor - DEC.
DEC has more warm bodies in front of their machines however].
That was about 1985. We also had a local Unix group and I think I
was the only SCO user among all the rest who had 3B's, Vaxen,
Suns, DataGeneral, HP, IBM, and on and on. Meeting were often at
some corporate vendors sites. All of my friends ran something
other than SCO - and they univerally bashed it. The Linux
bashing is nothing new to me.
The free/alternate Unix software paradigm has been around since the
early days when Stallman brought forth GNU - Gnu's Not Unix - or
Douglas Comer's XINU [ Unix spelled backwards].
I have seen the arguments among the Linuxen as to which was the
better distribution - and as you say almost with the same fervor.
>>We need more
>>communications going and create a synergy so that we all grow and
>>are able to survive.
>By the same logic, all the religions of the world, should have
>organized into some kind of OPEC-like cartel or standards
>organization, and unified the structure of religion. The last time
>someone tried that was Constantine in 400AD. Balkanization and
>entrenchment was the general trend after that.
That's not what I meant in the preceding paragraph. I think one of
the better defintions of synergy I heard was like this.
"Two people meet. They both have a dollar. The exchange dollars.
Each still has one dollar.
Two people meet and each has an idea - they exchange ides - each
now has two ideas".
Looking and learning at the other persons point of view, or another
software implementation sometimes will show you that your way was
not the best, or you can implement some of the ideas from the other
side into your work. That is nothing like setting standards, it's
an exchange of ideas. While it may not change another person's
beliefs it hopefully will give them insight into your thinking and
perhaps see it from your perspective.
Once you try to merge standards you wind up with compromise after
compromise so that sometimes the best gets lost and only the
mediocre items upon which everyone agrees are left.
[I can't find a comp.unix.social.commentary group so this will have
to do].
>Instead of doing R&D, and coming up with a new supply of secrets,
>the priests of Mithra simply sat on their collection until it was
>obsolete and depreciated in value.
Sounds like some defunct software companies, does it not.
>Yeah, the various Unix vendors should have gotten together and
>standardized everything in sight. Proprietary enhancements should
>have been distributed seperately from the one true Unix. Shrink
>wrap and portable Unix apps should have been available 10 years
>ago. Source code should have been slowly leaked to forstall the
>inevitable.
A lot of the change started when iNTEL released chips where the
segments were so large you could treat them as linear memory
addresses. Up until then Unix implementation on Motorola, National,
Zilog, Western Electric, and other vendors chips were far easier
to program than iNTEL. The '286 made the first Unix implementation
on iNTEL livable, and the '386 opened the door for cheaper
implementations.
I remember many from that era who would never go near an iNTEL chip,
and only the strongest of the almost 100 Unix HW/SW variants of the
late 1980s survived. Gawd there were a LOT of them. I suspect some
didn't see the writing on the wall until it was almost faded and by
then it was too late. Several Unix on iNTEL platforms were released.
Microport made one of the first big pushed with their 'free' Sys
V.2. It came pre-installed on the xxMB hard drive that you paid $995
for. Big sales gimmic but it attracted a lot of attention and base
price for the whole package including development system and text
processing package was under $600. [I found an invoice from a long
time ago where I paid something like $550 including tax for it all].
Nothing else was that cheap with all those features.
Somewhere I have an Apple A/UX developers t-shirt. Not that I ever
used the platform - but they were also pushing Unix at that time,
and I suspect - but can not remember for sure - that is was on
the Motorola platform, one of the more widely used chips for Unix
systems. Apple's current Unix implementation is on the PowerPC
platform. Looks interesting. There is a petition on the 'net to
get Apple to port it to the iNTEL platform. GUI is reminiscent of
NeXTStep but not quite. One of the real pluses of NeXTStep - and it
was hardware hungry - was that the dispaly was display Postscript
- hardware intensive and of course required licenses from Adobe so
that ran up the cost.
>I don't know what will happen with commerical Unix. I do know what
>happened to Mithraism. It's followers found a new paradigm, with
>mysteries and secrets that nobody could possibly understand.
Hm. Why does that sound like the Microsoft Foundations Library?
Bill
Bill Vermillion
Jim Richardson <james_ri...@coastal.dnr.state.ga.us> wrote:
>In article <39CB72B9...@aplawrence.com>,
> Tony Lawrence <to...@aplawrence.com> wrote:
>> crmalaker@[127.0.0.1] wrote:
>> > Tony Lawrence wrote:
>> > > Ayup. We'll see more and more of that, I'll bet. I used to
>> > > get that a lot with Sun morons who would change a SCO system
>> > > to csh. Dopes.
>> > Holy Korn shell Batman! Did the "Boy"-wonder screw up when he
>> > changed his 5.0.5 root account to ksh? Maybe he's lucky he
>> > hasn't rebooted in 3 months!
>> Actually the reason I call 'em dopes is because it usually went
>> like this:
>> Customer hires some Sun "guru" to work on their system. The Sun
>> guru immediately changes root's shell to csh 'cause he's too dumb
>> to know that csh is crap.
>One final point on sh .v ksh. Despite warnings, many folks make /usr a
>separate file system. On UW7, as an example, /bin is a link to /usr/bin
>while /sbin is a directory. In /sbin is the much discussed static sh,
>but ksh is in /bin (aka /usr/bin). If /usr is not available for some
>reason, ksh will not be available. My advice has been to leave root as
>sh. You can always change to any shell you want once all is working.
There is nothing to prevent you from copying the ksh to /sbin.
Providing of course you have a static version of it.
I'd also advise that if someone has any favorite forensic tools
that aren't distributed with the OS to put them in /sbin
Jeff Liebermann
>Jeff Liebermann (je...@comix.santa-cruz.ca.us) wrote:
>
>: >We need more
>: >communications going and create a synergy so that we all grow and
>: >are able to survive.
>
>: By the same logic, all the religions of the world, should have organized
>: into some kind of OPEC-like cartel or standards organization, and unified
>: the structure of religion. The last time someone tried that was
>: Constantine in 400AD. Balkanization and entrenchment was the general
>: trend after that.
>
>Hmm, do you mean Constantine I and the First Council of Nicaea in 325,
>or Constantine IV, who tried to unify the Eastern and Western churches
>in 680, or Constantine V and the synod of 754, or Constantine VI and
>the Second Council of Nicaea in 787 (reversing the icon veneration
>decision of 754!), or Constantine XI and the Eastern/Western unification
>proclamation of 1452 ...
I meant Constantine 1.0. The later versions only tried to optimize
Christianity, with little consideration for the other religions. See:
http://www.salve.edu/~romanemp/conniei.htm
for a detailed history. The part it neglects to mention was the
Constantine genuinely tried to unify ALL the various Roman religions, not
just the multiple Christian factions. Everyone had their place in his
conglomeration. For example, the bridge keepers in Rome were a major
commercial power, which couldn't be ignored. So, the manager of the new
strategic alliance became the chief bridge keeper. The name "Pontiff or
Pontifex" is derived from "pons" meaning "keeper of the bridges".
Constantine 1.0 endorsed and supported Christianity as the dominant
religion because Christianity was the fastest growing but also the
weakest. He could dominate Christianity much easier than any of the
better established personal operating systems. Christianity also showed
an amazing uptime in surviving 300 years of continuous persecution, and
involved no established management team to pay off with golden
parachutes.
Meanwhile, the popular Linux religion is being endorsed by many of the
larger hardware manufacturers. Strategic alliances and mutual
development conglomerations are epidemic. I suspect the small Linux
companies are wondering what hit them when they find a large company both
holding and pulling ALL the strings. If the LSB (Linux Standards Base)
or other group ever successfully unites all 111 Linux distributions, I
will impressed.
>Indeed, the track record isn't promising. And VI ended up being deposed
>and blinded by his mother (!). I don't know what the UNIX parallel to that
>would be, but it can't be too good
Nothing new here. Unix marketing has been running blind for years, with
no visible effects, and they couldn't even sell the products to their own
mothers.
Drivel: I've been involved in a few volunteer organizations and have had
the displeasure of trying to manage volunteers. Must be similar to
trying to manage unpaid Linux developers. It's amazing how well Linux
resembles reality.
David Mabo
Unixware, step-mother of SCO Unix?!
"J. L. Schilling" wrote:
>
> Jeff Liebermann (je...@comix.santa-cruz.ca.us) wrote:
>
> : >We need more
> : >communications going and create a synergy so that we all grow and
> : >are able to survive.
>
> : By the same logic, all the religions of the world, should have organized
> : into some kind of OPEC-like cartel or standards organization, and unified
> : the structure of religion. The last time someone tried that was
> : Constantine in 400AD. Balkanization and entrenchment was the general
> : trend after that.
>
> Hmm, do you mean Constantine I and the First Council of Nicaea in 325,
> or Constantine IV, who tried to unify the Eastern and Western churches
> in 680, or Constantine V and the synod of 754, or Constantine VI and
> the Second Council of Nicaea in 787 (reversing the icon veneration
> decision of 754!), or Constantine XI and the Eastern/Western unification
> proclamation of 1452 ...
>
> Indeed, the track record isn't promising. And VI ended up being deposed
> and blinded by his mother (!). I don't know what the UNIX parallel to that
> would be, but it can't be too good
>
> --
> Jonathan Schilling SCO, Inc. j...@sco.com
--
David H. Mabo, CPCM
Adaptix Corp. - Cincinnati, Ohio
Jeff Liebermann
>Once you try to merge standards you wind up with compromise after
>compromise so that sometimes the best gets lost and only the
>mediocre items upon which everyone agrees are left.
#begin rant(void)
I was once involved in a standards group.
There once was a standards committee,
That met in a part of the city.
So rather than think,
In went kitchen sink,
and left us a mess, such a pity.
I'll resist the temptation to foam at the mouth over the methods and
results of several standards efforts I've been involved or monitoring.
However, you're off a bit. Minimum level of agreement does not equate to
mediocrity. The 802.11b (high rate 2.4Ghz wireless) could barely agree
on anything. Some members were off trying to specify the form factor,
colour, and required labeling to correspond to their product. Others had
patents that they wanted to see included so they could sue their
competitors. However, all but one agreed that a common level of
interoperability was needed and a good thing. The rest of the debate was
over how common, and what it should be. Synergy is a nice term, but the
reality was Darwin's survival of the fittest.
It's interesting to note that the inclusion of technology into the Linux
code base, is largely controlled by a few individuals. Nobody gets
anything into the kernel without Linus's approval. Alan Cox has the last
say in networking. So much for democracy in Linux. So, why does it work
so well? Because the people involved communicate very well, have the
best of intentions, understand the technology fully, and a eminently fair
in their decisions.
We return you now to the standards committee model, where communications
is an ordeal, intentions are political, representatives are often
technically unqualified, and equitable arrangements are often settled in
the courts. I've seen a few minor Linux standards efforts collapse
because the players were unwilling to play the game by the rules of the
typical standards committee. I don't blame them.
So, why does Linux work? Well it doesn't in some areas. Linux seems to
be a collection of parts and pieces. If you built an automobile by
buying parts and pieces from the local auto parts store, you would have a
Linux automobile. Various parts and pieces would function perfectly, or
even in a superior manner. However, the entire assemblage would of
dubious reliability. Integration and testing are somewhat lacking.
(Note: This may change as the "Open Software Testing Lab" gets going
with support from big iron companies).
At the other end of the spectrum are the commercial Unixii. On the
whole, they integrate the parts together quite nicely. The overall
impression is a stable, reliable, functional, product. However, if you
dissect this animal, you'll find that the individual components are
rather ugly, rather old, and have lots of room for improvement. Some
parts are even unfathomable. This is the traditional Detroit automobile
model.
So, which is more popular? The open source model, where you assemble
your own automobile out of public domain components? Yes, there are kit
cars. Electric vehicles are often assembled this way. It has its place.
Or, is the proprietary architecture of the typical Detroit model, the
more popular route? Certainly it's a good thing to be able to get under
the hood and replace commercial products with open source replacement
components. The JC Whitney catalog is full of automotive bolt-on modules
and drivers. However, I'm not sure I wanna risk my life driving an open
source automobile.
>[I can't find a comp.unix.social.commentary group so this will have
>to do].
Unix geeks are not social animals and have no life outside Unix. A
social group would be empty. The closest I could find was:
soc.culture.scientists
soc.culture.misc
There's also a soc.genealogy.computing. Weird.
>I remember many from that era who would never go near an iNTEL chip,
>and only the strongest of the almost 100 Unix HW/SW variants of the
>late 1980s survived. Gawd there were a LOT of them. I suspect some
>didn't see the writing on the wall until it was almost faded and by
>then it was too late. Several Unix on iNTEL platforms were released.
There was one big difference between the Intel bunch and the others.
Intel went to heroic efforts to run previous legacy code on new
processors. The others would trash their customers software investment
with every new processor or machine.
>>I don't know what will happen with commerical Unix. I do know what
>>happened to Mithraism. It's followers found a new paradigm, with
>>mysteries and secrets that nobody could possibly understand.
http://dir.yahoo.com/Society_and_Culture/Religion_and_Spirituality/Faiths_and_Practices/Mithraism/
>Hm. Why does that sound like the Microsoft Foundations Library?
I don't engage in programming or other perversions, so I wouldn't know.
I couldn't even program my VCR to record some Olympic Games and get it
right. To me, the big mystery is why some people actually enjoy
programming. Apple recognized the problem long ago and called it
scripting instead of programming.
Bill Vermillion
Jeff Liebermann <je...@comix.santa-cruz.ca.us> wrote:
>On Fri, 22 Sep 2000 19:57:51 GMT, bi...@wjv.com (Bill Vermillion) wrote:
>>Once you try to merge standards you wind up with compromise after
>>compromise so that sometimes the best gets lost and only the
>^^^^^^^^^<
>>mediocre items upon which everyone agrees are left.
>#begin rant(void)
>I was once involved in a standards group.
>
> There once was a standards committee,
> That met in a part of the city.
> So rather than think,
> In went kitchen sink,
> and left us a mess, such a pity.
Cute.
>However, you're off a bit. Minimum level of agreement does not
>equate to mediocrity. ...
Note that I did say sometimes. I have a friend who is a fellow in
a standards group. While they may not be large - their decisions
affect a lot of equipment design for hardware - much of which is in
the $100K range. Those aren't medicore but carefully researched
engineering standards.
But in other arenas compromises often lead to dillution of original
concepts - something that pleases everyone doesn't neccesarily that
what is workable for everyone is best for everyone.
> Some members were off trying to specify the form factor, colour,
>and required labeling to correspond to their product. Others had
>patents that they wanted to see included so they could sue their
>competitors. However, all but one agreed that a common level of
>interoperability was needed and a good thing.
Protect you investment. That is normal.
>The rest of the debate was over how common, and what it should be.
>Synergy is a nice term, but the reality was Darwin's survival of
>the fittest.
He who has the most money and power sets the rules.
>So, which is more popular? The open source model, where you
>assemble your own automobile out of public domain components? Yes,
>there are kit cars.
And that almost the model for the British automobile industry at
one time. The same Lucas tail-light lens on a Jaguar was the same
as on an Triumph-TR3. All the instruments said either Smiths or
Jaeger. There are other places where parts are interchangeable.
Many years ago when I blew a clutch on my Alfa Romeo Gullietta
Sypder Veloce [ I loved that name ], it was more expedient to
replace the clutch plate with a $9 [as I remember] Studebaker part
than a $40 Alfa part and have to wait a week or 10 days for it to
arrive.
>Or, is the proprietary architecture of the typical Detroit model, the
>more popular route?
It give more individuality to the make - instead of all having the
same cosmetic tail-light as above - but it means that you have to
build more of any particular model to make styling/cosmetic changes
cost effective.
>Certainly it's a good thing to be able to get under the hood
>and replace commercial products with open source replacement
>components. The JC Whitney catalog is full of automotive bolt-on
>modules and drivers. However, I'm not sure I wanna risk my life
>driving an open source automobile.
But the JC Whitney model is not what I'd call open source because
many of the pieces/parts are made to replace OEM euqipment, while
the British model had the automobile manufacturers getting the
'periphals' from manufacturers who supplied parts for new vehicles.
Most of the Italian electricals were from Magnetti Marelli [ and oh
how I cussed their startes on cold Washington state mornings in a
car that seemed to be designed for a warm Mediteranean atomsphere.
It was an art to learn to tap-dance on the throttle firing up
something with dual 40MM dual Webers and no chokes to get it lit,
but once running what a jewel. It had been tweaked above specs a
bit and was getting about 110HP out of 78 cubic inches at 6grand.
Had to be the the favorite automobile of all I've owned with the
Sunbeam Tiger coming in close second.
That's on the opposite scale of your truck - which do I not recall
is also a diesel?
>>I remember many from that era who would never go near an iNTEL chip,
>>and only the strongest of the almost 100 Unix HW/SW variants of the
>>late 1980s survived. Gawd there were a LOT of them. I suspect some
>>didn't see the writing on the wall until it was almost faded and by
>>then it was too late. Several Unix on iNTEL platforms were released.
>There was one big difference between the Intel bunch and the others.
>Intel went to heroic efforts to run previous legacy code on new
>processors. The others would trash their customers software investment
>with every new processor or machine.
And when they tried to change that - they had problems. The P6
based CPU - the PentiumPro ran Windows 16 bit code slower than
32 bit code. Made it a wonderful chip for Unix and NT systems, but
didn't sell well because an identically clocked P5 ran faster
than the P6 - and thus the MMX came up on the market.
Sometimes you do have to break with the past. Trying to remain
backwards compatible could have more financical impact than
starting over.
There are always cross-compilers. The earliest MS programs
suffered from this converting the 8080 codes into the 8086
envirionments as I recall. But in those days MS was primarily
a language company.
>>Hm. Why does that sound like the Microsoft Foundations Library?
>I don't engage in programming or other perversions, so I wouldn't know.
All you have to do is see some of it and you'll realize that you
made the right decision.
>I couldn't even program my VCR to record some Olympic Games and get it
>right.
That means you just bought the wrong VCR. On my recently retired
unit if it had an extended power off and came up with 12:00:00
hitting date would bring up a prompt on the screen and show
1986. Spin the dial to current year [up to 2016], hit enter,
spin to month and the date and the day was set.
To program a record since it already knew what day it was it
displayed the rest of the months calendar with today highlight,
spin the knob to the day, enter, spin to station, spin to time.
It's all in the user interfaces. And now with the time on the new
VCRs being set by the clock from the local PBS stations [
technology by a grant from Sony as I recall] you should never have
to even see blinking time. As a matter of fact if I have an
extended power outage I set all the clocks to the VCR - it's the
only one I know is accurate except the computer.
>To me, the big mystery is why some people actually enjoy
>programming. Apple recognized the problem long ago and called it
>scripting instead of programming.
Well I guess it's how you define programming also. Isn't system
adminstration with shell scripts programming? I don't mind working
in databases either, and that's programming.
What I don't like are TheExtremleyLongVariableName concepts that
seem to be prevalent today, so I stay away from that area. :-)
Jeff Liebermann
>Note that I did say sometimes. I have a friend who is a fellow in
>a standards group. While they may not be large - their decisions
>affect a lot of equipment design for hardware - much of which is in
>the $100K range. Those aren't medicore but carefully researched
>engineering standards.
Carefully researched? What manner of science fiction be this? The stuff
I saw looked like it was extracted from various manufacturers white
papers and marketing propaganda. I made a promise (to myself) to
personally strangle the next moron who proclaims that their technology is
better because it supports more acronyms. Never mind that it barely
works, doesn't scale well, had predatory license fees, is sole sourced
for the major chips, and costs too much.
I'll admit that my involvement in standards production was many years
ago. Perhaps a miracle has occurred and the standards groups are
actually working with de facto standards (already in production with
demonstrateable hardware) than de jure standards (transient paper tigers
based on theory). I saw far too much of the latter. My current favorite
is the 1500(?) member Bluetooth conglomeration. The only hardware that's
available are $5,000 development systems. The science fiction $5 single
chip Bluetooth radio is many many years away. Intel is currently only
2.5 years late on their promised ship date. Much of the research is
based on modeling. The standard is already at 1.1 and will soon go to
Bluetooth 3 (don't ask why it's now three) thanks the members wanting to
do voice over Bluetooth. Few have even seen the chips, but the revisions
are already going strong. Meanwhile, the IEEE has assembled the 802.15
committee, who's sole purpose is to arbitrate between the multiple
mutually interfering 2.4Ghz technologies. To insure confusion, the FCC
just loosened up on the rules for both PN and FH spread spectrum for both
the HomeRF and 802.11b mobs, effectively doubling thruput but also
creating more interference potential for Bluetooth.
Don't even get me started on what I think of WAP (Wireless Access
Protocol).
>But in other arenas compromises often lead to dillution of original
>concepts - something that pleases everyone doesn't neccesarily that
>what is workable for everyone is best for everyone.
Ever try to please the 1500(?) members of the Bluetooth consortium? One
will take one hour, two will take two hours, three will take three hours,
ad nausium. Projects like this scale linearly.
>That's on the opposite scale of your truck - which do I not recall
>is also a diesel?
Yep. 1983 Dodge D50 diesel. 260,000 miles so far.
My previous vehicle was a 1970 Series IIa Land Rover.
http://www.cruzio.com/~jeffl/pics/land_rover1.jpg
Like all British automobiles, it was in a perpetual state of restoration.
Every Saturday was spent under the Rover doing something. Kinda reminds
me of spending my life configuring Windoze or "learning" Linux. I
decided my Saturdays were best spent on other things, so when I finally
blew the transmission with a very bad downshift, I bought a vehicle that
required little maintenance. I shopped around for a while looking under
the hood. Everything I saw was a spaggetti mess of wires and hoses. The
diesel was simple and only a little more expensive. What little
maintenance I do is rarely under the truck.
>That means you just bought the wrong VCR.
I paid $10 at a garage sale for it. My guess is circa 1995. I could get
an upgrade, but the legacy hardware serves 90% of my needs. When the
standards change so that my antique will not longer run the current
technology (i.e. Digital TV), I might consider an upgrade. Meanwhile,
I'm a Xenix kind of TV viewer.
>It's all in the user interfaces. And now with the time on the new
>VCRs being set by the clock from the local PBS stations [
>technology by a grant from Sony as I recall] you should never have
>to even see blinking time.
It has the EDS/SMTPE time set feature. I tried it and it sorta works.
I'm in a fringe area in the mountains. I get 2 strong TV stations, and 4
weak ones. PBS varies thanks to me planting my TV antenna in the exact
location where the KTEH (PBS) transmitter is behind a bunch of redwoods.
The result is that the closed captions and EDS time signals are flakey.
With EDS running, the digital clock resembles a slot machine display.
I also don't use the stock remote control. I have an IR remote control
emulation program for my HP 100LX palmtop. I have it programmed to
provide the most common functions for the TV, VCR, and various satellite
boxes (mostly SCPC) as macros. Works for me.
Oh well, back to stacking firewood...
Crossfire
the Lovely and Talented Geoff Johnson <gp...@optushome.com.au> wrote:
[...snip]
] If you want to know why OSR5 is dying its because you were too stodgy
] in your thinking to see a good product when you saw it and jump
] en-mass to UW7.
Coming from a Small - to - Medium - Business POV:
UW7 talked the talk but couldn't walk the walk until the release of 7.1.1.
We bought 7.0.0, 7.0.1, and 7.1.1. We tried 'em all, in-house. 7.0.0 took
me 3 days of on-and-off trying to even get to install, and then crashed
every time I did anything with any of the hardware. 7.0.1 installed easily
and was a bit more stable, but still had lots of HW-related problems.
7.1.1 installed fairly easily and has turned out to be quite stable--I
don't think my 7.1.1 box has crashed once since I installed it in February,
and I haven't exactly been babying it. My only objection as an administrator
to 7.1.1 is that there's a jillion supplements that need to be installed--at
least SCO released several of the more important ones as a single supplement.
From a selling standpoint, UW7 was entirely too expensive for what most
SMB people needed until the release of the Business Edition. Sure, it
had capacity for giant filesystems, lots of memory, and hotpluggable
hardware, but most small businesses don't need features like that and so
they aren't going to want to pay for them.
For us, stoginess had nothing to do with it; SCO shot themselves in the
foot by offering an unstable and very expensive OS and pretending like it
was a replacement for a stable and (relatively) inexpensive OS.
Personally, I like UW7, BTW. It's my main server for my domain and it's
great. I wouldn't WANT to go back to OSR5, though it does have a special
place in my heart... ;)
Just my $0.02. YMMV, IMHO, etc. Here, have a grain of salt.
Jon Reid
========================================================================
Email: jonLU...@MEATapeiros.com Web: http://www.apeiros.com/~jon
(DeSPAM my email to contact me.)
========================================================================
Peter
>
>UW7 talked the talk but couldn't walk the walk until the release of 7.1.1.
>We bought 7.0.0, 7.0.1, and 7.1.1. We tried 'em all, in-house. 7.0.0 took
>me 3 days of on-and-off trying to even get to install, and then crashed
>every time I did anything with any of the hardware. 7.0.1 installed easily
>and was a bit more stable, but still had lots of HW-related problems.
>7.1.1 installed fairly easily and has turned out to be quite stable--I
>don't think my 7.1.1 box has crashed once since I installed it in February,
>and I haven't exactly been babying it. My only objection as an administrator
>to 7.1.1 is that there's a jillion supplements that need to be installed--at
>least SCO released several of the more important ones as a single supplement.
>
>From a selling standpoint, UW7 was entirely too expensive for what most
>SMB people needed until the release of the Business Edition. Sure, it
>had capacity for giant filesystems, lots of memory, and hotpluggable
>hardware, but most small businesses don't need features like that and so
>they aren't going to want to pay for them.
>
>For us, stoginess had nothing to do with it; SCO shot themselves in the
>foot by offering an unstable and very expensive OS and pretending like it
>was a replacement for a stable and (relatively) inexpensive OS.
>
>Personally, I like UW7, BTW. It's my main server for my domain and it's
>great. I wouldn't WANT to go back to OSR5, though it does have a special
>place in my heart... ;)
>
So we using still UW2.1.1/.2 and now .3 without
any problems. Some users run it now over 3 years an all without
a crash. On all the production UW2* systems i had never a crash over
the last 5 years! :-))))
UW7 until now is a ping pong play.
But SCO is drying out UW21x and the same will come to OS and i believe
to UW7 under Caldera too.
I like UW2, using it from the first days and had never trouble with
itself only very bad driver support. But i think this is/was a SCO
speciality.
Cheers Peter
Bill Vermillion
Crossfire <jonLU...@MEATapeiros.com> wrote:
>On Thu, 21 Sep 2000 13:53:00 GMT,
>the Lovely and Talented Geoff Johnson <gp...@optushome.com.au> wrote:
>
>[...snip]
>] If you want to know why OSR5 is dying its because you were too stodgy
>] in your thinking to see a good product when you saw it and jump
>] en-mass to UW7.
>Coming from a Small - to - Medium - Business POV:
>UW7 talked the talk but couldn't walk the walk until the release
>of 7.1.1. We bought 7.0.0, 7.0.1, and 7.1.1. We tried 'em all,
>in-house. 7.0.0 took me 3 days of on-and-off trying to even get to
>install, and then crashed every time I did anything with any of the
>hardware. 7.0.1 installed easily and was a bit more stable, but
>still had lots of HW-related problems.
Took awhile to get that one right I guess.
...
>From a selling standpoint, UW7 was entirely too expensive for
>what most SMB people needed until the release of the Business
>Edition. Sure, it had capacity for giant filesystems, lots of
>memory, and hotpluggable hardware, but most small businesses don't
>need features like that and so they aren't going to want to pay for
>them.
All of those all the Sys V.4 featurers that were there in Unixware
from Novell, Unix from Sun, SGI, and the other Sys V.4 vendors.
SCO didn't add them. This was part of what you got with Sys V.4
and if you didn't need it - well that was part of the package. IMO
was quite late coming to the Sys V.4 game. It may not have been
needed by the small business users but the larger systems needed
V.4 offerings, and the small business market was rapidly moving
into peer-peer Windows and NT solutions. Well over 60% of the
small SCO users I have supported in the past are on MS based
platforms. Only the larger ones are still on SCO.
Whether the smaller users want/need them or not, SCO couldn't
afford to support a rapidly shrinking market segment could they?
>For us, stoginess had nothing to do with it; SCO shot themselves
>in the foot by offering an unstable and very expensive OS and
>pretending like it was a replacement for a stable and (relatively)
>inexpensive OS.
Even Volkswagen eventually dropped the 'beetle'.
>Personally, I like UW7, BTW. It's my main server for my domain and
>it's great. I wouldn't WANT to go back to OSR5, though it does have
>a special place in my heart... ;)
The special place I have is for the really old Xenix. So much power
in so little space. 6 user systems with only 1MB RAM and a 20MB
HD.
Crossfire
the Lovely and Talented Bill Vermillion <bi...@wjv.com> wrote:
] In article <slrn8t8cut.45...@olethros.apeiros.com>,
] Crossfire <jonLU...@MEATapeiros.com> wrote:
]
] >Coming from a Small - to - Medium - Business POV:
]
] >UW7 talked the talk but couldn't walk the walk until the release
] >of 7.1.1. We bought 7.0.0, 7.0.1, and 7.1.1. We tried 'em all,
] >in-house. 7.0.0 took me 3 days of on-and-off trying to even get to
] >install, and then crashed every time I did anything with any of the
] >hardware. 7.0.1 installed easily and was a bit more stable, but
] >still had lots of HW-related problems.
]
] Took awhile to get that one right I guess.
I'll admit I had problems installing 7.1.1, but there was a TA solution.
All I had to do was stop being frustrated and go look for it.
] >From a selling standpoint, UW7 was entirely too expensive for
] >what most SMB people needed until the release of the Business
] >Edition. Sure, it had capacity for giant filesystems, lots of
] >memory, and hotpluggable hardware, but most small businesses don't
] >need features like that and so they aren't going to want to pay for
] >them.
]
] All of those all the Sys V.4 featurers that were there in Unixware
] from Novell, Unix from Sun, SGI, and the other Sys V.4 vendors.
Ah yes, I remember that. (Doesn't much alter my point, but I do remember
that.)
] SCO didn't add them. This was part of what you got with Sys V.4
] and if you didn't need it - well that was part of the package. IMO
] was quite late coming to the Sys V.4 game. It may not have been
] needed by the small business users but the larger systems needed
] V.4 offerings, and the small business market was rapidly moving
] into peer-peer Windows and NT solutions. Well over 60% of the
] small SCO users I have supported in the past are on MS based
] platforms. Only the larger ones are still on SCO.
]
] Whether the smaller users want/need them or not, SCO couldn't
] afford to support a rapidly shrinking market segment could they?
In my (admittedly anecdotal) experience, the SMB segment was viable and
growing. We never had any lack of new clients for our solutions. Of course,
SMB folks can go the Microsoft route (*shudder*), but they didn't have to
because OpenServer was/is a good product for them.
I hope Caldera will continue to support OpenServer and UnixWare both,
as they target separate legitimate markets. I doubt that'll happen, though.
] >For us, stoginess had nothing to do with it; SCO shot themselves
] >in the foot by offering an unstable and very expensive OS and
] >pretending like it was a replacement for a stable and (relatively)
] >inexpensive OS.
]
] Even Volkswagen eventually dropped the 'beetle'.
And then decided to bring it back. (I doubt that'd happen to OpenServer
if Caldera decides to can it, of course.)
] >Personally, I like UW7, BTW. It's my main server for my domain and
] >it's great. I wouldn't WANT to go back to OSR5, though it does have
] >a special place in my heart... ;)
]
] The special place I have is for the really old Xenix. So much power
] in so little space. 6 user systems with only 1MB RAM and a 20MB
] HD.
Xenix was before I came into the field; I cut my teeth on HP-UX when I
was a student. HP-UX does NOT have a special place in my heart... ;)
] Bill Vermillion - bv @ wjv . com
A Shelton
> On Thu, 21 Sep 2000 14:37:03 GMT, bi...@wjv.com (Bill Vermillion) wrote:
>>There seems to be somewhat of a polarization between the young
>>Linux camps and the older traditonal Unix way.
> Somewhat of a polarization? Try snobbery bordering on illogical hatred.
> Search Slashdot for SCO and see what the Linux crowd thinks of SCO. At
> best, SCO is the evil predatory corporate giant, out to extort obscene
> fees, for what by should, by all rights, be free to the unwashed masses.
> Religious arguements between various Linux incantations only slightly
> less acrimonious.
Some of us wanted to buy a unix, for personal use, before Linux came
into existence. Talking to a hostile vendor (reasonable enough, he knows
i'm not a business) and working out it would cost me 4,500$ (australian)
for a minimum development system made me wonder how SCO was going to
compete with windows in the future. After that introduction finding
Linux, with an agressive momentum and ambition was simply golden.
(Note that I did find a copy of SCO in the interval in between)
Likewise SCO did itself no favors. Snippy marketing, demeaning comments
from users, lots of `toy unix' comments, and even a bit of FUD...what
did they expect?
> Commerical Unix is like the religions of Mithraism, where the high
> priests of the religion maintained a large collection of "secrets", which
> were gradually dispensed to the GUM (great unwashed masses). It was
> quite popular in its time, but lost favour when just about everyone knew
> these "secrets". Instead of doing R&D, and coming up with a new supply
> of secrets, the priests of Mithra simply sat on their collection until it
> was obsolete and depreciated in value.
People might (reasonably) think i'm another SCO bashing linux zealot.
But really I support OS variety, and at the moment that means surviving
against microsoft. A lot of linux people were fairly disappointed that
the `dominant' unix on intel seemed happy to milk a niche and make enemies
of the rapidly developing linux community while getting their base eroded
away. I remember them announcing their Unixware/OSR5 merge, projects Gemini
and Monterey, but I don't see anything that is going to start knocking NT
off servers. And Linus's observation that you can't protect the servers
when another company owns 100% of the desktops is still valid.
ps. Does anyone have the classic post to linux misc along the lines off,
"now that unixware is cheap even you students can use a real unix"? I
wish i'd kept it.
pps. There's 111 (probably) linux distributions. Only 5 (ish) of them
are important and the core of those is largely identical. Oh, and most
linux people don't exactly know how they intend to make money. But
watching them fight while trying to appear `open and friendly' is
awfully amusing.
--
Apparently I'm insane, but I'm one of the happy kinds. (dilbert)
Andrew Shelton ashe...@cs.rmit.edu.au
GCS(2.1)-d+H+sw+v-C++UL+>L+++E-N++WV--R++tv-b+D++e+fr*y?