What's VMS up to these days?

[A. W. Dunstan]

Many years ago I used VMS a great deal & thought it was the greatest thing
since sliced bread. But where I live I haven't seen anyone interested in
VMS in about 20 years, and most people have never even heard of it.

My 1st job out of college was for EDS, who used VMS for plant floor
automation for GM. When I left EDS I worked for a small company (KMS
Fusion) that did scientific research; they were a heavy user of VMS (and
RSX) and active in DECUS. We were at v5.5 or so, as I recall.

I've been following this newsgroup for a few months now & see a fair bit of
activity. Which leads me to wonder what VMS is used for these days? Who
uses it, and where? Is it still being developed? Does anyone do any
development of software to run on it?

Just curious...

--
Al Dunstan, Software Engineer
OptiMetrics, Inc.
3115 Professional Drive
Ann Arbor, MI 48104-5131

"There are two ways of constructing a software design. One way is to
make it so simple that there are obviously no deficiencies. And the
other way is to make it so complicated that there are no obvious
deficiencies."
- C. A. R. Hoare

[Jan-Erik Soderholm]

A. W. Dunstan wrote 2012-02-01 19:30:
> Many years ago I used VMS a great deal& thought it was the greatest thing

> since sliced bread. But where I live I haven't seen anyone interested in
> VMS in about 20 years, and most people have never even heard of it.
>
> My 1st job out of college was for EDS, who used VMS for plant floor
> automation for GM. When I left EDS I worked for a small company (KMS
> Fusion) that did scientific research; they were a heavy user of VMS (and
> RSX) and active in DECUS. We were at v5.5 or so, as I recall.
>

> I've been following this newsgroup for a few months now& see a fair bit of

> activity. Which leads me to wonder what VMS is used for these days?

More or less the same things as wen you were active. :-)

> Who uses it, and where?

Probably those who do, also used it 20 years ago.

> Is it still being developed?

Yes. Latest version (8.4) released summer 2010.
Of course, not the major leaps as when things like clusters was
released. Mostly adaptions to new hardware and storage.

> Does anyone do any development of software to run on it?

I do. COBOL and Rdb.

B.t.w, Sweden has a few large VMS users. Swedish users regulary
sends aprox 20 attendents to the OpenVMS Bootcamp. About the
same number as the rest of Europe together... :-)

Jan-Erik.

>
> Just curious...
>

[abrsvc]

On Feb 1, 2:02 pm, Jan-Erik Soderholm <jan-erik.soderh...@telia.com>
wrote:

> > Just curious...- Hide quoted text -
>
> - Show quoted text -

I support a lab information system currently running on OpenVMS. It
is the most stable platform around. I ahve 1 machine now with an
uptime of 1100+ days. Others are a little less. Fortran, Macro32 and
RDB being the main components.

Dan

[Arne Vajhøj]

On 2/1/2012 1:30 PM, A. W. Dunstan wrote:
> Many years ago I used VMS a great deal& thought it was the greatest thing

> since sliced bread. But where I live I haven't seen anyone interested in
> VMS in about 20 years, and most people have never even heard of it.
>
> My 1st job out of college was for EDS, who used VMS for plant floor
> automation for GM. When I left EDS I worked for a small company (KMS
> Fusion) that did scientific research; they were a heavy user of VMS (and
> RSX) and active in DECUS. We were at v5.5 or so, as I recall.
>

> I've been following this newsgroup for a few months now& see a fair bit of

> activity. Which leads me to wonder what VMS is used for these days? Who
> uses it, and where? Is it still being developed? Does anyone do any
> development of software to run on it?

I will assume that VMS is mostly used for critical stuff that
for whatever reason is very difficult to port to another
platform.

(otherwise some manager would have ordered a migration
years ago)

There are still VMS people around. More than enough for
the work. I think very few new people enter the VMS world.

I expect very few new from scratch projects to start
on VMS these days. But all those critical app that
are still around require maintenance and enhancements.

Fortran, Pascal, C, C++, PL/I, Cobol, Basic etc.etc..

Arne

[JohnF]

Jan-Erik Soderholm <jan-erik....@telia.com> wrote:
> A. W. Dunstan wrote 2012-02-01 19:30:

>> Just curious...

>> Does anyone do any development of software to run on it?
>
> I do. COBOL and Rdb.
> B.t.w, Sweden has a few large VMS users. Swedish users regulary
> sends aprox 20 attendents to the OpenVMS Bootcamp. About the
> same number as the rest of Europe together... :-) Jan-Erik.

Also just curious... what would you say explains Swedish VMS
usage? As Arne mentioned, US managers have all migrated long
ago, unless impractical. So I suppose there's some different
kind of managerial/business-analysis thinking in Sweden that
keeps VMS a viable platform. Could you elaborate about that?
I'd be interested in any reasonable (not religious/strident)
arguments to keep it going. Thanks,
--
John Forkosh ( mailto: j...@f.com where j=john and f=forkosh )

[JF Mezei]

JohnF wrote:

> Also just curious... what would you say explains Swedish VMS
> usage?

Long dark winters, saunas and georgeous women ? :-)

I think that Sweden's VMS usage is skewed by one company, OMS/OMX which
has written VMS software to run stock exchanges and not only uses it for
itself but has sold it to other stock exchanges as well.

[Subcommandante XDelta]

Sigh, :-)

Perhaps they can be convinced to buy VMS from HP, the Swedes aren't
fools.

An impractical fantasy, mayhaps, guilty as charged.

[Jan-Erik Soderholm]

OMX is just one of them

There is a large furniture 4-letter company that is a heavy
VMS user. Every shop around the world has VMS systems and there
is a large VMS center to run all backend processing.

The steel business is also large.

Companies Volvo, SAAB, Ericsson are/was large VMS shops.

[Marc Van Dyck]

on 1/02/2012, A. W. Dunstan supposed :

> Many years ago I used VMS a great deal & thought it was the greatest thing
> since sliced bread. But where I live I haven't seen anyone interested in
> VMS in about 20 years, and most people have never even heard of it.
>
> My 1st job out of college was for EDS, who used VMS for plant floor
> automation for GM. When I left EDS I worked for a small company (KMS
> Fusion) that did scientific research; they were a heavy user of VMS (and
> RSX) and active in DECUS. We were at v5.5 or so, as I recall.
>
> I've been following this newsgroup for a few months now & see a fair bit of
> activity. Which leads me to wonder what VMS is used for these days? Who
> uses it, and where? Is it still being developed? Does anyone do any
> development of software to run on it?
>
> Just curious...

I still manage 20 OpenVMS systems - they are involved in the processing
of all debit and credit card payments in Belgium, about one billion
payments per year. Cobol, C/C++, myself still doing some bits of
Pascal, and ACMS (yes!). All on Itanium blades and EVA SAN storage.
Not the most stable in the company (we also still have NSK which is
even more robust) but certainly the best performance/stability/cost
ratio. Although I must admit that since cross-breeding started to
happen between OpenVMS and HP-UX, this one (11iV3)now comes as a
very close second.

--
Marc Van Dyck

[Michael Kraemer]

In article <mn.12f77dc249...@invalid.skynet.be>, Marc Van Dyck

<marc.gr...@invalid.skynet.be> writes:

> Although I must admit that since cross-breeding started to
> happen between OpenVMS and HP-UX, this one (11iV3)now comes as a
> very close second.

How comes? Don't we hear here over and over again what a
"worthless puddle of #@$%&" HP-UX is?

[Bob Koehler]

In article <6eedneo9GfldGbTS...@megapath.net>, "A. W. Dunstan" <n...@spam.thanks> writes:
>
> I've been following this newsgroup for a few months now & see a fair bit of
> activity. Which leads me to wonder what VMS is used for these days?

Everything, though few use it for word processing or spreadsheets.

> Who
> uses it, and where?

That would be a long list. Smaller than it was in the 80's, but
still too long for a post here. Schools, military, general bussiness,
hobbyists, stock markets, ...

> Is it still being developed?

Yes, but the development team is now in India.

>Does anyone do any
> development of software to run on it?

Absolutely. There is a fairly long list of COTS applications, and
lots of people do in-house or custom software on it.

[Bob Koehler]

In article <4f29d684$0$281$1472...@news.sunsite.dk>, =?ISO-8859-1?Q?Arne_Vajh=F8j?= <ar...@vajhoej.dk> writes:

> I will assume that VMS is mostly used for critical stuff that
> for whatever reason is very difficult to port to another
> platform.
>
> (otherwise some manager would have ordered a migration
> years ago)

There have been many reports of such orders being given anyhow, and
failing.

[Bob Koehler]

In article <3dnki7la6nhtrgqqb...@4ax.com>, Subcommandante XDelta <v...@star.enet.dec.com> writes:
>
> Perhaps they can be convinced to buy VMS from HP, the Swedes aren't
> fools.

Would it actually be better to have VMS maintained in Sweden than
in India? Why?

No point in having VMS go the way of Saab.

[MG]

On 2-2-2012 15:57, Bob Koehler wrote:
> Would it actually be better to have VMS maintained in Sweden than
> in India? Why?

Please tell me you're joking...

- MG

[Jan-Erik Soderholm]

With an hourly cost 5 times of India, it's of course
a joke. :-) Every larger IT consulting company over here
has a lot of indian techies. Don't missunderstand me,
they are usualy *very* bright people!

> > Is it still being developed?
>

> Yes, but the development team is now in India.

Yes, and the development team is now in India.
Why the "but" ?

[Marc Van Dyck]

Michael Kraemer brought next idea :

Don't know - but I'm using both, so I can compare. For applications
that use a SAN and do a lot of I/O, we consider HP-UX far more stable
and efficient than Linux. There are some points for which I still
regret Tru64 - AdvFS and Tru Cluster, to name them - but for what
we are doing, HP-UX does the job whe expect it to do.

--
Marc Van Dyck

[Subcommandante XDelta]

On 2 Feb 2012 08:24:58 -0600, koe...@eisner.nospam.encompasserve.org

(Bob Koehler) wrote:

>In article <6eedneo9GfldGbTS...@megapath.net>, "A. W. Dunstan" <n...@spam.thanks> writes:
>>
>> I've been following this newsgroup for a few months now & see a fair bit of
>> activity. Which leads me to wonder what VMS is used for these days?
>
> Everything, though few use it for word processing or spreadsheets.
>
>> Who
>> uses it, and where?
>
> That would be a long list. Smaller than it was in the 80's, but
> still too long for a post here. Schools, military, general bussiness,
> hobbyists, stock markets, ...

Well, technically, it's not too long to post... :-)

In a properly managed and marketed VMS, such client categorisation and
statistics would be publically available, and such categorisation
would be mappable over time.

As for a list of specific clients of VMS, that is a dual edged sword,
especially for those seeking to steal VMS away to another OS, OTOH,
that is probably going to happen away (seeking to steal customers) and
the virtue of such a customer (Customer opt-in/out of course) list, is
that it demonstrates the extent of the VMS usage and could assuage
customer confidence in VMS's long term viability.

>> Is it still being developed?
>
> Yes, but the development team is now in India.

Burial team you mean.

>>Does anyone do any
>> development of software to run on it?
>
> Absolutely. There is a fairly long list of COTS applications, and
> lots of people do in-house or custom software on it.

Are the accurate lists of COTS's, and VAR's and ISv's available?

All these things are useful to be explicated.

[JF Mezei]

Marc Van Dyck wrote:
> Although I must admit that since cross-breeding started to
> happen between OpenVMS and HP-UX, this one (11iV3)now comes as a
> very close second.

Could you elaborate on what sort of VMS features were implemented in
HP-UX ?

The one big ticket item was to have been clustering, but LaCarly canned
this in favour of a deal with veritas.

[Phillip Helbig---undress to reply]

In article <jgd71h$h8f$1...@reader1.panix.com>, JohnF

<jo...@please.see.sig.for.email.com> writes:

> Also just curious... what would you say explains Swedish VMS
> usage? As Arne mentioned, US managers have all migrated long
> ago, unless impractical. So I suppose there's some different
> kind of managerial/business-analysis thinking in Sweden that
> keeps VMS a viable platform. Could you elaborate about that?
> I'd be interested in any reasonable (not religious/strident)
> arguments to keep it going. Thanks,

While I'm sure Jan-Erik can say much more about this than I can, it is
certainly the case in Sweden that the difference between the managers
and those at the bottom of the totem pole is less than about anywhere
else in the world, however this difference is measured. That might have
something to do with it.

But the real mystery question is whether your flush-right paragraph is a
coincidence or is intentional.

[Phillip Helbig---undress to reply]

In article <4f2a4da9$0$31231$c3e8da3$b280...@news.astraweb.com>, JF

Mezei <jfmezei...@vaxination.ca> writes:

> JohnF wrote:
>
> > Also just curious... what would you say explains Swedish VMS
> > usage?
>
> Long dark winters, saunas and georgeous women ? :-)

Unfortunately, in constrast to the German- and Dutch-speaking countries,
one rarely sees much of gorgeous women IN the sauna in Sweden, since the
rule (there are a few exceptions) is swimsuits required or separated by
sex (at least not both as sometimes in England), while in the the
German- and Dutch-speaking countries the rule (there are a few
exceptions) is mixed and nudity required.

> I think that Sweden's VMS usage is skewed by one company, OMS/OMX which
> has written VMS software to run stock exchanges and not only uses it for
> itself but has sold it to other stock exchanges as well.

IKEA is also another large Swedish company which uses a lot of VMS.
Also the Swedish national lottery runs on VMS. I'm sure Jan-Erik can
mention a few others.

[JF Mezei]

Bob Koehler wrote:

> Would it actually be better to have VMS maintained in Sweden than
> in India? Why?

The success of an OS isn't based on where it is maintained, it is based
on how much the owner wants to spend to improve the OS.


Once you ditch the original and highly experienced real VMS engineers,
whether you hire newbie indians or newbie swedes probably doesn't make
much of a difference.

The problem with India is that it is not only *seen* as low cost, but
also low quality, mostly because most companies that choose to outsource
to India usually choose to pay for low quality support.

But a company could move to India and pay for experienced graduates and
have top quality software done.

Sweden would have an advantage over India because it does have a pool of
experience VMS folks who have used and developped on VMS. (especially if
Saab has let go many IT worker with VMS experience)

But my guess is that HP would more likely turn to folks like Hoff to
help port VMS to the 8086.

<cue in instant denial of any official plans to port VMS to 8086>

[MG]

On 2-2-2012 21:58, JF Mezei wrote:
> The problem with India is that it is not only *seen* as low cost, but
> also low quality, mostly because most companies that choose to outsource
> to India usually choose to pay for low quality support.

Bingo.

- MG

[Single Stage to Orbit]

On Thu, 2012-02-02 at 15:58 -0500, JF Mezei wrote:
>
> But my guess is that HP would more likely turn to folks like Hoff to
> help port VMS to the 8086.
>
> <cue in instant denial of any official plans to port VMS to 8086>

Bugger 8086, why not port to ARM instead? :)
--
Tactical Nuclear Kittens

[Arne Vajhøj]

It has been mentioned.

But it is difficult to say if and how much it is worse
than other large IT projects.

Arne

[Subcommandante XDelta]

On Thu, 02 Feb 2012 23:07:27 +0000, Single Stage to Orbit
<alex....@munted.org.uk> wrote:

>On Thu, 2012-02-02 at 15:58 -0500, JF Mezei wrote:
>>
>> But my guess is that HP would more likely turn to folks like Hoff to
>> help port VMS to the 8086.
>>
>> <cue in instant denial of any official plans to port VMS to 8086>
>
>Bugger 8086, why not port to ARM instead? :)

Simple, kick Microsoft's ass from here to kingdom come.

Who's your Daddy? ;-)

[Subcommandante XDelta]

On Thu, 02 Feb 2012 16:45:20 +0100, MG <marc...@SPAMxs4all.nl>
wrote:

I'll bite, unaccustomed as the VLF is to barking up trees and going
out on a limb, and anyway, I need to give my new dentures a chomp in
the park.

"Indian Call Centre", elegantly summarises the problem.

Yes, it is terribly fashionable in management sub-cultures of modern
corporations to gut them and outsource the jobs of minions lesser than
them to the third-world, whilst, of course, that thin smear of
top-management, continues to earn outrageous incomes and grant
themselves ridicious bonuses, even if the esprit de corp of the
corporation is shot to shyte, and the long term viability of the
enterprise.

So is the exporting of VMS Engineering and whatnot to India a case of
"Fashion Victimisation"? - I do not think so, it's was a deliberate
act of bastardisation to kill off, or savagely attenuate, the residual
cultural legacy of VMS Engineering from the days of DEC - with the
objective of establishing a heavy lid on the preservation and
extension of VMS development and business.

They sacked, super-annuated or made redundant with payouts, many of
which with gagged gobs courtesy of the NDA's that they signed - a
generational purge of VMS advocacy and insurrection within the ranks
of VMS Engineering.

And replaced with what? a culturally crew cut clone army of PhD's and
ersatz beige bizoids, who despite what ever good character and
vertebral systems they might otherwise posess, are all going to be of
the essence of "yes sir, no sir, three bags full sir, anything you
want sir, how would you like your VMS butchered?"

- Otherwise... it's back to the slums for them! -

And they can kiss good-bye their dreams of the "HP way" corporate
life-style (Lite, extremely Lite). -

No vigorous debate and resistance at meetings about the direction of
VMS in India!

Totally supine, totally submissive to some nasty smear of management
back in the states at HP HQ that engineered and implemented this
deliberate hatchet job to kill off VMS.

Come the revolution, VMS Engineering is moved back out of India, and
any of the quality (spine, character and technical skills) are given
the green card treatment, commensurate and proportionate salaries and
the opportunity to taste the Actual American "dream."

No matter where you find your pool of PhD's pay them peanuts and you
get monkey code, when I initially heard of the bastard act of Indian
outsourcing, I not so much thought, "there goes the neighbourhood" as
"there goes the quality of the VMS code-base" - noise is going to
enter, and bug-recursion.

(What was the versions of VMS/VAX and VMS/AXP just prior to the move?)

Back to the Sweden v India question:

Not to put too fine a point on it, India is a slum ridden open sewer
that has recently caught the capitalist disease in concert with an
extremely corrupt governance system - whereas Sweden, like all the
Scandinavian, Nordic, Social Democracies are the anti-thesis of that.

You don't survive Nordic winters unless you dot those i's, cross those
t's, do your due diligence and pull together as a team, and that is
reflected in the contemporary social and governance cultures in
Scandinavia.

Things are thought out well, designed well and implemented well, deep
quality is an ingrained survival trait - so I find it of little wonder
that there is an natural affinity of the Swedish mind to VMS.

Moreover if VMS Engineering and associated camp followers were based
in Sweden the would be substantially less tolerance of bullshit in the
governance of VMS, coming from VMS HQ - people would speak their minds
frankly and forthrightly.

"Oh no! - back to the slums for us!" is not a behavioural constraint
in the Scandinavias, every one is valued as implied by the civil
society infrastructure investment.

I might be elliptically, nay, hyperbolically, off the mark and talking
through my proverbial, but I am happy to be stood corrected.

[Richard B. Gilbert]

On 2/2/2012 6:07 PM, Single Stage to Orbit wrote:
> On Thu, 2012-02-02 at 15:58 -0500, JF Mezei wrote:
>>
>> But my guess is that HP would more likely turn to folks like Hoff to
>> help port VMS to the 8086.
>>
>> <cue in instant denial of any official plans to port VMS to 8086>
>
> Bugger 8086, why not port to ARM instead? :)

How about, "Because H-P is not interested in doing anything more with
VMS than it is contractually obligated to do!"

The hand writing on the wall has been painfully clear for the last
fifteen or twenty years. VMS is an unwanted step child and H-P will
do only enough to meet their contractual obligations.

If somebody wanted VMS badly enough to purchase the rights to VMS and to
assume responsibility for maintenance and H-P's contractual obligations,
H-P would be glad to be rid of it!

They might even pay you to take if off their hands!

[JohnF]

Phillip Helbig---undress to reply <hel...@astro.multiclothesvax.de> wrote:
> JohnF <jo...@please.see.sig.for.email.com> writes:
>
>> Also just curious... what would you say explains Swedish VMS
>> usage? As Arne mentioned, US managers have all migrated long
>> ago, unless impractical. So I suppose there's some different
>> kind of managerial/business-analysis thinking in Sweden that
>> keeps VMS a viable platform. Could you elaborate about that?
>> I'd be interested in any reasonable (not religious/strident)
>> arguments to keep it going. Thanks,
>
> While I'm sure Jan-Erik can say much more about this than I
> can, it is certainly the case in Sweden that the difference
> between the managers and those at the bottom of the totem
> pole is less than about anywhere else in the world, however
> this difference is measured. That might have something to
> do with it.

Thanks, Phillip, very interesting conjecture indeed. Too bad
there's no way in heck that'll ever happen in the USA. I was
hoping for something that would be applicable here (New York
City). I guess your underlying idea is that the "underlings"
who actually know what's what, also have some say about what
happens. ...<thinking>... nope, I don't see how that'll ever
happen here, either:)

> But the real mystery question is whether your flush-right
> paragraph is a coincidence or is intentional.

That would be telling.

[Arne Vajhøj]

On 2/2/2012 3:58 PM, JF Mezei wrote:
> Bob Koehler wrote:
>> Would it actually be better to have VMS maintained in Sweden than
>> in India? Why?
>
> The success of an OS isn't based on where it is maintained, it is based
> on how much the owner wants to spend to improve the OS.
>
>
> Once you ditch the original and highly experienced real VMS engineers,
> whether you hire newbie indians or newbie swedes probably doesn't make
> much of a difference.

But what about indians with lots of experience.

It is my understanding that indians has been working on VMS
for many many years.

> Sweden would have an advantage over India because it does have a pool of
> experience VMS folks who have used and developped on VMS. (especially if
> Saab has let go many IT worker with VMS experience)

Did Saab use VMS?

Arne

[Single Stage to Orbit]

On Thu, 2012-02-02 at 20:40 -0500, Richard B. Gilbert wrote:
> > Bugger 8086, why not port to ARM instead? :)
>
> How about, "Because H-P is not interested in doing anything more with
> VMS than it is contractually obligated to do!"
>
> The hand writing on the wall has been painfully clear for the last
> fifteen or twenty years. VMS is an unwanted step child and H-P will
> do only enough to meet their contractual obligations.
>
> If somebody wanted VMS badly enough to purchase the rights to VMS and
> to assume responsibility for maintenance and H-P's contractual
> obligations, H-P would be glad to be rid of it!
>
> They might even pay you to take if off their hands!

We'll see.
--
Tactical Nuclear Kittens

[Arne Vajhøj]

On 2/2/2012 9:57 AM, Bob Koehler wrote:
> In article<3dnki7la6nhtrgqqb...@4ax.com>, Subcommandante XDelta<v...@star.enet.dec.com> writes:
>>
>> Perhaps they can be convinced to buy VMS from HP, the Swedes aren't
>> fools.
>
> Would it actually be better to have VMS maintained in Sweden than
> in India? Why?

It would certainly be a lot more expensive.

Arne

[Arne Vajhøj]

On 2/2/2012 12:30 AM, JohnF wrote:
> Jan-Erik Soderholm<jan-erik....@telia.com> wrote:
>> A. W. Dunstan wrote 2012-02-01 19:30:
>>> Just curious...

>>> Does anyone do any development of software to run on it?
>>

>> I do. COBOL and Rdb.
>> B.t.w, Sweden has a few large VMS users. Swedish users regulary
>> sends aprox 20 attendents to the OpenVMS Bootcamp. About the
>> same number as the rest of Europe together... :-) Jan-Erik.

>
> Also just curious... what would you say explains Swedish VMS
> usage? As Arne mentioned, US managers have all migrated long
> ago, unless impractical. So I suppose there's some different
> kind of managerial/business-analysis thinking in Sweden that
> keeps VMS a viable platform. Could you elaborate about that?
> I'd be interested in any reasonable (not religious/strident)
> arguments to keep it going.

There are probably many reasons.

Some guesses:
* less focus on next quarters earnings and more focus
on long term earnings
- more real owners
- day trading not so fashionable
result in more long term planning
* lots of big companies (MS rules the small company world)
* tradition of listening to the people at the floor
* strong starting position (both in general and at universities) back
in the 80's and 90's
* the do as everybody else philosophy works both ways: if
no one is leaving VMS then no one is thinking about it

Arne

[JF Mezei]

Arne Vajhøj wrote:

> It is my understanding that indians has been working on VMS
> for many many years.

That is correct. But most of the indian's work has been maintenance,
recompiling VMS applciations for new version of VMS (for instance, they
have had TPU/EVE for many years and haven't done anything except
recompile it).

There may have been a few highly experienced guys with abilities such as
Clustering over IP. But this small core wouldn't be able to handle all
of the VMS core/kernel.

And HP has been quite adament that nobody should compare the indian
staffing levels to staffing levels of the real VMS engineering. So we
have no idea if they replaced the real vms engineers with more or fewer
untrained newbie indians.

[Jan-Erik Soderholm]

I know that the flight/airplane part of SAAB was/is a large VMS user.
Can't tell about the automobile part...


> Arne
>

[Marc Van Dyck]

JF Mezei explained on 2/02/2012 :

I would not say "features" - but to give an example, HP-UX behaviour
with SANs was a nightmare in 11iV1, went a bit better with V2, and
now behaves almost identically as OpenVMS in V3. It certainly is not
a coincidence. And to me, such enhancements are definitely more
important than new bells, whistles or gadgets in a user interface,
because it positively impacts stability and operability.

--
Marc Van Dyck

[Phillip Helbig---undress to reply]

In article <jgfef2$cjd$1...@reader1.panix.com>, JohnF

I've seen people flesh out text with extra spaces (SMS's HELP files for
ZIP and UNZIP come to mind), but yours have no extra spaces. Which is
it:

o You have WAY too much time on your hands.

o You have a nifty DCL routine to do it.

o You're a wizard.

o Other.

[A. W. Dunstan]

abrsvc wrote:

> On Feb 1, 2:02 pm, Jan-Erik Soderholm <jan-erik.soderh...@telia.com>

> wrote:
>> A. W. Dunstan wrote 2012-02-01 19:30:
>>

>> > Many years ago I used VMS a great deal& thought it was the greatest
>> > thing since sliced bread. But where I live I haven't seen anyone
>> > interested in VMS in about 20 years, and most people have never even
>> > heard of it.
>>
>> > My 1st job out of college was for EDS, who used VMS for plant floor
>> > automation for GM. When I left EDS I worked for a small company (KMS
>> > Fusion) that did scientific research; they were a heavy user of VMS
>> > (and RSX) and active in DECUS. We were at v5.5 or so, as I recall.
>>
>> > I've been following this newsgroup for a few months now& see a fair

>> > bit of activity. Which leads me to wonder what VMS is used for these
>> > days?
>>

>> More or less the same things as wen you were active. :-)

>>
>> > Who uses it, and where?
>>

>> Probably those who do, also used it 20 years ago.

>>
>> > Is it still being developed?
>>

>> Yes. Latest version (8.4) released summer 2010.
>> Of course, not the major leaps as when things like clusters was
>> released. Mostly adaptions to new hardware and storage.

>>
>> > Does anyone do any development of software to run on it?
>>
>> I do. COBOL and Rdb.
>>
>> B.t.w, Sweden has a few large VMS users. Swedish users regulary
>> sends aprox 20 attendents to the OpenVMS Bootcamp. About the
>> same number as the rest of Europe together... :-)
>>
>> Jan-Erik.
>>
>>
>>
>>
>>

>> > Just curious...- Hide quoted text -
>>
>> - Show quoted text -
>
> I support a lab information system currently running on OpenVMS. It
> is the most stable platform around. I ahve 1 machine now with an
> uptime of 1100+ days. Others are a little less. Fortran, Macro32 and
> RDB being the main components.
>
> Dan

Interesting. One of my reasons for asking was that I so rarely hear mention
of VMS (outside this newsgroup, anyway). No ads looking for programmers or
sysadmins, certainly no ads encouraging people to give it a try & see how
good it is. Not much point in listing it on my resume...sigh...

--
Al Dunstan, Software Engineer
OptiMetrics, Inc.
3115 Professional Drive
Ann Arbor, MI 48104-5131

"There are two ways of constructing a software design. One way is to
make it so simple that there are obviously no deficiencies. And the
other way is to make it so complicated that there are no obvious
deficiencies."
- C. A. R. Hoare

[Johnny Billquist]

Depends on which SAAB you are talking about.
The airplane/defense manufacturer, or the car company?

The former definitely did, and probably still do. The car manufacturer -
I have no idea...

Johnny

[Johnny Billquist]

Here is another possible explanation: early adopters of new
technology/high tech country.

What would you have choose in the late 70s/early 80s?
DOS 1.0?
Unix?
RDOS?
MVS?

I'd say, at that time, DEC was the company you went to. Later adopters
of technology would probably have started looking more at Unix, and even
later, Windows.

Johnny

[Mike K.]

On Feb 3, 4:21 pm, Johnny Billquist <b...@softjar.se> wrote:
> Here is another possible explanation: early adopters of new
> technology/high tech country.
>
> What would you have choose in the late 70s/early 80s?
> DOS 1.0?
> Unix?
> RDOS?
> MVS?

That all depends on what you needed and how much money you had. MVS,
VM/370, VMS and Unix were all valid options for large-scale multiuser
systems. CP/M (and later MS-DOS) were both valid choices for a small
business, as was the Apple II. If you were using a Nova or a small
Eclipse to run lab equipment, RDOS was even still an option, though if
you were running anything larger, I'm sure that DG would be trying to
sell you on AOS. Really that's the beauty of that era, no matter what
your needs were, there was a system out there that fit it.

Mike

[glen herrmannsfeldt]

Johnny Billquist <b...@softjar.se> wrote:

(snip)

> What would you have choose in the late 70s/early 80s?
> DOS 1.0?
> Unix?
> RDOS?
> MVS?

For small computers it was CP/M for a long time.
I have an Epson QX-10 dating to about 1984. (Now, to get
it running again.)

Otherwise, there was Data General for minicomputers,
Nova on the low end, Eclipse on the higher end.
IBM with various system, System/7, System/1, System/38.
Xerox/Sigma, and HP with the HP-1000, HP-21xx, HP-3000.

Apple II for home, and some laboratory use. Also, Commodore
had some lab-style machines before the VIC-20 and C-64 for
home use.

> I'd say, at that time, DEC was the company you went to. Later adopters
> of technology would probably have started looking more at Unix,
> and even later, Windows.

-- glen

[Arne Vajhøj]

On 2/3/2012 2:25 PM, A. W. Dunstan wrote:
> Interesting. One of my reasons for asking was that I so rarely hear mention
> of VMS (outside this newsgroup, anyway). No ads looking for programmers or

> sysadmins, certainly no ads encouraging people to give it a try& see how

> good it is. Not much point in listing it on my resume...sigh...

Relative few jobs with employees that are not very likely to change jobs
means very few job ads.

Arne

[GreyCloud]

glen herrmannsfeldt wrote:
> Johnny Billquist <b...@softjar.se> wrote:
>
> (snip)
>> What would you have choose in the late 70s/early 80s?
>> DOS 1.0?
>> Unix?
>> RDOS?
>> MVS?
>
> For small computers it was CP/M for a long time.
> I have an Epson QX-10 dating to about 1984. (Now, to get
> it running again.)
>
> Otherwise, there was Data General for minicomputers,
> Nova on the low end, Eclipse on the higher end.
> IBM with various system, System/7, System/1, System/38.
> Xerox/Sigma, and HP with the HP-1000, HP-21xx, HP-3000.
>
> Apple II for home, and some laboratory use. Also, Commodore
> had some lab-style machines before the VIC-20 and C-64 for
> home use.

Back around 1992 I went to school at Iomega corp.
They used a C-64 on testing their read-write heads.
A very low cost solution.

[Arne Vajhøj]

Any different from other countries in the western world??

Arne

[Arne Vajhøj]

On 2/2/2012 7:35 PM, Subcommandante XDelta wrote:
> Come the revolution, VMS Engineering is moved back out of India, and
> any of the quality (spine, character and technical skills) are given
> the green card treatment, commensurate and proportionate salaries and
> the opportunity to taste the Actual American "dream."

I think you will need to find a parallel universe.

Arne

[Arne Vajhøj]

As far as I know then you have never worked in VMS engineering.

I seem to remember people that did work in VMS engineering
not share your opinions.

Arne

[JohnF]

Phillip Helbig---undress to reply <hel...@astro.multiclothesvax.de> wrote:

All the above? -- sadly the first, plenty of the second,
modestly (or not so much) the third, and everybody's the
fourth. As for the first, I'd hazard everyone in this ng
has that -- ditto second and fourth, perhaps third, too.

[JohnF]

Thanks, Arne. To keep things together, let me cut-and-paste
Phillip's remarks...

> > While I'm sure Jan-Erik can say much more about this than I
> > can, it is certainly the case in Sweden that the difference
> > between the managers and those at the bottom of the totem
> > pole is less than about anywhere else in the world, however
> > this difference is measured. That might have something to
> > do with it.
>
> Thanks, Phillip, very interesting conjecture indeed. Too bad
> there's no way in heck that'll ever happen in the USA. I was
> hoping for something that would be applicable here (New York
> City). I guess your underlying idea is that the "underlings"
> who actually know what's what, also have some say about what
> happens. ...<thinking>... nope, I don't see how that'll ever
> happen here, either:)

And also Johnny's remarks...
> Johnny Billquist...

> Here is another possible explanation: early adopters of new
> technology/high tech country.

> What would you have choose in the late 70s/early 80s?
> DOS 1.0?
> Unix?
> RDOS?
> MVS?

> I'd say, at that time, DEC was the company you went to.
> Later adopters of technology would probably have started
> looking more at Unix, and even later, Windows.

Johnny's point about VMS's demise seems the most obvious,
at least from my perspective watching it happen in the US.
Mostly price performance when Ken Olsen obstinately stuck
to his guns despite new cheaper, faster Unix workstations.
Distantly second, platform openness, that wasn't quite so
much an issue around 1990 when Digital's customer-centric
attitude made its proprietariness more acceptable.
But that's past. What I'd be more interested in now is
reasonable, rational talking points, to dissuade managers
from further attrition/migration. And in some fantasyland
future similar talking points to promote new VMS adoption.
Unfortunately, from my experience the above remarks won't
succeed either way. VMS being good (or better or best) is
a necessary but far-from-sufficient condition to adopt it
or just not abandon it. Technical points notwithstanding,
it's got to be ported to x86-64 to stand any chance. Even
then it's very slim at best. Open-sourcing might remove a
"very", but still slim. And neither's happening anyway so
it's just a lost cause, as far as I can see. However, I'd
be delighted to hear any kind of rebuttal.

[Paul Sture]

I came across this one in 1997 from a job agency

"Why don't you VMS people want to change jobs?"

My answer was that we were generally happy in our jobs.

--
Paul Sture

[JF Mezei]

Arne Vajhøj wrote:

> I seem to remember people that did work in VMS engineering
> not share your opinions.

They couldn't share them because they have been forbidden to comment on
the abilities and size of the new team in india.


We know that there were a few who did have experience with such such as
clustering over IP. But this says nothing on whether that group has
suffient EXPEREINECED resources to handle all of the VMS kernel,
clusteting, DCL etc. (not the type of bug that appeared in the DIRECTORY
command in 8.4)

When a coropoation makes such changes and refuses to brag about the new
team, it generally means that the new team has reduced capabilities, and
likely reduced headcount.

And this becomes a serious issue if you no longer have sufficient
experience within group to lead a porting effort to the 8086.

[Paul Sture]

On Fri, 03 Feb 2012 17:53:59 +0000, Phillip Helbig---undress to reply
wrote:

> I've seen people flesh out text with extra spaces (SMS's HELP files for
> ZIP and UNZIP come to mind), but yours have no extra spaces.

Grr. At one point I had a fellow system manager whose EDIT/TPU settings
converted tabs to spaces. He left a trail of files which looked as
though they had been modified but in fact weren't.

--
Paul Sture

[Paul Sture]

On Fri, 03 Feb 2012 22:21:35 +0100, Johnny Billquist wrote:

> Here is another possible explanation: early adopters of new
> technology/high tech country.
>
> What would you have choose in the late 70s/early 80s?
> DOS 1.0?
> Unix?
> RDOS?
> MVS?
>
> I'd say, at that time, DEC was the company you went to. Later adopters
> of technology would probably have started looking more at Unix, and even
> later, Windows.

I chose VMS in 1980, when it was a new and exciting thing, and
minicomputers were not only replacing mainframes (which I was working
on), but providing a much lower cost platform for the many companies
around then who had not really got into IT.



--
Paul Sture

[Arne Vajhøj]

He should have followed practice for the project.

But spaces is a lot more self contained than tabs.

Arne

[Johnny Billquist]

That was the proposed explanation, yes. And one part of that reason was
that labor costs in Sweden were so high, even back then. You really
wanted to automate and computerize.
And we also had (have?) some fairly high tech stuff in Sweden that not
all western world countries do, which also leads to being early adopters.

Johnny

[Arne Vajhøj]

Just the average age of VMS people and the geographical
distance between VMS sites is enough to reduce mobility.

Arne

[Arne Vajhøj]

On 2/4/2012 4:47 AM, JF Mezei wrote:
> Arne Vajhøj wrote:
>> I seem to remember people that did work in VMS engineering
>> not share your opinions.
>
> They couldn't share them because they have been forbidden to comment on
> the abilities and size of the new team in india.

Some did say that had very competent people.

> We know that there were a few who did have experience with such such as
> clustering over IP. But this says nothing on whether that group has
> suffient EXPEREINECED resources to handle all of the VMS kernel,
> clusteting, DCL etc. (not the type of bug that appeared in the DIRECTORY
> command in 8.4)
>
> When a coropoation makes such changes and refuses to brag about the new
> team, it generally means that the new team has reduced capabilities, and
> likely reduced headcount.

So your proof of your claim is the lack information that
contradicts it?

Hmm.

> And this becomes a serious issue if you no longer have sufficient
> experience within group to lead a porting effort to the 8086.

It is impossible to port VMS to 8086.

The reason VMS is not ported to x86-64 is money not skills. If
they money was there they could get people with the skills.

Arne

[Simon Clubley]

On 2012-02-04, Arne Vajhøj <ar...@vajhoej.dk> wrote:
> On 2/4/2012 4:47 AM, JF Mezei wrote:
>> Arne Vajhøj wrote:
>>> I seem to remember people that did work in VMS engineering
>>> not share your opinions.
>>
>> They couldn't share them because they have been forbidden to comment on
>> the abilities and size of the new team in india.
>
> Some did say that had very competent people.
>

It's a pity they were not assigned to producing the VMS patch kits after
responsibility for said patch kits was moved to India.

Simon.

PS: In case my point is too subtle: I don't subscribe to JF style
theories, but instead believe in taking a evidence based approach when
possible and there's now plenty of evidence with which the quality of
the VMS India team can be compared to the old Nashua VMS team.

PPS: And before anyone interprets this as a comment on a culture; it isn't.
It's an observation that the relocation of VMS engineering seems to have
been done too quickly/cheaply by HP. The same problems would have occurred
if responsibility had been handed to a inexperienced team based in the US.

--
Simon Clubley, clubley@remove_me.eisner.decus.org-Earth.UFP
Microsoft: Bringing you 1980s technology to a 21st century world

[Subcommandante XDelta]

On Sat, 4 Feb 2012 23:37:13 +0000 (UTC), Simon Clubley
<clubley@remove_me.eisner.decus.org-Earth.UFP> wrote:

>On 2012-02-04, Arne Vajhøj <ar...@vajhoej.dk> wrote:
>> On 2/4/2012 4:47 AM, JF Mezei wrote:
>>> Arne Vajhøj wrote:
>>>> I seem to remember people that did work in VMS engineering
>>>> not share your opinions.
>>>
>>> They couldn't share them because they have been forbidden to comment on
>>> the abilities and size of the new team in india.
>>
>> Some did say that had very competent people.
>>
>
>It's a pity they were not assigned to producing the VMS patch kits after
>responsibility for said patch kits was moved to India.
>
>Simon.
>
>PS: In case my point is too subtle: I don't subscribe to JF style
>theories, but instead believe in taking a evidence based approach when
>possible and there's now plenty of evidence with which the quality of
>the VMS India team can be compared to the old Nashua VMS team.

Such as? - which is not intended to be a contest of your assertion,
but out of genuine curiousity - such as?

>PPS: And before anyone interprets this as a comment on a culture; it isn't.
>It's an observation that the relocation of VMS engineering seems to have
>been done too quickly/cheaply by HP. The same problems would have occurred
>if responsibility had been handed to a inexperienced team based in the US.

Well that's a cultural problem with HP "Lost Patrol" HQ states side.

HP India is a situational problem, rather than necessarily a cultural
problem.

All up the managerial and technical chains in HP India they are going
to be risk adverse to advocate VMS research and development and
marketing, it's the pragma of survival, few would want to piss off the
yanks and be given the flick.

And no great insult is intended by the VLF, to any Indians from VMS
Engineering, in this post or any other, that are reading comp.os.vms,
(and I trust that they do) I am sure you have vertebral systems and
functional gonads, an appreciation and respect for what VMS is, and
could be, but Mark Hurd put VMS in a heavily gaffer taped box and sent
it packing to Coventry - what can you do?

Besides the technical arms of VMS Engineering in India taking mass
industrial action (a general strike) demanding that VMS be marketed
and that a port to i686/AMD64 be commissioned, what sort of future
does VMS Engineering in India, have, otherwise?

Impractical and phyrric in the circumstances - probably.

These are dire times.

[Arne Vajhøj]

I do not remember Sweden as being that much richer and
more advanced than Denmark, UK, Germany, France, US etc.,
but on the other side I was not that much in Sweden.

Arne

[JF Mezei]

Subcommandante XDelta wrote:

> Besides the technical arms of VMS Engineering in India taking mass
> industrial action (a general strike) demanding that VMS be marketed
> and that a port to i686/AMD64 be commissioned, what sort of future
> does VMS Engineering in India, have, otherwise?

Fro what I have been informaly told, one of the problems is high
turnover. So when they reduce headcount at VMS engineering, those
indians just ove to another department or some other outsourcing firm
next door.

The job market in India is not what it is like in the USA. folks do find
jobs easily.

[Phillip Helbig---undress to reply]

In article <4f2ddee5$0$283$1472...@news.sunsite.dk>,

=?ISO-8859-1?Q?Arne_Vajh=F8j?= <ar...@vajhoej.dk> writes:

> >>> Here is another possible explanation: early adopters of new
> >>> technology/high tech country.
> >>
> >> Any different from other countries in the western world??
> >
> > That was the proposed explanation, yes. And one part of that reason was
> > that labor costs in Sweden were so high, even back then. You really
> > wanted to automate and computerize.
> > And we also had (have?) some fairly high tech stuff in Sweden that not
> > all western world countries do, which also leads to being early adopters.
>
> I do not remember Sweden as being that much richer and
> more advanced than Denmark, UK, Germany, France, US etc.,
> but on the other side I was not that much in Sweden.

Richer? Not necessarily, though more evenly distributed than the others
(except perhaps Denmark). I don't know about Denmark, but people in the
other countries mentioned above tend to have much more fear of big
government than do the people in Sweden. I think many things were
computerized in Sweden long before they were in other countries. There
is not the fear of handing over information to the government and
letting it be used in a variety of computerized contexts. (The point
isn't the information gathered; this is roughly similar in all
countries. However, in Sweden there is a) very good communication
between various government agencies and b) information is by default
public. Things which are compulsory there are forbidden elsewhere.)

[Simon Clubley]

On 2012-02-04, Subcommandante XDelta <v...@star.enet.dec.com> wrote:
> On Sat, 4 Feb 2012 23:37:13 +0000 (UTC), Simon Clubley
><clubley@remove_me.eisner.decus.org-Earth.UFP> wrote:
>
>>It's a pity they were not assigned to producing the VMS patch kits after
>>responsibility for said patch kits was moved to India.
>>

>>PS: In case my point is too subtle: I don't subscribe to JF style
>>theories, but instead believe in taking a evidence based approach when
>>possible and there's now plenty of evidence with which the quality of
>>the VMS India team can be compared to the old Nashua VMS team.
>
> Such as? - which is not intended to be a contest of your assertion,
> but out of genuine curiousity - such as?
>

Clearly, you have not been reading this newsgroup over the last
couple of years. :-)

If you have a account on Eisner, you can also find a dedicated thread
(note 3629.*) discussing the patch issues in the VMS Notes area.

You should also check out the issues around the release of V8.4 discussed
in other threads on Eisner (as well as here in comp.os.vms).

If you don't have a Eisner account and don't want to create one, use
Google to search comp.os.vms for patch problem and similar searches
as the issues were also discussed here.

I've included a sampling of the messages from Eisner below.

Here's something I wrote to HP back in December 2009:

| <<< EISNER::DRA1:[NOTES$LIBRARY]VMS.NOTE;1 >>>
| -< VMS and bundled utilities >-
|================================================================================
|Note 3629.13 quality control issues with Alpha VMS 8.3 patches 13 of 79
|EISNER::CLUBLEY "Simon Clubley" 52 lines 16-DEC-2009 06:05
|--------------------------------------------------------------------------------
|
|From: EISNER::CLUBLEY "Simon Clubley" 16-DEC-2009 06:00:19.69
|To: OPENVMS....@HP.COM
|CC: CLUBLEY
|Subj: Major concerns about current VMS patch quality
|
|Hello,
|
|Over the last few months, since VMS engineering in the US was disbanded
|and it's functions moved elsewhere, there has been a disturbing series
|of incidents with the patches which have come out since then which seem
|to show a basic lack of understanding of VMS on the part of the people
|now entrusted with maintaining VMS.
|
|For example, in one patch, a delete command in the patch's installation
|routine wasn't properly specified showing a lack of basic understanding
|of DCL on the part of the engineers creating the patch as well as showing
|that the patch had not been properly tested prior to release.
|
|In the current patch been discussed on comp.os.vms, it appears that the
|engineers fixing the problems with the current patch did so by removing
|some modules from the patch kit before releasing it back under the exact
|name even though it had been modified.
|
|There are two problems with this: First that a modified patch kit was
|released under the exact same name as the previous version instead of
|been given a new version number, and second, presumably those modules
|were in the patch kit in the first place because they fixed other
|unrelated problems.
|
|These and other incidents, which have been discussed in places like
|Eisner and on comp.os.vms, are causing me (and apparently other people
|as well) major concerns about the quality of VMS in the future.
|
|As a final note, one bit of advice which was issued by HP recently
|in relation to one specific patch stated that even though the patch kit
|recommended terminating installation due to an error been encountered,
|the advice from HP was to continue with the installation anyway.
|
|VMS system managers are not used to been told to continue a procedure
|in the face of a recommendation to abort installation. While I understand
|that in this specific case, continuing was acceptable, I hope that it's
|not considered acceptable practice now to allow installation procedures
|to throw an error on the basis that the system manager is expected to
|ignore them and to continue anyway.
|
|What are your thoughts on this ?
|
|Thank you,
|
|Simon Clubley
|

BTW, the delete command problem was due to a version specification been
missed off the command. The kit made it all the way through "testing" and
out into the wild before the error was discovered.

If something that basic doesn't get caught during testing, then how can
you have confidence in the kit as a whole ?

I got a generic "we're sorry" answer, and a promise to investigate.
I never did get a proper answer.

Here's another one:

| <<< EISNER::DRA1:[NOTES$LIBRARY]VMS.NOTE;1 >>>
| -< VMS and bundled utilities >-
|================================================================================
|Note 3629.15 quality control issues with Alpha VMS 8.3 patches 15 of 79
|--------------------------------------------------------------------------------
|VMS update February 2010
|http://www.openvms.org/stories.php?story=10/02/09/5563218
|
|Includes a link to the following:
|http://www11.itrc.hp.com/service/cki/docDisplay.do?docId=emr_na-c02001423
|
|SUPPORT COMMUNICATION - SECURITY BULLETIN
|Document ID: c02001423
|Version: 1
|HPSBOV02505 SSRT100023 rev.1 - HP OpenVMS RMS, Local Escalation of Privilege
|NOTICE: The information in this Security Bulletin should be acted upon as soon
|as possible.
|Release Date: 2010-02-02
|Last Updated: 2010-02-02
|Potential Security Impact: Local escalation of privilege
|Source: Hewlett-Packard Company, HP Software Security Response Team
|
|VULNERABILITY SUMMARY
|A potential security vulnerability has been identified with certain RMS (Record
|Management Services) patch kits for HP OpenVMS running on ALPHA platforms. The
|vulnerability could be locally exploited resulting in an escalation of
|privilege.
|
|References: CVE-2010-0443
|SUPPORTED SOFTWARE VERSIONS*: ONLY impacted versions are listed.
|RMS patch kit VMS83A_RMS-V1000 dated September 2009 and update kit
|VMS83A_UPDATE-V1100 dated November 2009.

[snip]

Here's another one:

| <<< EISNER::DRA1:[NOTES$LIBRARY]VMS.NOTE;1 >>>
| -< VMS and bundled utilities >-
|================================================================================
|Note 3629.70 quality control issues with Alpha VMS 8.3 patches 70 of 79
|--------------------------------------------------------------------------------
|
|OpenVMS Alpha V8.4 VMS84A_UPDATE V4.0...
|
|<http://h20000.www2.hp.com/bizsupport/TechSupport/Document.jsp?objectID=c02720117>
|
|VMS84A_UPDATE-V0400 replaces APB.EXE, and doesn't WRITEBOOT, which means
|the new APB.EXE won't be used, though the old APB will usually continue
|to work until it gets overwritten or gets erased.
|
|The customer note above also includes text that implies APB.EXE might
|not be created contiguously, and recommends a COPY /CONTIGUOUS to
|resolve that.
|
|Remember to WRITEBOOT early, in other words.
|

I really got annoyed about this one.

This is exactly the kind of issue which will pass your post patch testing
but months/years later, when you have to reboot for a unrelated problem,
you find that your system won't boot anymore. Even worse, because
significant time has probably elapsed since installation, you waste time
looking at what you just did that caused you to take the system down
instead of looking at the faulty patch you installed in the past.

You could have significant downtime with this type of problem before you
realised what the real problem was.

And yes, I am aware that this also happened in the distant past when Nashua
was around, but the point is that Nashua learnt from this and made sure
it didn't happen again.

This is just a sampling of the problems.

The problem is not any one patch (even Nashua issued bad patches every once
in a while), but rather it establishes a pattern of a steady stream of
patches with problems, including basic ones, which should have never
occured or should have been caught during testing before release.

Simon.

[David Weatherall]

To be fair to Paul's colleague, he may well not have been aware. If his
TPU was set to _not_ use tabs and he opened a file to browse, unless he
used Gold/Q or Do QUIT instead of F10 or Ctrl-z, then the files could
be saved without him realising. I find it difficult to believe he did
an 'Eliminate Tabs' deliberately. If he did, however, he deserves any
opprobrium heaped upon him.

The Tab/Space conundrum is one of the reasons I asked for DIFF
/IG=WHITESPACE. Guy added it, God bless him. Unfortunately, I'm trapped
on VMS 7.3-2 so still don't get to use it :) Oh well.

Cheers - Dave

--

[Paul Sture]

On Sat, 04 Feb 2012 20:01:01 +0100, Johnny Billquist wrote:

> That was the proposed explanation, yes. And one part of that reason was
> that labor costs in Sweden were so high, even back then. You really
> wanted to automate and computerize.
> And we also had (have?) some fairly high tech stuff in Sweden that not
> all western world countries do, which also leads to being early
> adopters.

I noticed the same in Switzerland in comparison with the UK. The higher
labour costs do drive investment in machinery. This is perhaps most
evident just observing the building / construction machinery in use.



--
Paul Sture

[Johnny Billquist]

Richer? Not sure where that come into this. I don't think anything I
wrote should have led you to make that connection. As for
technologically advanced, part of that was required by the long
"tradition" of being self-reliant and neutral. I don't remember seeing
any Danish manufactured fighter jets for instance. Since we had to do so
many things ourself, we have (had) a fairly high technology level. That,
in combination with the high labor costs (taxes anyone?) led to Swedish
companies being very interested in automation and computerization very
early. But Sweden have been in a decline since the 70s, I think.

"Why not visit Sweden this year. See the wonderful telephone system."
-- Monthy Python

I remember when I was much younger, and having to talk with operators
when making international calls. Very weird when being used to the fully
automated telephone system in Sweden.

But I'm only trying to find explanations without actually have all the
knowledge or information. Others are free to speculate and find other
answers. :-)

Johnny

--
Johnny Billquist || "I'm on a bus
|| on a psychedelic trip
email: b...@softjar.se || Reading murder books
pdp is alive! || tryin' to stay hip" - B. Idol

[Johnny Billquist]

Good point. Switzerland is actually similar to Sweden in many ways
(although not all). And I accidentally moved to Switzerland from Sweden. :-)

[Subcommandante XDelta]

Oh, dear, well bang goes my back of a beercoaster theory, though this
actuality is just as bad or worse. :-(

At least in my dystopic thesis, the expertise and experience was
present, just deeply sat on.

In your scenario it's a revolving door kiddies splash pool, barely
able to produce patches that don't need patching let alone tackle an
architecture migration.

You don't develop or maintain an operating system without going deep
and staying deep, with management ferrying in supplies by bathysphere.

[Subcommandante XDelta]

On Sun, 5 Feb 2012 10:56:04 +0000 (UTC), Simon Clubley

<clubley@remove_me.eisner.decus.org-Earth.UFP> wrote:

>On 2012-02-04, Subcommandante XDelta <v...@star.enet.dec.com> wrote:
>> On Sat, 4 Feb 2012 23:37:13 +0000 (UTC), Simon Clubley
>><clubley@remove_me.eisner.decus.org-Earth.UFP> wrote:
>>
>>>It's a pity they were not assigned to producing the VMS patch kits after
>>>responsibility for said patch kits was moved to India.
>>>
>>>PS: In case my point is too subtle: I don't subscribe to JF style
>>>theories, but instead believe in taking a evidence based approach when
>>>possible and there's now plenty of evidence with which the quality of
>>>the VMS India team can be compared to the old Nashua VMS team.
>>
>> Such as? - which is not intended to be a contest of your assertion,
>> but out of genuine curiousity - such as?
>>
>
>Clearly, you have not been reading this newsgroup over the last
>couple of years. :-)

Clearly not! :-)

I know I've been thrashing avout like a rhinoceros in a china shop
with an itanic chip on my shoulder, and a blindfold.

Not so much out of the loop as resident in the Oort Cloud.

>If you have a account on Eisner, you can also find a dedicated thread
>(note 3629.*) discussing the patch issues in the VMS Notes area.
>
>You should also check out the issues around the release of V8.4 discussed
>in other threads on Eisner (as well as here in comp.os.vms).
>
>If you don't have a Eisner account and don't want to create one, use
>Google to search comp.os.vms for patch problem and similar searches
>as the issues were also discussed here.
>
>I've included a sampling of the messages from Eisner below.

Thanks for the briefings, context for droll sarcasm now established.
:-)

[Paul Sture]

Quite possibly back then. Even more so now.

--
Paul Sture

[Paul Sture]

On Mon, 06 Feb 2012 13:11:56 +1100, Subcommandante XDelta wrote:

> On Sat, 04 Feb 2012 23:55:45 -0500, JF Mezei
> <jfmezei...@vaxination.ca> wrote:
>
>>Subcommandante XDelta wrote:
>>
>>> Besides the technical arms of VMS Engineering in India taking mass
>>> industrial action (a general strike) demanding that VMS be marketed
>>> and that a port to i686/AMD64 be commissioned, what sort of future
>>> does VMS Engineering in India, have, otherwise?
>>
>>
>>Fro what I have been informaly told, one of the problems is high
>>turnover. So when they reduce headcount at VMS engineering, those
>>indians just ove to another department or some other outsourcing firm
>>next door.
>>
>>The job market in India is not what it is like in the USA. folks do find
>>jobs easily.

I'm not too sure about that. The version I heard half a dozen years ago
when I asked an Indian project manager (who was normally based in India
but over here in Europe to oversee a team) was that it's highly
competitive; if you don't do what management requests there are 200
people waiting to fill your position. IIRC this was said in the context
of asking developers to put in long shifts to hit project deadlines.

(in my experience if you are banging your head late into the evening
against a technical problem, it's best to have a good meal and a good
night's sleep, starting afresh the next day, but I digress.)

> Oh, dear, well bang goes my back of a beercoaster theory, though this
> actuality is just as bad or worse. :-(
>
> At least in my dystopic thesis, the expertise and experience was
> present, just deeply sat on.
>
> In your scenario it's a revolving door kiddies splash pool, barely able
> to produce patches that don't need patching let alone tackle an
> architecture migration.
>
> You don't develop or maintain an operating system without going deep and
> staying deep, with management ferrying in supplies by bathysphere.

Some form of job security is a prerequisite, no? Being in fear of losing
your job at the end of the week is not conducive to concentrating on the
technical problems at hand.

--
Paul Sture

[Paul Sture]

On Sun, 05 Feb 2012 12:56:50 +0000, David Weatherall wrote:

> Arne Vajhøj wrote:
>
>> On 2/4/2012 7:57 AM, Paul Sture wrote:
>> > On Fri, 03 Feb 2012 17:53:59 +0000, Phillip Helbig---undress to reply
>> > wrote:
>> > > I've seen people flesh out text with extra spaces (SMS's HELP files
>> > > for ZIP and UNZIP come to mind), but yours have no extra spaces.
>> >
>> > Grr. At one point I had a fellow system manager whose EDIT/TPU
>> > settings converted tabs to spaces. He left a trail of files which
>> > looked as though they had been modified but in fact weren't.
>>
>> He should have followed practice for the project.
>>
>> But spaces is a lot more self contained than tabs.
>>
>> Arne
>
> To be fair to Paul's colleague, he may well not have been aware. If his
> TPU was set to _not_ use tabs and he opened a file to browse, unless he
> used Gold/Q or Do QUIT instead of F10 or Ctrl-z, then the files could be
> saved without him realising. I find it difficult to believe he did an
> 'Eliminate Tabs' deliberately. If he did, however, he deserves any
> opprobrium heaped upon him.

It was something of a culture clash with someone from a different
division who got transferred in.

> The Tab/Space conundrum is one of the reasons I asked for DIFF
> /IG=WHITESPACE. Guy added it, God bless him. Unfortunately, I'm trapped
> on VMS 7.3-2 so still don't get to use it :) Oh well.
>

Did that only arrive in 8.n?

Spaces give more consistent output for printing, especially in the case
of compiler listings, but for system management stuff I used to stick
with tabs (which feature in SYLOGICALS.TEMPLATE, for example).

--
Paul Sture

[Paul Sture]

On Sun, 05 Feb 2012 17:01:51 +0100, Johnny Billquist wrote:

> On 2012-02-05 14:33, Paul Sture wrote:
>> On Sat, 04 Feb 2012 20:01:01 +0100, Johnny Billquist wrote:
>>
>>> That was the proposed explanation, yes. And one part of that reason
>>> was that labor costs in Sweden were so high, even back then. You
>>> really wanted to automate and computerize.
>>> And we also had (have?) some fairly high tech stuff in Sweden that not
>>> all western world countries do, which also leads to being early
>>> adopters.
>>
>> I noticed the same in Switzerland in comparison with the UK. The
>> higher labour costs do drive investment in machinery. This is perhaps
>> most evident just observing the building / construction machinery in
>> use.
>
> Good point. Switzerland is actually similar to Sweden in many ways
> (although not all). And I accidentally moved to Switzerland from Sweden.
> :-)

I'll admit I don't know much about Sweden, but referring back to a
previous comment about management styles, things can be quite democratic
in Switzerland.

--
Paul Sture

[Paul Sture]

On Sun, 05 Feb 2012 16:59:15 +0100, Johnny Billquist wrote:

> Richer? Not sure where that come into this. I don't think anything I
> wrote should have led you to make that connection. As for
> technologically advanced, part of that was required by the long
> "tradition" of being self-reliant and neutral. I don't remember seeing
> any Danish manufactured fighter jets for instance. Since we had to do so
> many things ourself, we have (had) a fairly high technology level. That,
> in combination with the high labor costs (taxes anyone?) led to Swedish
> companies being very interested in automation and computerization very
> early. But Sweden have been in a decline since the 70s, I think.

I don't think you should underestimate the influence of being neutral.
Switzerland for example still has heavy farming subsidies which date back
to WW II when it was surrounded (as was Sweden then).

> "Why not visit Sweden this year. See the wonderful telephone system."
> -- Monthy Python
>
> I remember when I was much younger, and having to talk with operators
> when making international calls. Very weird when being used to the fully
> automated telephone system in Sweden.

I was pleasantly surprised by the French telephone system when I got
there in the early 1980s. In my schooldays the French system had been
regarded as a joke, but by the time I arrived they had modernised it, and
it was a far superior system to the UK one at the time.

As late as 1976 (and maybe later) there were places in the UK that you
could only reach by calling the operator and asking them to connect you.

--
Paul Sture

[David Weatherall]

Paul Sture wrote:

> On Sun, 05 Feb 2012 12:56:50 +0000, David Weatherall wrote:
>
> > Arne Vajhøj wrote:
> >
> >> On 2/4/2012 7:57 AM, Paul Sture wrote:
> >> > On Fri, 03 Feb 2012 17:53:59 +0000, Phillip Helbig---undress to
> reply >> > wrote:
> >> > > I've seen people flesh out text with extra spaces (SMS's HELP
> files >> > > for ZIP and UNZIP come to mind), but yours have no extra
> spaces. >> >
> >> > Grr. At one point I had a fellow system manager whose EDIT/TPU
> >> > settings converted tabs to spaces. He left a trail of files
> which >> > looked as though they had been modified but in fact
> weren't. >>
> >> He should have followed practice for the project.
> >>
> >> But spaces is a lot more self contained than tabs.
> >>
> >> Arne
> >
> > To be fair to Paul's colleague, he may well not have been aware. If

> > his TPU was set to not use tabs and he opened a file to browse,

> > unless he used Gold/Q or Do QUIT instead of F10 or Ctrl-z, then the
> > files could be saved without him realising. I find it difficult to
> > believe he did an 'Eliminate Tabs' deliberately. If he did,
> > however, he deserves any opprobrium heaped upon him.
>
> It was something of a culture clash with someone from a different
> division who got transferred in.
>
> > The Tab/Space conundrum is one of the reasons I asked for DIFF
> > /IG=WHITESPACE. Guy added it, God bless him. Unfortunately, I'm
> > trapped on VMS 7.3-2 so still don't get to use it :) Oh well.
> >
>
> Did that only arrive in 8.n?
>
> Spaces give more consistent output for printing, especially in the
> case of compiler listings, but for system management stuff I used to
> stick with tabs (which feature in SYLOGICALS.TEMPLATE, for example)

I agree. I use tabs as much as a result of the space saved on an RK05
:) as much as anything else. It is also a nice and easy way of getting
a consistent indent. When coding (Fortran or DCL or even C/Pascal) tabs
to the comments mean that changing code indent levels doesn't mean
realigning the comment fields. One of the reaons I cursed a colleague
who had the same TPU bahaviour as yours :)

Guy implemented '/ig-white' in V8.? He was going to get it back into
7.3-2 but it isn't there on our machines. Pity!


--

Cheers - Dave

[glen herrmannsfeldt]

David Weatherall <nos...@nowheren.no.how> wrote:

(snip)

> I agree. I use tabs as much as a result of the space saved on an RK05
> :) as much as anything else. It is also a nice and easy way of getting
> a consistent indent. When coding (Fortran or DCL or even C/Pascal) tabs
> to the comments mean that changing code indent levels doesn't mean
> realigning the comment fields. One of the reaons I cursed a colleague
> who had the same TPU bahaviour as yours :)

I don't mind tabs in C, where they are part of the standard,
but they are not allowed in standard Fortran. Their use has caused
many questions in comp.lang.fortran where programs didn't
work as expected.

-- glen

[abrsvc]

I believe that tabs ARE allowed in fortran, just not in the free form.
The use of tabs was quite common for F77 source. This would place the
statement within the correct "columns" for fixed source eliminating
the 6 spaces normally required to get past the line number and
continuation areas. Please note that F90 and beyond typically avoid
this.

I think it is the combination of tabs and spaces that caused
confuision. I know that I often "solved" problems at client sites by
making the use of spacing consistant.

Dan

[Ken Fairfield]

On Monday, February 6, 2012 12:37:42 PM UTC-8, abrsvc wrote:
> On Feb 6, 2:55 pm, glen herrmannsfeldt <g...@ugcs.caltech.edu> wrote:

[...]

> > I don't mind tabs in C, where they are part of the standard,
> > but they are not allowed in standard Fortran. Their use has caused
> > many questions in comp.lang.fortran where programs didn't
> > work as expected.

> I believe that tabs ARE allowed in fortran, just not in the free form.

If anything, they were/are disallowed in FIXED source
form. In principle, they'd be no problem in free source
form where white-space is significant (as opposed to the
rules for fixed source form).

I don't have the standard to hand, but IIRC tabs are
still an extension in statements (again, as opposed
to comments and/or character literals). But as noted
above, they are unlikely to cause any problems in
"modern" free source form.

> The use of tabs was quite common for F77 source. This would place the
> statement within the correct "columns" for fixed source eliminating
> the 6 spaces normally required to get past the line number and
> continuation areas. Please note that F90 and beyond typically avoid
> this.

Try posting this to comp.lang.fortran and I think you'll
find a very different reaction from the knowledgeable/
old-timers there... The treatment of tabs in statements
(as opposed to comments) could be very problematic. Even
on VMS.

I recall being incredibly puzzled by a fellow student's code
which would fail to compile after I made minor edits. His
code was making use of a VMS (ne VAX) Fortran extension that
counted a tab in column 1 as a single character against the
80 character card imagelimit, so he could, using EDT, run his
statements all the way to column 80 on a terminal screen.

I was, on the other hand, using Wylvax, a line-by-line
editor, which would convert tabs to blanks on exit (or
at least, in the lines I edited). All of a sudden,
statements were truncated without an apparent visual
cue.

-Ken

[Michael Kraemer]

Johnny Billquist schrieb:

>
> But Sweden have been in a decline since the 70s, I think.
>

was that before or after Abba?

[glen herrmannsfeldt]

abrsvc <dansabr...@yahoo.com> wrote:

(snip)

> I believe that tabs ARE allowed in fortran, just not in the free form.
> The use of tabs was quite common for F77 source. This would place the
> statement within the correct "columns" for fixed source eliminating
> the 6 spaces normally required to get past the line number and
> continuation areas. Please note that F90 and beyond typically avoid
> this.

It is worse. Last I remember, the DEC compilers accepted tabs,
moved to column 8, but counted it as column 7. Since fixed form
ends at column 72, it is very confusing to have different
columns after a tab.

> I think it is the combination of tabs and spaces that caused
> confuision. I know that I often "solved" problems at client sites by
> making the use of spacing consistant.

The Fortran standards, any version, do not make any allowance
for tabs in either programs or input data. Implementations might
allow it as an extension, as DEC did.

-- glen

[Johnny Billquist]

On 2012-02-03 22.39, Mike K. wrote:

> On Feb 3, 4:21 pm, Johnny Billquist<b...@softjar.se> wrote:
>> Here is another possible explanation: early adopters of new
>> technology/high tech country.
>>

>> What would you have choose in the late 70s/early 80s?
>> DOS 1.0?
>> Unix?
>> RDOS?
>> MVS?
>
> That all depends on what you needed and how much money you had. MVS,
> VM/370, VMS and Unix were all valid options for large-scale multiuser
> systems. CP/M (and later MS-DOS) were both valid choices for a small
> business, as was the Apple II. If you were using a Nova or a small
> Eclipse to run lab equipment, RDOS was even still an option, though if
> you were running anything larger, I'm sure that DG would be trying to
> sell you on AOS. Really that's the beauty of that era, no matter what
> your needs were, there was a system out there that fit it.

Late 70s, early 80s were not a time for Unix. It was still very hard to
even get hold of if you weren't academia, as well as being rather
immature. It had recently been ported to the VAX, but was still rather
poor both in features and performance compared to VMS.
MVS and other offerings from IBM was something that was so batch
oriented that it worked for specific customers, but every new business
case meant major development of various components, as well as not
really suitable for a lot of new, interactive, more real time type of
applications.

CP/M did exist, and was sometimes usable. One big problem was that the
hardware you had for those kind of systems were horrible when it came to
reliability. Apple II had the same kind of problems. It was ok to use in
an office with controlled and friendly environment, on your desk. But
that is not where a lot of things were happening back then.

I wonder if AOS had even came to market at that time. Didn't RDOS also
exist in a version for the Eclipse? Anyway, the Eclipse was not a nice
machine (in my opinion), and never really took off. Nova was DGs big
success, and it was definitely an option. But RDOS was not that good.

There were definitely a lot of systems out there back then, but not
really many that were that good. Many of them came and went in a very
short time.

Johnny

[Paul Sture]

On Mon, 06 Feb 2012 19:44:37 +0000, David Weatherall wrote:

> Paul Sture wrote:
>
>>
>> Spaces give more consistent output for printing, especially in the case
>> of compiler listings, but for system management stuff I used to stick
>> with tabs (which feature in SYLOGICALS.TEMPLATE, for example)
>
> I agree. I use tabs as much as a result of the space saved on an RK05 :)
> as much as anything else. It is also a nice and easy way of getting a
> consistent indent. When coding (Fortran or DCL or even C/Pascal) tabs to
> the comments mean that changing code indent levels doesn't mean

> realigning the comment fields. One of the reasons I cursed a colleague
> who had the same TPU behaviour as yours :)

COBOL comment sections which ended up with an asterisk in column 79 (or
80?) fell into that category too.

Another reason for using tabs was the Digital COBOL "Terminal Format" for
sources. A leading tab on a source line was much faster to bash in than
the traditional format of x spaces (a doddle of course on a card punch
with tabs set up for COBOL source), which you had to count carefully to
avoid the compiler throwing a fit. There was a utility to convert
between the two formats, making it easy to see the space "wasted" using
the traditional format (aka ANSI IIRC), and that also applied to RM05s.

You had to select this conversion utility at compiler installation time.
Another source of woe was that the installation procedure defaulted to
not installing it, so if you were working on someone else's machine it
might not be there.

> Guy implemented '/ig-white' in V8.? He was going to get it back into
> 7.3-2 but it isn't there on our machines. Pity!

I was sure there was another way. Perhaps it's brain fade. If it had
been too time consuming I would have written some DCL using f$edit to
collapse multiple spaces and tabs and compare the results.

--
Paul Sture

[Bob Koehler]

In article <9paot5...@mid.individual.net>, "David Weatherall" <nos...@nowheren.no.how> writes:

> Paul Sture wrote:
>>
>> Spaces give more consistent output for printing, especially in the
>> case of compiler listings, but for system management stuff I used to
>> stick with tabs (which feature in SYLOGICALS.TEMPLATE, for example)

Spaces give more consistent output only when you work with people
or software that forgets that printers are set up to do tabs every
8 columns. Honor that standard and you'll have little trouble with
tab based spacing.

> I agree. I use tabs as much as a result of the space saved on an RK05
> :) as much as anything else. It is also a nice and easy way of getting
> a consistent indent. When coding (Fortran or DCL or even C/Pascal) tabs
> to the comments mean that changing code indent levels doesn't mean
> realigning the comment fields. One of the reaons I cursed a colleague
> who had the same TPU bahaviour as yours :)

DEC Fortran tab-format is a time saver. And when you had folks
writing 120,000 lines of Fortran, that added up. I wonder if ANSI
never adopted it (DEC submitted it) because of all the folks who
had trouble with stuff that didn't honor the modulo-8 tabs?

[Paul Sture]

On Tue, 07 Feb 2012 11:25:43 +0100, Johnny Billquist wrote:

> Late 70s, early 80s were not a time for Unix. It was still very hard to
> even get hold of if you weren't academia, as well as being rather
> immature. It had recently been ported to the VAX, but was still rather
> poor both in features and performance compared to VMS.

I remember asking whether Unix should be considered when we were looking
at new systems in the late 1970s and was given a firm "No", on the
grounds of cost and/or availability.

Then there was the somewhat later story, perhaps apocryphal, of someone
who upon learning that his bank was implementing Unix systems, took all
but one dollar out of his account there. When asked why, he said that he
wanted to minimise his risk when the file system garbled everything, but
was prepared to risk a dollar in the event that he suddenly found himself
with several thousand in his account :-)

> MVS and other offerings from IBM was something that was so batch
> oriented that it worked for specific customers, but every new business
> case meant major development of various components, as well as not
> really suitable for a lot of new, interactive, more real time type of
> applications.

We said no to ICL because they wanted us to go with a system which only
supported batch. The service we could offer with our existing PDP
solution meant we could guarantee delivery of replacement parts to any
dealer in the UK by 9:00 am the following day as long as we received the
order by our cut off time in the afternoon. The cut off time was
something like 14:00, but for one-off emergencies we could extend that to
about 15 minutes before the daily pickup at 16:00.

We could not have offered that service using an overnight batch system.
Emergency orders could have stretched to more like 48 hours turnaround.

--
Paul Sture

[Paul Sture]

Probably before and during Abba:

http://en.wikipedia.org/wiki/Kingdom_of_Sweden#Post-war_era

"Sweden, like countries around the globe, entered a period of economic
decline and upheaval, following the oil embargoes of 1973–74 and 1978–79.

In the 1980s pillars of Swedish industry were massively restructured.
Shipbuilding was discontinued, wood pulp was integrated into modernized
paper production, the steel industry was concentrated and specialized,
and mechanical engineering was robotized."

--
Paul Sture

[Fritz Wuehler]

> On 2012-02-03 22.39, Mike K. wrote:
> >> What would you have choose in the late 70s/early 80s?
> >> DOS 1.0?

no

> >> Unix?
no

> >> RDOS?

what is that, Data General? I think I used it and didn't like it.

> >> MVS?

Yeah most of the world ran on it and I still prefer it.

> > That all depends on what you needed and how much money you had. MVS,
> > VM/370,

so far so good

> VMS and Unix were all valid options for large-scale multiuser systems

Uh no. VMS and UNIX were not then and are not now nearly the same scale as
MVS and VM. Companies like banks and insurance companies ran their whole
operation on one MVS or VM box. No major business has ever been able to run
on one VAX or one machine running UNIX.

> MVS and other offerings from IBM was something that was so batch
> oriented that it worked for specific customers, but every new business
> case meant major development of various components, as well as not
> really suitable for a lot of new, interactive, more real time type of
> applications.

That is not true. Since the late 1970s CICS and IMS were both usable enough
that major operations like worldwide travel agencies and reservations
systems, banks, and phone companies all hosted their *online* systems on MVS
CICS or IMS.

> There were definitely a lot of systems out there back then, but not
> really many that were that good. Many of them came and went in a very
> short time.

The best of the lot, MVS and VM are still in production, still in
development, and still making billions a year. There are a lot less sites
now than then, but they're still moving ahead.

[glen herrmannsfeldt]

Bob Koehler <koe...@eisner.nospam.encompasserve.org> wrote:

(snip)

> Spaces give more consistent output only when you work with people
> or software that forgets that printers are set up to do tabs every
> 8 columns. Honor that standard and you'll have little trouble with
> tab based spacing.

(snip)

> DEC Fortran tab-format is a time saver. And when you had folks
> writing 120,000 lines of Fortran, that added up. I wonder if ANSI
> never adopted it (DEC submitted it) because of all the folks who
> had trouble with stuff that didn't honor the modulo-8 tabs?

For me, the problem was that tab went to column 8, but counted
it as column 7, or, if the first character was numeric, column 6.
(I believe that is right, I haven't actually done this
for many years.) Given that, converting tabs to blanks could
result in invalid Fortran statements.

As noted in a post on COBOL, card keypunches had a SKIP key,
when used with the appropriate control card, let you skip
quickly to column 7. It didn't punch a tab key on the card.

The system I used even before VAX/VMS existed, WYLBUR, allowed
tabs to be used for data entry and terminal output, but normally
didn't store them in the file. The WYLBUR internal file format
compresses out blanks, so they don't take more space than tabs.
They are expanded to blanks before sending to the compiler.

The comes up sometimes in comp.lang.fortran. I believe that
even through Fortran 2008 tabs haven't been included in
the standard. Most compilers don't yet support all the features
of Fortran 2003.

-- glen

[abrsvc]

On Feb 7, 1:28 pm, glen herrmannsfeldt <g...@ugcs.caltech.edu> wrote:

You are correct, the use of TABs was not a standard but an extension
that was supported by many compilers. My error and mis-statement.
The confusion around where TAB brought the line in relation to
continuation characters etc for FORTRAN was a source for many
problems. This was the case when inadvertantly both the TAB and
spacing was used. Attempts to line up the code using spaces resulted
in errors.

Dan

[Bob Koehler]

In article <jgrqh0$iva$1...@speranza.aioe.org>, glen herrmannsfeldt <g...@ugcs.caltech.edu> writes:
>
> For me, the problem was that tab went to column 8, but counted
> it as column 7, or, if the first character was numeric, column 6.
> (I believe that is right, I haven't actually done this
> for many years.) Given that, converting tabs to blanks could
> result in invalid Fortran statements.

There were a great many conversion utilities which knew this
and would convert accordingly. But if all your systems were
DEC, or if ANSI had adopted it as a standard, you'ld have
no problem.

Later, HP-UX, Solaris, ..., all touted "VAX compatable Fortran",
and they knew how to use DEC tab formatted code. I beleive the
current incarnation of gfortran also knows.

[Paul Sture]

On Tue, 07 Feb 2012 08:07:16 -0600, Bob Koehler wrote:

> In article <9paot5...@mid.individual.net>, "David Weatherall"
> <nos...@nowheren.no.how> writes:
>> Paul Sture wrote:
>>>
>>> Spaces give more consistent output for printing, especially in the
>>> case of compiler listings, but for system management stuff I used to
>>> stick with tabs (which feature in SYLOGICALS.TEMPLATE, for example)
>
> Spaces give more consistent output only when you work with people or
> software that forgets that printers are set up to do tabs every 8
> columns. Honor that standard and you'll have little trouble with tab
> based spacing.

I was thinking of the compilers available years ago. I just compiled a
terminal format COBOL program using an Alpha running VMS 8.3 and it
appears from an examination of the listing that the tabs and spaces are
in the right place for a decent printout. The compiler actually outputs
a tab in the first byte of each listing record, then the correct number
of spaces before the compiler generated line number.

I must admit I don't have an LNO3 handy for the acid test of printing it
out, but my neighbours are probably thankful for that :-)

>> I agree. I use tabs as much as a result of the space saved on an RK05
>> :) as much as anything else. It is also a nice and easy way of getting
>> a consistent indent. When coding (Fortran or DCL or even C/Pascal) tabs
>> to the comments mean that changing code indent levels doesn't mean
>> realigning the comment fields. One of the reaons I cursed a colleague
>> who had the same TPU bahaviour as yours :)
>
> DEC Fortran tab-format is a time saver. And when you had folks
> writing 120,000 lines of Fortran, that added up. I wonder if ANSI
> never adopted it (DEC submitted it) because of all the folks who had
> trouble with stuff that didn't honor the modulo-8 tabs?

That time saving was also true with DEC COBOL "Terminal format". I
suspect other languages also benefited.

--
Paul Sture

[Paul Sture]

On Tue, 07 Feb 2012 18:28:48 +0000, glen herrmannsfeldt wrote:

> As noted in a post on COBOL, card keypunches had a SKIP key, when used
> with the appropriate control card, let you skip quickly to column 7. It
> didn't punch a tab key on the card.

The card keypunches I witnessed indeed did not generate a tab character;
they just left empty spaces, much as an old fashioned typewriter would do.

--
Paul Sture

[Phillip Helbig---undress to reply]

In article <jgpmjf$haj$1...@solani.org>, Michael Kraemer <M.Kr...@gsi.de>
writes:

> Johnny Billquist schrieb:
>
> > But Sweden have been in a decline since the 70s, I think.
>
> was that before or after Abba?

Aaahhh yes, ABBA: all that youthful sexual energy enticingly packed in
skin-tight lycra---and that's just the blokes!

I've never lived in Sweden, but I've spent about 8 months or so there
over the years. The best time was probably about 1955--1975 (no, I
wasn't there then). The country has a population of about 8 million,
and managed to produce Abba, Ericsson, (part of) ABB, SKF, SAAB, Volvo,
Marabou, IKEA, Husqvarna. Sweden is still not a member of NATO and
joined the EU (by a very close vote) only in 1995 (and has managed to
avoid the common currency via a loophole (made use of in order to
conform to plebiscite results) which has since been closed for other
countries). It used to be a really interesting place. Some of this
still survives, for example the efficient bureaucracy which makes use of
the "personal number" as a primary key, the idea that information is
public by default (including every person's tax return---tax paid,
income, savings), measuring wind speed in meters per second. I don't
know if it's still the case that telephone keypads are like computer
keypads (i.e. low numbers at the bottom), not like telephone keypads in
most of the rest of the world.

SAAB parallels Sweden in many ways: they used to be really innovative,
unique, immediately recognizable. Starting around the time GM became
involved, quality dropped, there was little innovation, they looked like
many other cars. I still enjoy being there and like many things about
it, but I think the era shaped by people like Tage Erlander is gone for
good.

[Phillip Helbig---undress to reply]

In article <jgs9c9$ibg$1...@online.de>, hel...@astro.multiCLOTHESvax.de

(Phillip Helbig---undress to reply) writes:

> Aaahhh yes, ABBA: all that youthful sexual energy enticingly packed in
> skin-tight lycra---and that's just the blokes!

> I still enjoy being there and like many things about
> it, but I think the era shaped by people like Tage Erlander is gone for
> good.

I just realize that I mourn for Sweden like I mourn for DEC. Nostalgia
ain't what it used to be.

[JF Mezei]

Bob Koehler wrote:

> There were a great many conversion utilities which knew this
> and would convert accordingly. But if all your systems were
> DEC, or if ANSI had adopted it as a standard, you'ld have
> no problem.

My experience was with Cobol. VAX Cobol had a lot of flexibility in the
format of the code. (the language has rules about indentation to signify
that a portion "belongs" to the portion with less indentation etc)


When head office decided to adopt computers, they looked at ours and
were very impressed. Because DEC squandered the opportunity, Data
General won the contract. They were supplied with copies of some of our
programs and said that both the C and Cobol compiled cleanly without
changes.

Head office believed them and I was prevented from seeing the results.


DG cobol not only required strict indentation with spaces (no tabs) but
also that everything be uppercase. And their version of cobol lacked
the report writer feature which our programs relied heavily on.


And interestingly, getting Cobol to call some of our C routines on DG
was hell whereas it was quite easy on VAX.


The morale of the story is that when you start to use vendor specific
features that make your life easier, you make it much harder to port
when that vendor screws you.

[JF Mezei]

Phillip Helbig---undress to reply wrote:

> I just realize that I mourn for Sweden like I mourn for DEC. Nostalgia
> ain't what it used to be.

No reason to mourn Sweden. Check out the picture of Sweden's princess
Madeleine ...

http://www.dslreports.com/speak/slideshow/26841208?c=1724007&ret=L2ZvcnVtL3JlbWFyaywyNjg0MTIwOA%3D%3D



:-)

[Paul Sture]

On Tue, 07 Feb 2012 16:33:20 +0100, Fritz Wuehler wrote:

>> On 2012-02-03 22.39, Mike K. wrote:
>> >> What would you have choose in the late 70s/early 80s? DOS 1.0?
>
> no
>
>> >> Unix?
> no
>
>> >> RDOS?
>
> what is that, Data General? I think I used it and didn't like it.
>
>> >> MVS?
>
> Yeah most of the world ran on it and I still prefer it.
>
>> > That all depends on what you needed and how much money you had. MVS,
>> > VM/370,
>
> so far so good
>
>> VMS and Unix were all valid options for large-scale multiuser systems
>
> Uh no. VMS and UNIX were not then and are not now nearly the same scale
> as MVS and VM. Companies like banks and insurance companies ran their
> whole operation on one MVS or VM box. No major business has ever been
> able to run on one VAX or one machine running UNIX.

What I came across in the eighties and nineties were shops which used
mainframes for the commercial side of things and VMS for production and
warehousing. Banks and insurance companies also had use for VMS
alongside their mainframes.

--
Paul Sture

[Paul Sture]

Thank you for brightening up my morning :-)

--
Paul Sture

[Bob Koehler]

IIRC, the SKIP key and TAB key were the same, or one in the location
where you expected the other.

I was one of the few programmers who knew how to program a drum card,
and my drum card was set up for FORTRAN IV: 1 alphanumeric, 4
numerics, one shifted key (I prefered # as my continuation
character), more alphanumerics out to column 72, then numerics to the
end of the card.

It was so much more convenient to bang that skip key than to count
spaces.

[Paul Sture]

I never programmed one of the punch machines but I did program some ICL
workstations which mimicked them. I never used #, since UK keyboards
generated a £ instead (and guess what - not all printers could do that).

I also used a sleeker and lighter version of this handpunch, but that was
either for correcting data cards or writing wrappers around data cards.

http://bit.ly/yL1451

It was surprising how quickly you could learn to use one of these, and at
decent speed too.

--
Paul Sture

[Nomen Nescio]

> What I came across in the eighties and nineties were shops which used
> mainframes for the commercial side of things and VMS for production and
> warehousing. Banks and insurance companies also had use for VMS
> alongside their mainframes.

That was not my experience (during the seventies and eighties) but perhaps
it was regional. I never saw a shop that had anything but IBM computing
hardware (peripherals may have varied but were majority IBM) and this
includes numerous banks and state and local government shops. My friends
who worked for insurers remember the situation as I do as well. I didn't
work in New England although I went to uni there and that makes a big
difference for sure. DEC had good market penetration in New England and gave
lots of freebies (including PDP7s!) to local universities so it was probably
much more common to see their stuff and other non-IBM in that region. I
think the rest of the country was almost 100% IBM for large businesses until
the nineties where they started thinking they could save money by throwing
out the mainframe. They never were able to, they just added more varied
machines and networks and had to keep what they had before as well.

[glen herrmannsfeldt]

Bob Koehler <koe...@eisner.nospam.encompasserve.org> wrote:

(snip, I wrote)

>>> As noted in a post on COBOL, card keypunches had a SKIP key, when used
>>> with the appropriate control card, let you skip quickly to column 7. It
>>> didn't punch a tab key on the card.

(snip)

> IIRC, the SKIP key and TAB key were the same, or one in the location
> where you expected the other.

Yes, I believe the key is SKIP and not TAB, but those using drum
cards knew what it did.

> I was one of the few programmers who knew how to program a drum card,
> and my drum card was set up for FORTRAN IV: 1 alphanumeric, 4
> numerics, one shifted key (I prefered # as my continuation
> character), more alphanumerics out to column 72, then numerics to the
> end of the card.

I think I knew how to make one, but most of the time there
was already one there. As you note, it does a numeric shift
for column 6. Reminds me, once I tried to use / as a continuation
character, which would have been fine, except that with numeric
shift on it is instead 0. I think I even knew at the time that 0
does not indicate a continuation, but forgot that numeric
shift was on.

In addition to the usual numeric row at the top, there are also
ten keys in the shape of a numeric keypad, that are otherwise
non-numeric characters. (For the convenience of punching cards
with only numerical data. There is an Alpha shift key to override
the drum card shift, if you remember to use it.)

> It was so much more convenient to bang that skip key than to count
> spaces.

Yes, but it doesn't take long for your thumb to remember how
to count to six.

-- glen