Amazing, two new articles on Computerworld.com actual mention OpenVMS!
Ken Robinson
issues" <http://www.computerworld.com/action/article.do?command=viewArticleBasic&articleId=9025228&intsrc=hm_list>
and "Hurd: HP will show users how to run IT"
<http://www.computerworld.com/action/article.do?command=viewArticleBasic&articleId=9025227&intsrc=hm_list>
Posted with no comment....
Ken
JF Mezei
> Posted with no comment....
Well, did anyone think that *I* would be able to widthhold comments
about this ?
Ann Livermore:
##
In particular for our OpenVMS customers, we are very focused on our
installed base. We want our installed base to be happy, and if or when
they ever want to migrate, we want them to migrate to another HP platform.
##
This pretty much confirms the Stallard memo of May 7th 2002 not being a
glitch but truly representing HP corporate policy.
If that Livermore isn't an official announcement that HP intends to kill
VMS, I don't know what it.
Dirk Munk
> Ken Robinson wrote:
>
>> Posted with no comment....
>
> Well, did anyone think that *I* would be able to widthhold comments
> about this ?
>
> Ann Livermore:
> ##
> In particular for our OpenVMS customers, we are very focused on our
> installed base. We want our installed base to be happy, and if or when
> they ever want to migrate, we want them to migrate to another HP platform.
> ##
Yep, it's the same kind of bullshit that I even heard in the Digital
days. "We love our VMS customers, but should they want to move to
Windows-NT or Unix we will be most happy to .. blah blah blah blah
DaveG
I read the words "if and when they ever want to migrate". I don't
interpret that statement as being an "official announcement" to kill
VMS.
Rich Alderson
>> Ken Robinson wrote:
>>> Posted with no comment....
I don't think it's the phrase you quote, I think it's the statement that "we
are very focused on our installed base" that has people tolling the knell: If
they are focused on the installed base, that doesn't bode well for expansion of
same.
--
Rich Alderson | /"\ ASCII ribbon |
ne...@alderson.users.panix.com | \ / campaign against |
"You get what anybody gets. You get a lifetime." | x HTML mail and |
--Death, of the Endless | / \ postings |
Richard Maher
> I don't think it's the phrase you quote, I think it's the statement that
"we
> are very focused on our installed base" that has people tolling the knell:
You're right! I'm sure it was only last week that she was trumpeting an
aggressive green-field expansion strategy for vms. What an absolute 180
degree back-flip! I feel betrayed :-(
Ok so *this* time VMS is *really* dead? You're willing to concede that the
last couple of thousand pronouncements and obituaries over the last 15 years
may have been a tad premature, but this time *for sure* hey Bullwinkle?
Alright, I'll agree; we're all washed up and we heard it here first. You
were right all along! No need for next month's announcement then? And the
week after that's? and on and on and fucking on ?
Yes by all means teach HP a lesson and change jobs and hardware platforms
*now*, and while you're all at it, why not change newsgroups as well?
Regards Richard Maher
"I'm feeling much better now. . " - VMS as cov eagerly carries it on their
shoulders to the plague cart.
"Rich Alderson" <ne...@alderson.users.panix.com> wrote in message
news:mddwsxz...@panix5.panix.com...
JF Mezei
> I read the words "if and when they ever want to migrate". I don't
> interpret that statement as being an "official announcement" to kill
> VMS.
It is close to an announcement as can be made without actually killing
VMS. It is a warning to get the hell off VMS ASAP because HP's
intentions are to convince VMS customers to migrate to other HP platforms.
Combine this with the fact that VMS was the only major product whose
future was not announced until after the merg3er was signed on May 7th,
and on that day, after months of waiting, we are told by Stallard that
he expects VMS customer to migrate to HP-UX. Combine that with the total
lack of marketing even after VMS ran on that IA64 thing, not even
allowing a press release to the news wires about VMS now running on HP
blades.
Combine this with Cerner having struck a deal with HP to drop VMS and
port to HP-UX, and Gembase having been told that VMS was dead and thus
dropping plans for further upgrades.
If that Ann Livermore thing think she has what it takes to make a good
CEO, and this absolute blunder was truly intentional, then she is far
more incompetant than anyone could have imagined. If she is not
incompetant, then this was not a blunder and was a very strong message
to the VMS user community.
Tom Linden
<jfmezei...@vaxination.ca> wrote:
> Ken Robinson wrote:
>
>> Posted with no comment....
>
> Well, did anyone think that *I* would be able to widthhold comments
> about this ?
>
> Ann Livermore:
> ##
> In particular for our OpenVMS customers, we are very focused on our
> installed base. We want our installed base to be happy, and if or when
> they ever want to migrate, we want them to migrate to another HP
> platform.
> ##
From a bottom line perspective this is really stupid. Why would HP want
to migrate a
high margin VMS customer to a lower margin HP-UX customer, or even lower
Winodws. Can
you think of any other business that operates in this fashion?
>
> This pretty much confirms the Stallard memo of May 7th 2002 not being a
> glitch but truly representing HP corporate policy.
>
> If that Livermore isn't an official announcement that HP intends to kill
> VMS, I don't know what it.
It demonstrates a serious lack of business acumen.
--
PL/I for OpenVMS
www.kednos.com
David J Dachtera
Well, you can take that two (or more) ways, JF.
"Another HP platform" could mean hardware platform (I64, x86-64) or software
platform (OpenVMS-I64, PH-UX) ... and various combinations ...
--
David J Dachtera
dba DJE Systems
http://www.djesys.com/
Unofficial OpenVMS Marketing Home Page
http://www.djesys.com/vms/market/
Unofficial Affordable OpenVMS Home Page:
http://www.djesys.com/vms/soho/
Unofficial OpenVMS-IA32 Home Page:
http://www.djesys.com/vms/ia32/
Unofficial OpenVMS Hobbyist Support Page:
http://www.djesys.com/vms/support/
David J Dachtera
> [snip]
"...but I'm not dead yet!"
"You will be."
"...but I'm gettin' better!"
David J Dachtera
>
> [snip] Can
> you think of any other business that operates in this fashion?
Government?
> >
> > This pretty much confirms the Stallard memo of May 7th 2002 not being a
> > glitch but truly representing HP corporate policy.
> >
> > If that Livermore isn't an official announcement that HP intends to kill
> > VMS, I don't know what it.
>
> It demonstrates a serious lack of business acumen.
... and you expected ... what?
Tom Linden
<djes...@spam.comcast.net> wrote:
>> [snip] Can
>> you think of any other business that operates in this fashion?
> Government?
You missed the point, a company has two products, A has 20% margins and
sells 100 units
B has 10% margin and sells 500 units.
--
PL/I for OpenVMS
www.kednos.co
Michael Kraemer
>
> If that Livermore isn't an official announcement that HP intends to kill
> VMS, I don't know what it.
Relax. That's more or less what she said already 2 years ago,
nothing new here.
JF Mezei
> Relax. That's more or less what she said already 2 years ago,
> nothing new here.
Confirmation that HP has a "stay the course" policy and won't change it.
It is perfectly happy to actively sink the VMS ship (along with itanic).
Getting Cerner to drop VMS in favour of HP-UX as well as Livermore's
sayings are no longer just passive "let VMS live happily in its
corner". HP is now actively sending signals that VMS is on its way out.
And what is new is that major ISVs have begun to abandon ship.
I fear this is the last chance to save VMS. The next couple of months
will be critical to try to reverse HP's policy.
And out of principle, we must ensure this causes the biggest possible
loss of customers to HP. If customers decide to remain with HP despite
this action, it will set a very dangerous precedent for other vendors to
handle customers in unesthical ways and lie to them about tru strategies
for certain products.
Our only hope is to convince Hurd that he has been mislead by
Stallard/Livermore and to overrule their desire to act as pirates and
plunder as many VMS customers as possible onto the HP-UX ship.
Bill Gunshannon
Well, over the summer I am taking a "Stat for Business" course. We
are into "Decision Making" as of last night. Of course, the examples
we work with are all business related. And. if what I am learning is
really how they do it, I am not surprised that American businesses are
so far in the crapper. If I applied the logic used in these methods
to VMS it would have been killed long ago.
for example, we had a thing on "Deicision making when no probability
data is available." There are three methods. One is extremely pess-
imistic. One is extremely optomistic. And the third looks totally
random and could probably be done just as well with a dartboard with
out all the paper and pencil work.
Hmmmm...... Maybe I can ask some questions here and then see if I
can lay it out according to the methods in my textbook. Then everyone
can either laugh, cry or both.
bill
--
Bill Gunshannon | de-moc-ra-cy (di mok' ra see) n. Three wolves
bi...@cs.scranton.edu | and a sheep voting on what's for dinner.
University of Scranton |
Scranton, Pennsylvania | #include <std.disclaimer.h>
AEF
C'mon. You're putting the worst possible spin on it. She says she
wants their installed base to be happy, AND if or when, not "now". So
that means if the customers want to stay with VMS, she wants them to
be happy and they can stay with VMS.
Now if she said "we want to move them off VMS", then you'd have a
case.
AEF
Bill Todd
...
> that means if the customers want to stay with VMS, she wants them to
> be happy
Of course: right up until the day that the same thing happens to VMS
that happened to Alpha, and for the same reasons. She surely wouldn't
want to preannounce VMS's demise while people are still willing to pay
good money for it - far better to string things out until what they're
paying is no longer sufficient to keep up even a pretense.
As contrasted, of course, with actually *investing* in the future of a
product that *has* an intended future.
VMS is certainly not dead yet, and for some people that's enough. It's
just the people for whom that is *not* enough who should delve behind
the happy talk - not that this has been all that difficult since the
Alphacide: the signs of senescence have been plentiful, and if it
weren't for a continually diminishing but not yet completely irrelevant
cadre of faithful friends the family would have pulled the plug long ago.
- bill
Dr. Dweeb
You need to learn to parse "corporate speak"
Dweeb (now an MS jockey)
Dr. Dweeb
> Richard Maher wrote:
>> [snip]
>
> "...but I'm not dead yet!"
>
> "You will be."
>
> "...but I'm gettin' better!"
"No you're not"
"I feel fine"
Dr. Dweeb
VMS does not count for HP. Get used to living in the twilight zone.
Dweeb
Dr. Dweeb
Parsing error - see previous post-
This is not physics, this is the careful art of legal and marketing vetted
corporate speak where the words and message are not intended to be the same.
Dweeb
> AEF
Doug Phillips
It says what it says. No spin is needed. "very focused on our
installed base" and "we want them to migrate to another HP platform."
No pursuing *new* customers or migrating *to* OpenVMS. Throwing in
"happy" and "if or when" doesn't change the message. No mention of
OpenVMS *wins*; if there are any, why not jump at this opportunity to
say so? If there are currently *new* customer's with installations in
process, and *new* customers in the process of considering or actually
buying OpenVMS, HP is apparently not focused on them and don't care
whether they're happy or not.
"If or when" could have been a positive if she'd said 'migrating
_to_or_ from OpenVMS', but she didn't.
Her assumption and desire should be obvious to anyone who reads that
statement. Obviously, her idea of what makes the installed OpenVMS
customer base *happy* is a bit skewed from reality.
If you were serious about promoting OpenVMS, is that how you would
have responded to the question?
Bill Gunshannon
JF Mezei <jfmezei...@vaxination.ca> writes:
> DaveG wrote:
>> I read the words "if and when they ever want to migrate". I don't
>> interpret that statement as being an "official announcement" to kill
>> VMS.
>
>
> It is close to an announcement as can be made without actually killing
> VMS. It is a warning to get the hell off VMS ASAP because HP's
> intentions are to convince VMS customers to migrate to other HP platforms.
>
> Combine this with the fact that VMS was the only major product whose
> future was not announced until after the merg3er was signed on May 7th,
> and on that day, after months of waiting, we are told by Stallard that
> he expects VMS customer to migrate to HP-UX. Combine that with the total
> lack of marketing even after VMS ran on that IA64 thing, not even
> allowing a press release to the news wires about VMS now running on HP
> blades.
>
> Combine this with Cerner having struck a deal with HP to drop VMS and
> port to HP-UX, and Gembase having been told that VMS was dead and thus
> dropping plans for further upgrades.
Yep. Just saw an ad for a Cerner guy. No experience with Cerner apps
needed, will send to school. Requires Unix experience!!
JF Mezei
> C'mon. You're putting the worst possible spin on it. She says she
> wants their installed base to be happy,
To make the installed base happy, HP needs to stop making allusions to
them wanting VMS customers to port to other HP products.
To make the installed base happy, HP needs to market VMS so that VMS
becomes a palatable platform to higher management.
The current method HP is employing is not making the installed base
HAPPY. Talk to all those who are seeing ISV support for their middle
ware pulled from under them.
Dr. Dweeb
R.I.P.
VMS fed my family for more than 20 years. I have got over what Palmaer,
Curly, Carley and now Hurd have done.
Move on, and as a matter of principle, never buy anything from HP again. I
certainly have not spent 1 cent there since the CompaQ days, and have made
sure that others did likewise. It makes no difference I know, but it makes
me feel better.
Of course spending at M$ is worse, but what is a poor boy gonna do? I have
to pay for the grits somehow!
I know of few VMS sites that are not actively engaged in deserting VMS -
they have no choice and they know it. It is just a matter of planning and
execution. Most seem to be going "anywhere but HP", but the ink salesmen
couldn't care less I guess. They certainly treat the customers that way.
cheers
Dweeb
Richard Maher
For those of you who, after reading this thread, may be thinking that Ann
Livermore (I've got no idea who she is in HP and don't care) was responding
to a question such as "What are your feelings with regard to the future
strategic directionsof VMS?" or "How does your VMS strategy differ from your
other OS stable?" or "Just speak randonmly on any issue for 30secs", here's
the actual question she was reponding to: -
[That expertise and quality are the issue. Last fall I spoke with two
high-profile HP partners who are OpenVMS consultants. They were very
concerned about services being outsourced to places like China and India,
because they said the support technicians there simply don't have the
expertise that's required and aren't getting the training they need. What's
your response?]
I read that like the reporter, and whoever he was talking to (VMS people "In
particular"), accused HP of shafting the installed base by outsourcing
support to Bangalore, Mumbai and Brisbane. So she says something as bland as
[We value the installed base enough to want to keep them sweet so that they
stick with us if and when they move architectures.] or [Your call is
important to us!] and it gets translated in COV to [I'm sick of that
grave-dodging bastard VMS! Kill it - And kill it now!] Absolute disbelief
:-(
But whatever you do, listen to Dr. Dweeb; VMS *is* dead! (and has been for
years) Move on! So there can be absolutely *no* reason to keep telling
everybody about its death for the next days, months, YEARS now can there?
Cheers Richard Maher
"Ken Robinson" <kenr...@gmail.com> wrote in message
news:7dd80f60706191324k5c0...@mail.gmail.com...
David J Dachtera
>
> On Tue, 19 Jun 2007 18:45:09 -0700, David J Dachtera
> <djes...@spam.comcast.net> wrote:
>
> >> [snip] Can
> >> you think of any other business that operates in this fashion?
> > Government?
>
> You missed the point, a company has two products, A has 20% margins and
> sells 100 units
> B has 10% margin and sells 500 units.
Actually, I think it's more like:
A - 48% margin, 1,000 units
B - 10% margin, 2,000,000 units
AEF
wrote:
> Tom Linden wrote:
>
> > On Tue, 19 Jun 2007 18:45:09 -0700, David J Dachtera
> > <djesys...@spam.comcast.net> wrote:
>
> > >> [snip] Can
> > >> you think of any other business that operates in this fashion?
> > > Government?
>
> > You missed the point, a company has two products, A has 20% margins and
> > sells 100 units
> > B has 10% margin and sells 500 units.
>
> Actually, I think it's more like:
>
> A - 48% margin, 1,000 units
>
> B - 10% margin, 2,000,000 units
Now ... imagine the huge profit gains from switching even a small
percentage of B systems to A systems!
>
> --
> David J Dachtera
[...]
AEF
David J Dachtera
Ah, yes... Great minds... Eh?
Tom Linden
<djes...@spam.comcast.net> wrote:
> Tom Linden wrote:
>>
>> On Tue, 19 Jun 2007 18:45:09 -0700, David J Dachtera
>> <djes...@spam.comcast.net> wrote:
>>
>> >> [snip] Can
>> >> you think of any other business that operates in this fashion?
>> > Government?
>>
>> You missed the point, a company has two products, A has 20% margins and
>> sells 100 units
>> B has 10% margin and sells 500 units.
>
> Actually, I think it's more like:
>
> A - 48% margin, 1,000 units
>
> B - 10% margin, 2,000,000 units
>
We're getting a bit sloppy here, and I have checked the numbers for a
while,
but HP's margins on PC's were somewhere around 2 to 3%, and I don't
rememeber
what the numbers were for Windows servers and HP-UX servers, although I
don't
think any of them were over 10%
Bill Gunshannon
The only problem with trying to do that is that System A can't do
the things System B does.
Rich Alderson
> I know of few VMS sites that are not actively engaged in deserting VMS -
> they have no choice and they know it. It is just a matter of planning and
> execution. Most seem to be going "anywhere but HP", but the ink salesmen
> couldn't care less I guess. They certainly treat the customers that way.
Not unlike the Great 36-Bit Migration(TM) following the 1983 cancellation of
the Jupiter project (the next PDP-10 from DEC, called the "DECSYSTEM-2080" and
later the "DECSYSTEM-4050"). There were hundreds of people who decided that if
they had to move off the -10 architecture, they might as well move off Digital
hardware altogether.
Y'all have my sympathy.
--
Rich Alderson | /"\ ASCII ribbon |
ne...@alderson.users.panix.com | \ / campaign against |
"You get what anybody gets. You get a lifetime." | x HTML mail and |
--Death, of the Endless | / \ postings |
Dr. Dweeb
Actually, I usually use the term "palliative care", but I was feeling
particularly gloomy that evening. Even so, I, like you, have an entire
career built around VMS and have seen myself forced to move to areas where
jobs *actually exist* because VMS *IS* dead in the mindset of the "industry
pundits", journalists and most CIOs and about 90% of people who "used" to
work with VMS too it would seem.
Few companies are investing in new VMS systems, its all maintenance and
migration away from VMS, except for the few sites where Rdb and/or VMS is so
entrenched that the cost and risk of moving outweighs all other factors.
VMS's merits are irrelevant.
The perception is there and has been since the Palmer days, and things have
been going downhill ever since.
Btw. After 6 months with SQLServer2000 I can tell you that is is one of the
worst pieces of junk I have used for a long time. I am not convinced the
coming upgrade will improve things very much. Oh, it has lots of cool
programming bobs and knobs, but the engine is crud. But I digress.
I actually have the option of moving to VMS/Rdb (or AIX/DB2 or Oracle or
whatever) if that is the right choice. Technically it probably is given our
requirements, but I cannot bring myself to make a VMS recommendation because
the ecosystem for VMS and Rdb no longer exists and my faith in HPs
stewardship is zero - no, less than zero in fact. And the handfull of guys
in Rdb engineering while doing a great job, are at the mercy of the VMS
ecosystem.
Thera are no professional Ops people around and fewer application developers
and zero new entrants in either. The VMS owners have not spent a dime on
development tools for VMS for such a long time that no "modern" developer
takes VMS seriously, and therefore there is no interest. It is a career
dead end and this is never going to change. Sad, but true.
Beyond the entrenched clients, there is nothing but boring "keep it afloat"
and/or "figure out how we can trash this old fashioned crap" projects.
Hardly something to cause me to go to work with any enthusiasm.
If I can figure out how to view that video, I might get the board of
directors to view it - so that they will know just how inferior the
technology upon which their business relies actually is.
Cheers
Dr. Dweeb.
John Smith
> Ken Robinson wrote:
>
>> Posted with no comment....
>
> Well, did anyone think that *I* would be able to widthhold comments
> about this ?
>
> Ann Livermore:
> ##
> In particular for our OpenVMS customers, we are very focused on our
> installed base. We want our installed base to be happy, and if or when
> they ever want to migrate, we want them to migrate to another HP
> platform. ##
>
> This pretty much confirms the Stallard memo of May 7th 2002 not being
> a glitch but truly representing HP corporate policy.
>
> If that Livermore isn't an official announcement that HP intends to
> kill VMS, I don't know what it.
As good as VMS is, as much as we like it and want the market for it to be
grown, with pronouncements like that how does any reasonably sane ISV decide
that this is a market where they should invest their time/money?
And as we all know, applications sell systems.
It's time the HP apologists realize that they are playing Lucy in football
season to the customer's Charlie Brown.
Or put more bluntly, HP has an endless supply of anal lube for those
gullible enough.
--
OpenVMS - The never-advertised operating system with the dwindling ISV
base.
John Smith
AEF,
Perhaps you are confusing HP with IBM, whose motto is "THINK". HP's is
NIH.
--
OpenVMS - The never-advertised operating system with the dwindling ISV &
customer
base.
Doug Phillips
wrote:
> Hi,
>
> For those of you who, after reading this thread, may be thinking that Ann
> Livermore (I've got no idea who she is in HP and don't care) was responding
> to a question such as "What are your feelings with regard to the future
> strategic directionsof VMS?" or "How does your VMS strategy differ from your
> other OS stable?" or "Just speak randonmly on any issue for 30secs", here's
> the actual question she was reponding to: -
>
> [That expertise and quality are the issue. Last fall I spoke with two
> high-profile HP partners who are OpenVMS consultants. They were very
> concerned about services being outsourced to places like China and India,
> because they said the support technicians there simply don't have the
> expertise that's required and aren't getting the training they need. What's
> your response?]
>
> I read that like the reporter, and whoever he was talking to (VMS people "In
> particular"), accused HP of shafting the installed base by outsourcing
> support to Bangalore, Mumbai and Brisbane. So she says something as bland as
> [We value the installed base enough to want to keep them sweet so that they
> stick with us if and when they move architectures.] or [Your call is
> important to us!] and it gets translated in COV to [I'm sick of that
> grave-dodging bastard VMS! Kill it - And kill it now!] Absolute disbelief
> :-(
>
You are correct about the question; it was about support quality. Now,
tell me what Ms. Livermore's statement:
"We want our installed base to be happy, and if or when they ever want
to migrate, we want them to migrate [from OpenVMS] to another HP
platform."
has to do with the question about support? A Freudian slip, maybe?
The reporter didn't ask about migration. She didn't have to say
anything about HP wanting OpenVMS users to migrate to some other
platform. I guess she just couldn't help herself.
Now, what would do find more telling: An rehearsed answer when asked
if HP has long-term plans for OpenVMS, or a casual remark in a
response to a completely different topic?
Like-
Sam: Hi, Joe. How do you like your new car?
Joe: It's great, Sam. Plenty of room for the golf clubs in the trunk
(boot), and your wife really likes the back seat!
> But whatever you do, listen to Dr. Dweeb; VMS *is* dead! (and has been for
> years) Move on! So there can be absolutely *no* reason to keep telling
> everybody about its death for the next days, months, YEARS now can there?
>
> Cheers Richard Maher
>
[top posting not corrected]
> "Ken Robinson" <kenrb...@gmail.com> wrote in message
>
> news:7dd80f60706191324k5c0...@mail.gmail.com...> The articles are "Q&A: HP's Ann Livermore addresses company's services
> > issues"
>
> <http://www.computerworld.com/action/article.do?command=viewArticleBas...
> icleId=9025228&intsrc=hm_list>> and "Hurd: HP will show users how to run IT"
>
> <http://www.computerworld.com/action/article.do?command=viewArticleBas...
AEF
> In article <1182387477.151964.284...@c77g2000hse.googlegroups.com>,
> AEF <spamsink2...@yahoo.com> writes:
>
>
>
> > On Jun 20, 8:47 pm, David J Dachtera <djesys...@spam.comcast.net>
> > wrote:
> >> Tom Linden wrote:
>
> >> > On Tue, 19 Jun 2007 18:45:09 -0700, David J Dachtera
> >> > <djesys...@spam.comcast.net> wrote:
>
> >> > >> [snip] Can
> >> > >> you think of any other business that operates in this fashion?
> >> > > Government?
>
> >> > You missed the point, a company has two products, A has 20% margins and
> >> > sells 100 units
> >> > B has 10% margin and sells 500 units.
>
> >> Actually, I think it's more like:
>
> >> A - 48% margin, 1,000 units
>
> >> B - 10% margin, 2,000,000 units
>
> > Now ... imagine the huge profit gains from switching even a small
> > percentage of B systems to A systems!
>
> The only problem with trying to do that is that System A can't do
> the things System B does.
Oh yeah? Try defining logical names on other OSes and see what
happens. :-)
Hey, maybe some customers don't need those missing things. Maybe VMS
has some things others don't, like recovering in 13.xx seconds after a
gas explosion!
Bill, you've been vindicated about Linux. It had the worse recovery
time, but not much worse than Windows. HP-UX recovered significantly
faster.
>
> bill
>
> --
> Bill Gunshannon | de-moc-ra-cy (di mok' ra see) n. Three wolves
> b...@cs.scranton.edu | and a sheep voting on what's for dinner.
> University of Scranton |
> Scranton, Pennsylvania | #include <std.disclaimer.h>
AEF
AEF
Why do you say that? I said "imagine", I didn't say "expect".
Don't they push their Proliant series? Where's the NIH there?
AEF
AEF
She is answering a more general question that covers the specific
question, but it is a bit of "dodgery". She said she wants the
installed base to be happy. So one can conclude that that means she
implying that she would want to improve support to make them happy.
The remainder of the answer about if and when to migrate was partly
out of left field, but still implying she wants the customers to be
happy and stay with hp even if they migrate. Well, this should be
obvious. Of course HP wants to retain its customers, or at least it
makes sense to. Now while she may have other plans, she didn't SAY
anything about killing VMS. If, as some here do, always act so
cynically then what difference does it makes what she says? It will
always be interpreted by the cynical in a bad way.
> The reporter didn't ask about migration. She didn't have to say
> anything about HP wanting OpenVMS users to migrate to some other
> platform. I guess she just couldn't help herself.
She's just saying the obvious: She wants to hold on to the customers
by making them happy. Suppose a customer on its own wants to migrate.
She'll try to win that business. And why not? That doesn't make US
happy (that this hypothetical customer is migrating away from VMS),
but that's beside the point. If she said that HP is going to start
advertising VMS would the cynical act any differently? In which case
why make all the fuss over her words? If she said instead that she
would make sure to improve support would the cynical here be any
happier?
>
> Now, what would do find more telling: An rehearsed answer when asked
> if HP has long-term plans for OpenVMS, or a casual remark in a
> response to a completely different topic?
It's not completely different. She is saying that she wants to hold on
to her customers by keeping them happy and adding an example.
[...]
AEF
JF Mezei
> She's just saying the obvious: She wants to hold on to the customers
> by making them happy.
Sorry, that doesn't work in real life. She should have responded with
something like:
"We want to provide our customers with first quality service because it
not only ensures customers stay with HP but also attracts new customers."
Sorry, but Livermore's statement was an intentional message to the ISV
community. She took that opportunity to insert the message that HP does
not intent to grow the VMS marketplace and migrate customers from VMS
when she didn't need to mention this. This was intentional.
Let me ask you this: if you had your own company, and one fo your
employees spoke to the press in such a way that it made you lose
millions in business, wouldn't you quickly go to the media and correct
this error and try to reduce the damage cause by your employee ? (and
possible demote/fire/side-mote that employee.
Dr. Dweeb
> AEF wrote:
>> She's just saying the obvious: She wants to hold on to the customers
>> by making them happy.
>
> Sorry, that doesn't work in real life. She should have responded with
> something like:
>
> "We want to provide our customers with first quality service because
> it not only ensures customers stay with HP but also attracts new
> customers."
>
> Sorry, but Livermore's statement was an intentional message to the ISV
> community. She took that opportunity to insert the message that HP
> does not intent to grow the VMS marketplace and migrate customers
> from VMS when she didn't need to mention this. This was intentional.
>
I would suggest that is "the corporate line" and was inserted in order to
promote said corporate position.
The reasons for this being the corporate line are of course open for
conjecture, but here in cov, we have a pretty good handle on how it feels to
be at the reciving end of HP and can guess at the HP intent.
Dweeb
AEF
> AEF wrote:
> > She's just saying the obvious: She wants to hold on to the customers
> > by making them happy.
>
> Sorry, that doesn't work in real life. She should have responded with
> something like:
>
> "We want to provide our customers with first quality service because it
> not only ensures customers stay with HP but also attracts new customers."
If she said that you would claim she was lying. So what?
> Sorry, but Livermore's statement was an intentional message to the ISV
> community. She took that opportunity to insert the message that HP does
> not intent to grow the VMS marketplace
Yes, I'll give you that much. If HP is planning on growing the VMS
market, you wouldn't get that impression from what she said. (But how
then to explain the new video?)
> and migrate customers from VMS when she didn't need to mention this. This was intentional.
If she had said only "when", you'd be right. She said "if and when".
I'm going by what she said. I'm not saying that's HP's true
intentions. I'm only interpreting what she said at face value. If you
don't take it at face value you can spin it anyway you want, so what's
the point? I'm saying you're jumping to conclusions. That doesn't
guarantee that you're wrong about HP's intentions with regard to VMS,
but I find your argument rather weak.
Besides, you been claiming VMS faced imminent death for years as Rob
Young has pointed out months (years?) ago. What's different this time?
> Let me ask you this: if you had your own company, and one fo your
> employees spoke to the press in such a way that it made you lose
> millions in business, wouldn't you quickly go to the media and correct
> this error and try to reduce the damage cause by your employee ? (and
> possible demote/fire/side-mote that employee.
Yes, but is this really the case?
Consider this: Why is VMS so prominent in the gas-explosion video only
two clicks away from www.hp.com? Hey, they supposedly spent over $1M
on that! I assume they're expecting to make more than that in profits
as a result.
AEF
JF Mezei
> Besides, you been claiming VMS faced imminent death for years as Rob
> Young has pointed out months (years?) ago. What's different this time?
I have not claimed VMS faces imminent death. But what is clear is that
VMS is slowly being poisoned and eventually, HP will just put it out of
its misery.
You need to look at it from HP's point of view. They have no intentions
of growing VMS. Their only intentions are to move as many VMS customers
to their selected platform as possible.
If they sell VMS to Bruden or Process (or a joint venture of both), then
HP will lose a greater percentage of customers to those companies who
will also sell support than if HP lies to customers with fake roadmaps
and fully intends to convert them to HP-UX.
Messges such as Livermore's are intended to get ISVs to rething their
commitment to VMS. When an ISV like Cerner announces it is dropping VMS
support, the stain goes on Cerner and not HP so people aren't so mad at HP.
It is our role as VMS loyalist to ensure that HP doesn't get away with
this tactic by pointing out that in the case of Cerner, surely there has
been a deal by HP to urge Cerner to move to HP-UX. It was certaintly
not customer requests that pushed that. So if the blame for the loss of
Cerner lais flat on HP's face in the public eye, then HP will learn that
it cannot do this anymore.
And we need to convince HP that there is NOTHING WRONG WITH GROWING VMS.
If HP sells VMS to Bruden/Process, at least customers will not yhold a
grudge against HP and may continue to buy HP's coloured water in
expensive small containers.
If HP continues to act to quicken VMS's demise, then customers will
revold against HP and HP will stand to lose sales of other products.
WE MUST MAKE IT CLEAR TO HURD THAT VMS CUSTOMERS WILL NOT BE HP
CUSTOMERS UNLESS HP PROMOTES VMS.
Main, Kerry
> -----Original Message-----
> From: Dr. Dweeb [mailto:sp...@dweeb.net]
> Sent: June 21, 2007 4:51 PM
> To: Info...@Mvb.Saic.Com
> Subject: Re: And the question was? (Re: Amazing, two new articles on
> Computerworld.com actual mention OpenVMS!)
>
[snip...]
>
> Few companies are investing in new VMS systems, its all maintenance
> and
> migration away from VMS, except for the few sites where Rdb and/or VMS
> is so
> entrenched that the cost and risk of moving outweighs all other
> factors.
Shhh.. please don't tell Cust's with mission critical environments like
the following:
http://www.vista-control.com/itanium_success.htm
"Los Alamos, February 15th. 2007 After implementing mission-critical
systems on Windows-based computers for many years, a customer
experienced a virus in one of these systems that shut down production
for two days while the infected systems were diagnosed, restored and
tested. The impact was that plant production was severely impacted at no
small cost. Despite internal opposition because of the established
standard, Vsystem on HP Itanium servers running OpenVMS was chosen for
the next system to be replaced."
"Mission-critical applications require rock-solid 365/24 operation. The
costs of any down-time far out-weigh any initial savings of using
Windows-based computer systems let alone the additional costs of
on-going security upgrades and other security costs are considered."
Won't happen overnight, but there are a few things to consider ..
- about every 5-10 years the industry shifts from distributed to
centralized computing strategies. OpenVMS's real strengths are in
centralized computing. Right now, there is a massive push to
re-centralize servers and solutions. From a centralized perspective, ask
your new company how they plan to do app stacking or manage all of their
batch jobs so that you do not care what system the job runs on.
- Companies are getting tired of the 5-20 security patches released each
and every month for Windows and Linux. They are staring to realize that
perhaps they can not afford these platforms.
Actually, it is more like existing OpenVMS old timers who have not kept
up with recent developments like NetBeans for OpenVMS, Distributed
NetBeans, Availability Manager for OpenVMS (GUI mgmt), Java etc ..
Reference:
http://h71000.www7.hp.com/openvms/products/ips/netbeans/overview.html
"NetBeans is a modular, integrated development environment (IDE) for
Java and JavaBeans development. Written in 100 percent pure Java(tm), it
was open-sourced by Sun Microsystems. Its popularity is a result of its
versatility, extensible architecture, and relative ease of use. Key
features are:
- Support for the Java, C/C++, XML, and HTML
- Support for JSP, XML, RMI, CORBA, JINI, JDBC, and servlet technologies
- Support for Ant, CVS, and other version control systems
- Pluggable support for compilers, debuggers and execution services
- GUI form designer and other visual design tools
- Wizards for code generation and management tools
- Syntax-highlighting source editor
The goal of the NetBeans "community"-based development philosophy is to
maximize extensibility with the most generic and flexible framework
possible so that new tools can be added to a solid, stable, efficient,
and backward-compatible foundation.
Because the NetBeans tools platform is developed as an open-source
project, it has the potential to be the basis for other development
tools no matter what language or technology is being used. To this end,
NetBeans features open APIs to permit extensibility - for example,
version control support and Javadoc search and generation."
http://h71000.www7.hp.com/openvms/products/ips/netbeans/distnb.html
"Distributed NetBeans for OpenVMS allows you to run the NetBeans IDE on
your desktop system and develop applications on a remote OpenVMS Alpha
or Integrity server system."
> If I can figure out how to view that video, I might get the board of
> directors to view it - so that they will know just how inferior the
> technology upon which their business relies actually is.
>
Simple way to get their attention - If they test their important app's
before rolling out security patches (like any self respecting enterprise
would do), ask them if they can afford Windows/Linux.
> Cheers
> Dr. Dweeb.
>
Regards
Kerry Main
Senior Consultant
HP Services Canada
Voice: 613-592-4660
Fax: 613-591-4477
kerryDOTmainAThpDOTcom
(remove the DOT's and AT)
OpenVMS - the secure, multi-site OS that just works.
Doug Phillips
> > -----Original Message-----
> > From: Dr. Dweeb [mailto:s...@dweeb.net]
> > Sent: June 21, 2007 4:51 PM
> > To: Info-...@Mvb.Saic.Com
> > Subject: Re: And the question was? (Re: Amazing, two new articles on
> > Computerworld.com actual mention OpenVMS!)
>
> [snip...]
>
>
> > Few companies are investing in new VMS systems, its all maintenance
> > and
> > migration away from VMS, except for the few sites where Rdb and/or VMS
> > is so
> > entrenched that the cost and risk of moving outweighs all other
> > factors.
>
> Shhh.. please don't tell Cust's with mission critical environments like
> the following:http://www.vista-control.com/itanium_success.htm
Good news.
I haven't looked, but is this on the HP *corporate* "Success Stories"
page? The one from the "servers" link?
[snip]
> > Beyond the entrenched clients, there is nothing but boring "keep it
> > afloat"
> > and/or "figure out how we can trash this old fashioned crap" projects.
> > Hardly something to cause me to go to work with any enthusiasm.
>
> Actually, it is more like existing OpenVMS old timers who have not kept
> up with recent developments like NetBeans for OpenVMS, Distributed
> NetBeans, Availability Manager for OpenVMS (GUI mgmt), Java etc ..
>
I can't speak for all of the old-timers, but this old-timer thinks
OpenVMS is still the best OS in its class. What I feel is frustration
and disappointment because I know it could have become *the* "desk-top
to data-center" standard and had a chance to do so before M$ borged.
Okay, I'm over that now. But if VMS was a student, I'm sure "doesn't
work up to its potential" would have been on more than one report
card.
Unfortunately, the VMS owners decided to only compete in the data-
center niche, and its edge there is rapidly being closed by
advancements in the "standard" platforms.
"Recent developments" and chip road-maps point to an x86 instruction
set world. It's the applications, st***d. Always has been; always
will. Hard to consolidate on a platform that doesn't run all the
app's. Like it or not, M$ Windows is the 800 pound gorilla now, and
he isnt't going with Itanium.
Is it too late for VMS? Only HP can answer that question and their
answers aren't very reassuring.
I know there are some very good and enthusiastic HP'ers whose job it
is to do VMS and I only hope rather than lose that enthusiasm, they'll
try harder to move it up through the ranks.
If top management is in a "show me" stage with VMS, then I hope
someone showed them that video and they see/read/hear about the
testimonials. Or do most people think top management is omniscient?
Wrong! Top management is mostly *insulated*. Too often in large
companies the levels below the top add a bit too much color to the
filters in both directions.
I can imagine some PHB's watching the fail-over video and saying: "We
have to get HP-UX, Windows and Linux up faster. Work on it!"
Dr. Dweeb
>> -----Original Message-----
>> From: Dr. Dweeb [mailto:sp...@dweeb.net]
>> Sent: June 21, 2007 4:51 PM
>> To: Info...@Mvb.Saic.Com
>> Subject: Re: And the question was? (Re: Amazing, two new articles on
>> Computerworld.com actual mention OpenVMS!)
>>
>
> [snip...]
>
>
>>
>> Few companies are investing in new VMS systems, its all maintenance
>> and
>> migration away from VMS, except for the few sites where Rdb and/or
>> VMS is so
>> entrenched that the cost and risk of moving outweighs all other
>> factors.
>
> Shhh.. please don't tell Cust's with mission critical environments
> like the following:
Kerry, I just left one.
A multi-billion doillar p.a. houshhold name. They are expending a VERY
large amount of money to migrate away from HP (VMS/Rdb). Anything but HP at
any cost (and I mean *any* cost). They have absolutely sound reasons for
doing so, none of which have to do with VMS quality or functionality.
Dweeb
Dr. Dweeb
I see nothing to contradict my statement (quoted for effect).
>> If I can figure out how to view that video, I might get the board of
>> directors to view it - so that they will know just how inferior the
>> technology upon which their business relies actually is.
>>
>
> Simple way to get their attention - If they test their important app's
> before rolling out security patches (like any self respecting
> enterprise would do), ask them if they can afford Windows/Linux.
>
You need to read my previous posts on this matter. I am the last one to
disagree with you here!!!!! I will refer you in the event that you missed
my commentary.
Dweeb.
Main, Kerry
> From: Dr. Dweeb [mailto:sp...@dweeb.net]
> Sent: June 22, 2007 8:58 PM
> To: Info...@Mvb.Saic.Com
> Subject: Re: And the question was? (Re: Amazing, two new articles on
> Computerworld.com actual mention OpenVMS!)
>
> Main, Kerry wrote:
> >> -----Original Message-----
> >> From: Dr. Dweeb [mailto:sp...@dweeb.net]
> >> Sent: June 21, 2007 4:51 PM
> >> To: Info...@Mvb.Saic.Com
> >> Subject: Re: And the question was? (Re: Amazing, two new articles
> on
> >> Computerworld.com actual mention OpenVMS!)
> >>
> >
> > [snip...]
> >
> >
> >>
> >> Few companies are investing in new VMS systems, its all maintenance
> >> and
> >> migration away from VMS, except for the few sites where Rdb and/or
> >> VMS is so
> >> entrenched that the cost and risk of moving outweighs all other
> >> factors.
> >
> > Shhh.. please don't tell Cust's with mission critical environments
> > like the following:
>
>
> A multi-billion doillar p.a. houshhold name. They are expending a
> VERY
> large amount of money to migrate away from HP (VMS/Rdb). Anything but
> HP at
> any cost (and I mean *any* cost). They have absolutely sound reasons
> for
> doing so, none of which have to do with VMS quality or functionality.
>
> Dweeb
>
Well, as you stated, it sounds like this Cust has issues with HP that
extend well beyond OpenVMS.
Certainly not something HP wants (no vendor wants unhappy Cust's), and I
would hope the account team is working to correct whatever the issue is.
David J Dachtera
Unfortunately - and top HP management doesn't get this, either - the trouble
STARTS at the top.
JF Mezei
> Well, as you stated, it sounds like this Cust has issues with HP that
> extend well beyond OpenVMS.
VMS customers did not choose to be HP customers. They were forced into
it. And HP has yet to do anything to truly make us feel comfortable
within HP. It wouldn't take much, but HP refuses to do it.
> Certainly not something HP wants (no vendor wants unhappy Cust's), and I
> would hope the account team is working to correct whatever the issue is.
What percentage of the "intalled base" have an HP contact ?
Main, Kerry
> From: Dr. Dweeb [mailto:sp...@dweeb.net]
> Sent: June 22, 2007 9:03 PM
> To: Info...@Mvb.Saic.Com
> Subject: Re: And the question was? (Re: Amazing, two new articles on
> Computerworld.com actual mention OpenVMS!)
>
[snip...]
>
> "The VMS owners have not spent a dime on development tools for VMS for
> such
> a long time that no "modern" developer takes VMS seriously, and
> therefore
> there is no interest. It is a career"
>
> I see nothing to contradict my statement (quoted for effect).
>
Re: "modern developer"
Well, if modern tools means SOA type technologies, then imho, I have to
smile at this.
While there are no doubt a few areas where these technologies might be
useful, the sad reality is that App developers and tools vendors are
telling the world "move to SOA or be lost forever", but they have no
idea of how really tough this is and/or the huge culture shift in their
entire business that this requires.
Way to much hype, but "modern" App folks are off saying "we need SOA
tools, we need SOA technologies ..we need SOA staff .."
Reality check - think about the issues that DCE and common data
dictionary initiatives had. Anyone ever experienced how difficult it is
for multiple App groups to share stds, processes, and agree on data
models?
SOA is all this with even more process and culture changes required. And
all this at a time when the business is demanding massive cuts in IT
spending.
Anyway, perhaps a bit OT and not so much related to OpenVMS, but when
ever I hear this "modern developer" term, it starts my blood boiling. At
some point, someone has to tell the King (industry "modern developer")
they are wearing no clothes.
Bill Gunshannon
>> Kerry, I just left one.
>> A multi-billion doillar p.a. houshhold name. They are expending a
>> VERY
>> large amount of money to migrate away from HP (VMS/Rdb). Anything but
>> HP at
>> any cost (and I mean *any* cost). They have absolutely sound reasons
>> for
>> doing so, none of which have to do with VMS quality or functionality.
>> Dweeb
>>=20
>
> Well, as you stated, it sounds like this Cust has issues with HP that
> extend well beyond OpenVMS.
What makes you say that? We have had numerous people say that because
of HP's traetment of OpenVMS they would never buy another HP product.
Sounds pretty much the same to me.
>
> Certainly not something HP wants (no vendor wants unhappy Cust's), and I
> would hope the account team is working to correct whatever the issue is.
If it ain't PC's or ink, I don't think HP cares one way or the other.
bill
--
Bill Gunshannon | de-moc-ra-cy (di mok' ra see) n. Three wolves
bi...@cs.scranton.edu | and a sheep voting on what's for dinner.
Main, Kerry
> From: bi...@triangle.cs.uofs.edu [mailto:bi...@triangle.cs.uofs.edu] On
> Behalf Of Bill Gunshannon
> Sent: June 23, 2007 8:00 AM
> To: Info...@Mvb.Saic.Com
> Subject: Re: And the question was? (Re: Amazing, two new articles on
> Computerworld.com actual mention OpenVMS!)
>
> <FA60F2C4B72A584DBFC6...@tayexc19.americas.cpqcorp.ne
Mmm, re-read his statement "They have absolutely sound reasons for doing
so, none of which have to do with VMS quality or functionality."
Sounds like lots of non-OpenVMS issues to me.
> >
> > Certainly not something HP wants (no vendor wants unhappy Cust's),
> and I
> > would hope the account team is working to correct whatever the issue
> is.
>
> If it ain't PC's or ink, I don't think HP cares one way or the other.
>
> bill
>
Yeah, you are likely correct, who cares about all the billions of $'s
that the other depts bring in eh?
That's like saying Universities only care about Arts and Science
students (typically the biggest number by far) - following your analogy,
Universities do not care about Engineering and other Colleges / Depts.
Bill Gunshannon
>> >> Kerry, I just left one.
>> >> A multi-billion doillar p.a. houshhold name. They are expending a
>> >> VERY
>> >> large amount of money to migrate away from HP (VMS/Rdb). Anything
>> but
>> >> HP at
>> >> any cost (and I mean *any* cost). They have absolutely sound
>> reasons
>> >> for
>> >> doing so, none of which have to do with VMS quality or
>> functionality.
>> >> Dweeb
>> >>=3D20
>> >
>> > Well, as you stated, it sounds like this Cust has issues with HP
>> that
>> > extend well beyond OpenVMS.
>> What makes you say that? We have had numerous people say that because
>> of HP's traetment of OpenVMS they would never buy another HP product.
>> Sounds pretty much the same to me.
>
> Mmm, re-read his statement "They have absolutely sound reasons for doing
> so, none of which have to do with VMS quality or functionality."
What in what I said above has anything to do with "VMS quality or
functionality"?
>
> Sounds like lots of non-OpenVMS issues to me.
You don't think uncertainty about VMS's future is a VMS issue?
You don't think HP's refusal to promote VMS in order to secure
it's future is a VMS issue?
>
>> >
>> > Certainly not something HP wants (no vendor wants unhappy Cust's),
>> and I
>> > would hope the account team is working to correct whatever the issue
>> is.
>>=20
>> If it ain't PC's or ink, I don't think HP cares one way or the other.
>>=20
>> bill
>>=20
>
> Yeah, you are likely correct, who cares about all the billions of $'s
> that the other depts bring in eh?
>
> That's like saying Universities only care about Arts and Science
> students (typically the biggest number by far) - following your analogy,
> Universities do not care about Engineering and other Colleges / Depts.
Actually, your closer than you might think. They pore a lot more
mone into the School of Management here than they do the School of
Art's and Sciences. In the department that needs to work the hardest
to keep up with changes in technology, we are the one's who see our
budget eroding more and more each year.
All I was pointing out above is the same thing that everyone has
continuously pointed out. HP seems happiest pouring money into
the things with the lowest margin. Think what those billions
could look like if they involved things where the margin was
greater than 4% and most of the proifit didn't just go back out
the door to Bill Gates bank account.
Richard B. Gilbert
Do the multiplication. margin * volume
I think it will show that VMS is not where the money is.
Even if VMS ran on commodity hardware, who really needs it? Competing
O/Ss get the job done somehow or other. Linux is free. Solaris is
free. Windows is bundled with virtually every PC. Sure, there are a
few niches where VMS is clearly the O/S for the job, but not enough of them!
VMS could survive in its niche if HP gave a shit. But they don't! And
won't!
Richard Maher
> Well, if modern tools means SOA type technologies, then imho, I have to
> smile at this.
I found (and love!) this recent post in the comp.lang.java.programmer
newsgroup: -
> a funny but true Ads by IBM
>
> try {
>
> installTomcat();
> integrateAxis();
> addSecurity();
> throwInMessaging();
> } catch (TooComplicatedException e) {
And a couple of links I post here previously (How hard can it be?): -
Here's a lovely "little" Hello-World EJB example from HP
thenselves to help you make up your minds: -
http://devresource.hp.com/drc/technical_papers/ejb_jboss/
And here's one where the same people who used to discuss the criticality of
"seperating form and function" now discuss the "evils of point to point
integrations".
http://devresource.hp.com/drc/technical_white_papers/p2pSOA/p2pSOA_HP...
> While there are no doubt a few areas where these technologies might be
> useful,
Kerry, are you really denying that most IT system's requirements fit a
similar profile to: -
1) Combine my chosen News from the Athens Chronicle with the weather from
Bogata
2) Translate it all into Swahili
3) Fix foreign currency fluctuations
4) Serve it up with my CornFlakes
Surely, this is what IT managers have been crying out for :-)
There was also a guy recently in the COBOL newsgroup (from NZ post I think)
discussing the beauty of a SOAP based address validation service. I suppose
they're as likely as anyone to have a resonable list of valid addresses, but
how does your client application rate the prospective service providers?
What SLA do you have with these people so that when your application freezes
due to an NZ Post outage, or data corruption, you can sue them? What if down
the track they decide to start charging (or increase the charges) for the
service? When a street name changes do we all get a UDP broadcast telling us
to update our databases?
Is the beauty of XML not the schema and a service provider's ability to
change the metadata on the fly? So Suburb is now 50 bytes instead of 40 -
It's just magical how all our databases, screens and printouts automatically
adjust to reflect the change :-) Most sites I've seen have trouble with
controlled data redundancy in multiple databases when they own the data, but
I'm sure someone's thought all this through?
But let's talk Transactions. The fact that I believe WS-AT to be the
greatest piece of over-engineering since the Spruce Goose, doesn't mean to
won't fly! (But if VMS engineering really believed in Transactional Web
Services why would they have thrown Jim Johnson into the arms of Microsoft?)
So now to protect data integrity and preserve the ACID properties of our
distributed transactions we must lock all target rows (potentially freezing
great swathes of the database) until our service provider can tell us to
Commit or Abort? I'm guessing here that no one envisages a user's home PC
coordinating one of these transactions :-)
Better stick with Business Activity transactions and that whole
"Compensation Transaction" thingy? Atomicity's had its day; just another
legacy throw-back to be done away with.
"But SOAP is still a valid *internal* application deployment paradigm" -
Retreating Expensive Consultant
So all this is really what HP is stuffing down the throats of its
long-suffering VMS client base? After betraying them with Rally, ONC/RPC,
DCE/RPC, COM, BridgeWords, DECadmire, Forte, ACMSxp you are now forcing them
into yet another 180 degree backflip with WSIT and Axis :-( When all they
really wanted was to Webify and integrate their existing Applications and
data (or at least provide a decent GUI).
Cheers Richard Maher
"Main, Kerry" <Kerry...@hp.com> wrote in message
news:FA60F2C4B72A584DBFC6...@tayexc19.americas.cpqcorp.net...
AEF
> AEF wrote:
> > Besides, you been claiming VMS faced imminent death for years as Rob
> > Young has pointed out months (years?) ago. What's different this time?
>
> I have not claimed VMS faces imminent death. But what is clear is that
> VMS is slowly being poisoned and eventually, HP will just put it out of
> its misery.
Or just use Google Groups and search cov for "death JF" and read
message 5 (by Rob Young).
You're right! You said "impending death", not "imminent death". Sorry.
> You need to look at it from HP's point of view. They have no intentions
> of growing VMS. Their only intentions are to move as many VMS customers
> to their selected platform as possible.
While they're obviously not big on growing VMS, I don't think they're
actively trying to move customers away from it.
> If they sell VMS to Bruden or Process (or a joint venture of both), then
> HP will lose a greater percentage of customers to those companies who
> will also sell support than if HP lies to customers with fake roadmaps
> and fully intends to convert them to HP-UX.
Roadmaps are not contracts.
>
> Messges such as Livermore's are intended to get ISVs to rething their
> commitment to VMS. When an ISV like Cerner announces it is dropping VMS
> support, the stain goes on Cerner and not HP so people aren't so mad at HP.
She said what's obvious: any company wants to keep its customers and
that means keeping them happy enough not to jump ship.
>
> It is our role as VMS loyalist to ensure that HP doesn't get away with
> this tactic by pointing out that in the case of Cerner, surely there has
> been a deal by HP to urge Cerner to move to HP-UX. It was certaintly
> not customer requests that pushed that. So if the blame for the loss of
> Cerner lais flat on HP's face in the public eye, then HP will learn that
> it cannot do this anymore.
I don't know enough about this deal to comment. But again you appear
to be putting the worst possible spin on it.
>
> And we need to convince HP that there is NOTHING WRONG WITH GROWING VMS.
Agreed. I wrote my letter to Hurd and now we have the disaster-proof
video. Hah! ;-D
>
> If HP sells VMS to Bruden/Process, at least customers will not yhold a
> grudge against HP and may continue to buy HP's coloured water in
> expensive small containers.
>
> If HP continues to act to quicken VMS's demise, then customers will
> revold against HP and HP will stand to lose sales of other products.
>
> WE MUST MAKE IT CLEAR TO HURD THAT VMS CUSTOMERS WILL NOT BE HP
> CUSTOMERS UNLESS HP PROMOTES VMS.
I already wrote him and told him that with other OSes customers can
always go somewhere else.
AEF
Peter 'EPLAN' LANGSTOeGER
>She said what's obvious: any company wants to keep its customers and
>that means keeping them happy enough not to jump ship.
I expect more actions and more positive words if HP intends to keep its
(VMS) customers (because a lot ARE still forced to jumping ship)
Sigh. History is repeating itself over and over (and yet we still hope)...
--
Peter "EPLAN" LANGSTOEGER
Network and OpenVMS system specialist
E-mail pe...@langstoeger.at
A-1030 VIENNA AUSTRIA I'm not a pessimist, I'm a realist
JF Mezei
> While they're obviously not big on growing VMS, I don't think they're
> actively trying to move customers away from it.
How would you qualify the deal made with Cerner to stop developping for
VMS and instead target HP-UX ?
> Roadmaps are not contracts.
Exactly. While it is very nice for VMS to have a roadmap, it isn't
something you can bet your business on.
> She said what's obvious: any company wants to keep its customers and
> that means keeping them happy enough not to jump ship.
No. She added the mention that if/when VMS customers want to migrate, HP
wants to keep them as customers. This is like Microsoft stating
publically that it wouldn't mind Windows customers migrating to MACOS if
they continued to buy Office versions.
She could have just said that HP wants VMS customers to be happy and
remain with HP and keep it at that. She went out of her way to mention
the "migrating from VMS" bit. That was uncalled for.
> I don't know enough about this deal to comment. But again you appear
> to be putting the worst possible spin on it.
OK, I will put a good spin : HP has kept some developpers still working
on VMS and VMS is still being devevopped.
I wonder how soon VMS will ship with indian fonts built-in ??? :-)
AEF
wrote:
> In article <1182689234.575093.170...@o61g2000hsh.googlegroups.com>, AEF <spamsink2...@yahoo.com> writes:
>
> >She said what's obvious: any company wants to keep its customers and
> >that means keeping them happy enough not to jump ship.
>
> I expect more actions and more positive words if HP intends to keep its
> (VMS) customers (because a lot ARE still forced to jumping ship)
> Sigh. History is repeating itself over and over (and yet we still hope)...
Well, I'm sure hp would prefer to keep these customers. And I'm not
the least bit surprised that an hp exec like Ann would say things like
she said. But yes, it may be that HP either isn't willing to do more
or thinks things are fine the way they are. My primary point was about
what Ann Livermore SAID because that is what JF based his maximized
FUD upon. And I was saying that his conclusion does not follow from
her comments. It might be true for other reasons, but not because of
her comments.
But!: I think with the new video HP isn't going to pull the plug on
VMS anytime soon. Hopefully they'll come up with more positive things
about VMS as we saw in the disaster-proof video. And now many people
will see VMS in a very positive light for a change. We should be
happy! At least for now.
We should say, "That was great! Encore! More! More!", not wallow in
our sorrows. I believe that will be a lot more helpful than "Well,
this was wrong and that was wrong and this wasn't enough ... " Fooey!
Let's celebrate the video and encourage more such things from HP.
Remember the video: OpenVMS mentioned first. OpenVMS came IN first.
Hurray for OpenVMS!
Thank you HP. Thank you Mr. Hurd.
Re the fish: Fish recovery time: 294 seconds.
Hmmmm. Instead of catapulting the fish, how about a flying data center
recovery?! ... OK, that's too silly. (It's getting late!)
AEF
>
> --
> Peter "EPLAN" LANGSTOEGER
> Network and OpenVMS system specialist
> E-mail p...@langstoeger.at
Tom Linden
<jfmezei...@vaxination.ca> wrote:
> AEF wrote:
>> While they're obviously not big on growing VMS, I don't think they're
>> actively trying to move customers away from it.
>
> How would you qualify the deal made with Cerner to stop developping for
> VMS and instead target HP-UX ?
>
>> Roadmaps are not contracts.
>
> Exactly. While it is very nice for VMS to have a roadmap, it isn't
> something you can bet your business on.
Roadmaps are trial balloons written by people trying to define rather than
understand the market. It is a form of polling.
>
>> She said what's obvious: any company wants to keep its customers and
>> that means keeping them happy enough not to jump ship.
>
> No. She added the mention that if/when VMS customers want to migrate, HP
> wants to keep them as customers. This is like Microsoft stating
> publically that it wouldn't mind Windows customers migrating to MACOS if
> they continued to buy Office versions.
>
> She could have just said that HP wants VMS customers to be happy and
> remain with HP and keep it at that. She went out of her way to mention
> the "migrating from VMS" bit. That was uncalled for.
>
>> I don't know enough about this deal to comment. But again you appear
>> to be putting the worst possible spin on it.
>
> OK, I will put a good spin : HP has kept some developpers still working
> on VMS and VMS is still being devevopped.
>
> I wonder how soon VMS will ship with indian fonts built-in ??? :-)
--
PL/I for OpenVMS
www.kednos.com
John Smith
No sane organization will buy VMS (the development team et al.) unless they
had unshakeable confidence that Itanic won't be sunk by the 100th
anniversary of its namesake's demise. HP couldn't really care about which
processor PH-UX runs on - x86-64 ought to be fine for it; Tandem - well
those companies that use it are mostly price insensitive and are used to
being well lubed, so a modified x86-64 for its use could probably be
produced for a lot less than Intel sinking more money into Itanic. VMS is
the bastard stepchild of the bunch.
Let's assume that 100% of the remaining customer base and their 0.6 Gorham's
of remaining systems (W.A.G.) made the decision today to all migrate from
VMS asap. Figure five years for the most lengthy migration, but the bulk
would be completed in under 3 years - whether via porting to unix with some
of the available toolkits, or application replacement (ie. moving to a
unix/windows supported package), or via a wholesale re-write. Anyone
wanting to buy the VMS business from HP would base their bid on the
declining balance of support revenues over a 3-year horizon, discounted at
60%+ to accomodate the risk of faster migrations. As to 'new' sales
revenues - forget it.
--
OpenVMS - The never-advertised operating system with the dwindling ISV
base.
Bill Todd
...
HP couldn't really care about which
> processor PH-UX runs on - x86-64 ought to be fine for it; Tandem - well
> those companies that use it are mostly price insensitive and are used to
> being well lubed, so a modified x86-64 for its use could probably be
> produced for a lot less than Intel sinking more money into Itanic.
Both prospects are at least somewhat hampered by the fact that both
those environments are big-endian (and are old enough that some
significant amount of code - including third-party code which may not be
amenable to being massaged - may well depend upon this), while x86-64 is
little-endian.
(VMS, of course, is also little-endian, so porting it to x86-64 would
not encounter this particular obstacle.)
- bill
Dr. Dweeb
I am always amused by this "3 year" idea. Applications with 50 man years of
developement and bugs and lost code and documentation are not particularly
amenable to immediate conversion - and there are PLENTY of these.
Also, any mission critical system on Rdb that needs to be moved will likely
take "quite a while". This is not just a software issue - it is a manpower
and technical skillset for the development team and the operations team, and
then there is the little matter of performance and integrity - non-trivial
things. There are examples where the "5 year plan" was still such after 5
years. There are also examples where "no expense spared" projects went
tits-up because it turned out to be "just too hard".
HP know this and have calculated what it will mean.
OVMS will be around on palliative care for a long time to come. What won't
be around is a supply of experienced (or new) talented developers, data
specialists, operations people etc. with the skills. Those with the skills
will find themselves the buggywhip manufacturers of the IT world.
Dweeb
Tom Linden
wrote:
> John Smith wrote:
>
> ...
>
> HP couldn't really care about which
>> processor PH-UX runs on - x86-64 ought to be fine for it; Tandem - well
>> those companies that use it are mostly price insensitive and are used to
>> being well lubed, so a modified x86-64 for its use could probably be
>> produced for a lot less than Intel sinking more money into Itanic.
>
> Both prospects are at least somewhat hampered by the fact that both
> those environments are big-endian (and are old enough that some
> significant amount of code - including third-party code which may not be
> amenable to being massaged - may well depend upon this), while x86-64 is
> little-endian.
That is not a show-stopper. The 8086 had the byte swap instruction and
with some
modification it could certainly fully support both modes, as does power.
>
> (VMS, of course, is also little-endian, so porting it to x86-64 would
> not encounter this particular obstacle.)
>
> - bill
--
PL/I for OpenVMS
www.kednos.com
Michael Kraemer
> That is not a show-stopper. The 8086 had the byte swap instruction and
> with some
> modification it could certainly fully support both modes, as does power.
but this is only useful if you know that all your data is of
the same type, e.g. 32bit integers.
If the data stream comprises a mixture of ASCII text, doubles and ints
the reader software has to have an idea how the data is structured.
Michael Kraemer
> Both prospects are at least somewhat hampered by the fact that both
> those environments are big-endian (and are old enough that some
> significant amount of code - including third-party code which may not be
> amenable to being massaged - may well depend upon this), while x86-64 is
> little-endian.
IMHO these endianness issues shouldn't be such a problem.
Any half-decent app developped in the past 10, 15 or maybe even 20 years,
which stores its data in binary format has sort of a byte-order
indicator in its output data, so the reader could swap the bytes
accordingly.
And those who haven't probably deserve to die anyway.
AEF
>
> No sane organization will buy VMS (the development team et al.) unless they
> had unshakeable confidence that Itanic won't be sunk by the 100th
> anniversary of its namesake's demise. HP couldn't really care about which
> processor PH-UX runs on - x86-64 ought to be fine for it; Tandem - well
> those companies that use it are mostly price insensitive and are used to
> being well lubed, so a modified x86-64 for its use could probably be
> produced for a lot less than Intel sinking more money into Itanic. VMS is
> the bastard stepchild of the bunch.
>
> Let's assume that 100% of the remaining customer base and their 0.6 Gorham's
> of remaining systems (W.A.G.) made the decision today to all migrate from
> VMS asap. Figure five years for the most lengthy migration, but the bulk
> would be completed in under 3 years - whether via porting to unix with some
> of the available toolkits, or application replacement (ie. moving to a
> unix/windows supported package), or via a wholesale re-write. Anyone
> wanting to buy the VMS business from HP would base their bid on the
> declining balance of support revenues over a 3-year horizon, discounted at
> 60%+ to accomodate the risk of faster migrations. As to 'new' sales
> revenues - forget it.
>From personal experience I have found that these types of estimates
are missing a factor of 2.5. So make that 12.5, 7.5, and 7.5 years
respectively!
>
> --
> OpenVMS - The never-advertised operating system with the dwindling ISV
> base.
AEF
Stephen Hoffman
> Re: "modern developer"
>
> Well, if modern tools means SOA type technologies, then imho, I have to
> smile at this.
>
> While there are no doubt a few areas where these technologies might be
> useful, the sad reality is that App developers and tools vendors are
> telling the world "move to SOA or be lost forever", but they have no
> idea of how really tough this is and/or the huge culture shift in their
> entire business that this requires.
There are some seriously cool tools in this area, powerful and
effective tools and techniques. Surprisingly little code, for what you
get, too. Stuff I've been using, and that have made me more efficient
and effective than I was with older techniques and tools. Enhancements
and improvements are the nature of the business, after all.
> Way to much hype, but "modern" App folks are off saying "we need SOA
> tools, we need SOA technologies ..we need SOA staff .."
There's always been hype and FUD. Key here is whether a
client-server, three-tier client-server, middleware, SOA or whatever you
choose to call the particular design or tool or technique moves the
business forward.
Dismissing these tools out of hand is as hazardous as blindly leaping
into the unknown. (My grandfather used a wooden hammer and nails. I
use a nail gun. Had my grandfather had the opportunity, he would have
purchased and used the first nail gun he could have afforded. I still
have a hammer around, and it gets some use. That older compiler I have
around, too, still gets some use.)
> Reality check - think about the issues that DCE and common data
> dictionary initiatives had. Anyone ever experienced how difficult it is
> for multiple App groups to share stds, processes, and agree on data
> models?
If you're working in an organization that's fairly rigid, that can
certainly be the case.
I've been targeting modular designs for components for a while, and
object oriented (OO) design -- for all its faults -- does tend to force
encapsulation and modular programming. While it's possible to implement
modular programming in most languages, OO can be a particular advantage
in specific development environments; it tends to force a degree of
discipline.
The first large-scale project I worked on that used modular
components was back around 1987 or so, and the encapsulated UI and
internal communications and the networking most definitely expedited
development and deployment. I was then writing what would now be known
as middleware, or SOA. And the effort involved here gotten both easier
and -- due to the ever-increased scale of the projects -- harder since
then.
Had I to do it over again, I'd be using XML and such.
> SOA is all this with even more process and culture changes required. And
> all this at a time when the business is demanding massive cuts in IT
> spending.
Pressure to increase output and to decrease costs is nothing new.
The same pressures also apply to tasks including purchasing servers,
purchasing companies, or to most any other business expenses. (And then
there's the furor over whether IT is a competitive advantage -- and in
various cases, it is not. Nicholas Carr ignited a maelstrom on this
topic a while back.)
> Anyway, perhaps a bit OT and not so much related to OpenVMS, but when
> ever I hear this "modern developer" term, it starts my blood boiling. At
> some point, someone has to tell the King (industry "modern developer")
> they are wearing no clothes.
Watch out: some few of those developers are wearing some very serious
and very modern armor. Some of what I've been working with recently is
a whole lot less coding effort for a whole lot more output; it's far
more efficient. And far more competitive.
--
www.HoffmanLabs.com
Services for OpenVMS
Bill Todd
> On Tue, 26 Jun 2007 11:38:51 -0700, Bill Todd <bill...@metrocast.net>
> wrote:
>
>> John Smith wrote:
>>
>> ...
>>
>> HP couldn't really care about which
>>> processor PH-UX runs on - x86-64 ought to be fine for it; Tandem - well
>>> those companies that use it are mostly price insensitive and are used to
>>> being well lubed, so a modified x86-64 for its use could probably be
>>> produced for a lot less than Intel sinking more money into Itanic.
>>
>> Both prospects are at least somewhat hampered by the fact that both
>> those environments are big-endian (and are old enough that some
>> significant amount of code - including third-party code which may not
>> be amenable to being massaged - may well depend upon this), while
>> x86-64 is little-endian.
>
> That is not a show-stopper.
Yes, it often is, at least for environments affected by the problem -
and for considerably more reasons than Michael touched upon.
> The 8086 had the byte swap instruction
If you honestly believe that that makes a scintilla of difference, the
depth of your ignorance in this area is rather surprising.
and
> with some
> modification it could certainly fully support both modes, as does power.
"With some modification" being, of course, the operable constraint: do
you have any reason whatsoever to believe that any such modification to
x86 is in the cards? It's complex enough for a RISC architecture, and
for an ISA as convoluted as x86, adding it in at all - let alone without
compromising performance - would likely be a significant challenge.
- bill
Bill Todd
> Bill Todd schrieb:
>
>> Both prospects are at least somewhat hampered by the fact that both
>> those environments are big-endian (and are old enough that some
>> significant amount of code - including third-party code which may not
>> be amenable to being massaged - may well depend upon this), while
>> x86-64 is little-endian.
>
> IMHO these endianness issues shouldn't be such a problem.
Easy for you to say. Simply verifying that no problem exists could
require more investigation than, e.g., Y2K issues did.
> Any half-decent app developped in the past 10, 15 or maybe even 20 years,
> which stores its data in binary format has sort of a byte-order
> indicator in its output data, so the reader could swap the bytes
> accordingly.
Even assuming such an app (and one should not forget that both
environments in question are over 20 years old), unless it and any
associated legacy libraries were *coded* to be endian-independent just
knowing the byte order of the data wouldn't help much: at a minimum,
you'd need a subsidiary application to rearrange the data, and if there
were any endian-sensitive code in the application or the libraries that
still wouldn't suffice.
> And those who haven't probably deserve to die anyway.
Again, easy for you to say. The problem, of course, is that it's not
just the application that dies, it's the customer environment that
depends upon it - an environment that may not have the resources (or
even the capability independent of required resources) to work around
the problem.
There's a reason that IBM mainframes still support 360-era applications,
you know.
- bill
Tom Linden
wrote:
Michael, you are not thinking like PL/I programmer:-) It should be
transparent.
A structure is a structure is a structure. I/O statements know about them
JF Mezei
>> IMHO these endianness issues shouldn't be such a problem.
>
> Easy for you to say. Simply verifying that no problem exists could
> require more investigation than, e.g., Y2K issues did.
Consider Apple. It is using a kernel that runs on both endiannesses. And
it was able to rapidly deploy its apps onto the new target with
different endianness AND provide a PowerPc emulator on the 8086 that not
only interpret the PowerPC instructions on the fly, but also handles
endianness expectations of the original image.
Where there is a will, there is a way.
Now, whether HP would want to convert its HP-UX kernel or provide a
HP-UX compatibility layer on Linux so that HP apps would compile cleanly
is a different question.
HP also has the rights to True64 which might be portable to the 8086 in
the right endianness which would require less effort than porting all of
HP-UX.
Michael Kraemer
> Michael, you are not thinking like PL/I programmer:-) It should be
> transparent.
I know well that one can map mixed data to PL/I structures.
But so can C/C++:
fwrite( struct, sizeof(struct), 1, FILE * );
not that much different from
WRITE FILE(file) FROM(struct);
> A structure is a structure is a structure. I/O statements know about them
Unless one runs into the issue whether an integer is
16, 32 or 64 bit long on the platform which wrote the data,
but of different size on another platform which tries to read it.
And I wasn't aware that modern PL/I could do appropriate byte swap
on structure members automagically.
Michael Kraemer
> Again, easy for you to say. The problem, of course, is that it's not
> just the application that dies, it's the customer environment that
> depends upon it - an environment that may not have the resources (or
> even the capability independent of required resources) to work around
> the problem.
Maybe, but moving to a different-endian archictecture usually
means binary incompatibility anyway, so unless one has a very good
emulator one would have to rebuild the app + environment anyway.
For that, one needs source text + docs, and if these aren't available
or are too difficult to rebuild on a target platform, you're had.
Even in my little world I've seen such migration issues, and usually
the answer is to drop the old stuff and look for a new app with
at least similar functionality. Often this might even be a win at the
bottom line because requirements change over a 20 years and are
most probably better matched by new apps than revamped old ones.
> There's a reason that IBM mainframes still support 360-era applications,
> you know.
Yes, surely a lot of customers appreciate that,
but this was only possible because there was no need
to dump/replace the /360/370/390 architecture. IBM always managed
to revamp it in a backward compatible manner,
and still stayed profitable.
Michael Kraemer
> Michael, you are not thinking like PL/I programmer:-) It should be
> transparent.
> A structure is a structure is a structure. I/O statements know about them
plus, I forgot, of course there's also the issue of padding to
account for "odd" structure members.
May/will also depend on architecture, endianness, compiler.
So one can't assume that PL/I will automagically solve
the problems when migrating to a binary incompatible hardware.
Tom Linden
wrote:
> Tom Linden schrieb:
It can't unless you use a builtin like REVERSE, although it would be easy
to add
as an option to the ENVIRONMENT attribute. I misunderstood, I didn't
reazlize
you were talking about a mixed environment
Tom Linden
wrote:
> Tom Linden schrieb:
The compiler pads if you specifiy ALIGNED on the structure declaration,
otherwise not.
Bob Koehler
> The compiler pads if you specifiy ALIGNED on the structure declaration,
> otherwise not.
That may be true for your PL/I compilers, but it's not true in
general.
I worked with an HP C compiler under HP-UX where alignment depended
on surounding data, so even putting a structure in an .h file and
including the .h file in two different programs resulted in two
differen pads. That compiler had no alignment/padding control
directives or compiler options.
Bill Gunshannon
koe...@eisner.nospam.encompasserve.org (Bob Koehler) writes:
> In article <op.tuk1i...@murphus.linden>, "Tom Linden" <tom-r...@kednos.com> writes:
>> The compiler pads if you specifiy ALIGNED on the structure declaration,
>> otherwise not.
>
> That may be true for your PL/I compilers, but it's not true in
> general.
True, as things like C and Pascal don't have any language concept of
"alignment".
>
> I worked with an HP C compiler under HP-UX where alignment depended
> on surounding data, so even putting a structure in an .h file and
> including the .h file in two different programs resulted in two
> differen pads. That compiler had no alignment/padding control
> directives or compiler options.
I believe the "ALIGNED" he is using above is a languagte delclaration,
not a compiler option. But I could be wrong, I haven't done any PL/I
in over 20 years.
bill
--
Bill Gunshannon | de-moc-ra-cy (di mok' ra see) n. Three wolves
bi...@cs.scranton.edu | and a sheep voting on what's for dinner.
University of Scranton |
Scranton, Pennsylvania | #include <std.disclaimer.h>
Larry Kilgallen
> In article <6LXlI8...@eisner.encompasserve.org>,
> koe...@eisner.nospam.encompasserve.org (Bob Koehler) writes:
>> In article <op.tuk1i...@murphus.linden>, "Tom Linden" <tom-r...@kednos.com> writes:
>>> The compiler pads if you specifiy ALIGNED on the structure declaration,
>>> otherwise not.
>>
>> That may be true for your PL/I compilers, but it's not true in
>> general.
>
> True, as things like C and Pascal don't have any language concept of
> "alignment".
VMS implementations of C and Pascal certainly have that concept.
Bill Gunshannon
Within the language or as compiler options? I am not aware of
antything in the standards for either language that allowed for
alignment and if it is not in the standards (real or de facto)
then it is n't the language and it certainly doesn't increase
the liklihood that the program could be easily ported. How
would you create a C struct containing single character elements
and declare them to be "aligned"?
Oh yeah, MACRO-11 had it. :-)
Michael Kraemer
> VMS implementations of C and Pascal certainly have that concept.
Various C compiler's outside VMS have flags to control alignment within
structures, but this is an all-nothing-switch and not part of the language.
PL/I OTOH has the ALIGNED attribute as part of the language
which can be sticked to individual structure members.
Doug Phillips
> Bill Todd schrieb:
>
> > Again, easy for you to say. The problem, of course, is that it's not
> > just the application that dies, it's the customer environment that
> > depends upon it - an environment that may not have the resources (or
> > even the capability independent of required resources) to work around
> > the problem.
>
> Maybe, but moving to a different-endian archictecture usually
> means binary incompatibility anyway, so unless one has a very good
> emulator one would have to rebuild the app + environment anyway.
> For that, one needs source text + docs, and if these aren't available
> or are too difficult to rebuild on a target platform, you're had.
Why, to think that any computer company would expect their customers
to convert and rebuild all of their software, or buy all new app's
just to upgrade to newer hardware is ridiculous. I mean, what company
in their right mind would do that? Twice?
> Even in my little world I've seen such migration issues, and usually
> the answer is to drop the old stuff and look for a new app with
> at least similar functionality. Often this might even be a win at the
> bottom line because requirements change over a 20 years and are
> most probably better matched by new apps than revamped old ones.
>
And, if you have to buy all new app's and all new hardware, you are no
longer locked in to the vendor who caused you all this grief. Makes
sense to me.
> > There's a reason that IBM mainframes still support 360-era applications,
> > you know.
>
> Yes, surely a lot of customers appreciate that,
> but this was only possible because there was no need
> to dump/replace the /360/370/390 architecture. IBM always managed
> to revamp it in a backward compatible manner,
> and still stayed profitable.
There are many reasons why IBM is profitable, and thinking about how
best to hold on to their customers is one.
Michael Kraemer
> Why, to think that any computer company would expect their customers
> to convert and rebuild all of their software, or buy all new app's
> just to upgrade to newer hardware is ridiculous. I mean, what company
> in their right mind would do that? Twice?
I don't get it what you want tell me here.
Such things have happened in the past, more than once.
HP: 68k => PA => itanic => x86 ?
SGI: Mips => itanic => x86
DEC: Vax => alpha => itanic => ?
The only customers who got away w/o a major migration
in the past 15..20 years are those of IBM and Sun, IIRC.
> And, if you have to buy all new app's and all new hardware, you are no
> longer locked in to the vendor who caused you all this grief. Makes
> sense to me.
Who says one has to throw away all apps in such cases ?
If they are reasonably recent one may find versions on
the new platform as well.
Doug Phillips
> Doug Phillips schrieb:
>
> > Why, to think that any computer company would expect their customers
> > to convert and rebuild all of their software, or buy all new app's
> > just to upgrade to newer hardware is ridiculous. I mean, what company
> > in their right mind would do that? Twice?
>
> I don't get it what you want tell me here.
Sorry. I thought the sarcasm was obvious. Please place: <sarcasm> and
</sarcasm> around my statement.
> Such things have happened in the past, more than once.
> HP: 68k => PA => itanic => x86 ?
> SGI: Mips => itanic => x86
> DEC: Vax => alpha => itanic => ?
And, how has that worked out for those companies? That was kind of my
admittedly obscure point.
How is DEC and SGI doing now? We know where DEC is. SGI is losing
money but they say: "our business is still in the transition phase."
O.k., we'll wait-see.
Anyway, migrating because your platform has been killed is not the
same as choosing to migrate to another platform.
> The only customers who got away w/o a major migration
> in the past 15..20 years are those of IBM and Sun, IIRC.
>
And, Microsoft/Intel, of course. You could argue that Wintel cust's
had to "migrate" their software, I guess, but the argument is pretty
weak compared to any of those examples. The only comparable migration
woes I've seen in the Windows world were for those WinNT customers
that believed in Alpha and had to move back to Intel.
I'd say, Sun is moving towards x86, but is letting their customers
choose the pace.
> > And, if you have to buy all new app's and all new hardware, you are no
> > longer locked in to the vendor who caused you all this grief. Makes
> > sense to me.
>
> Who says one has to throw away all apps in such cases ?
> If they are reasonably recent one may find versions on
> the new platform as well.
If your current software won't run, and you have to buy new, then you
*are* throwing out the old; even if the application is the same.
Normally, this involves spending money at least for new licenses, but
probably for the new software, too. If you're lucky you might get a
discount to buy the same stuff that you've already bought once, or buy
it at the "upgrade" price.
Maybe your experience is different, but changing hardware platforms
has never fallen under any of the software update services I've ever
subscribed to; just moving up in performance on the same platform is
usually covered (of course there might be additional processor license
costs).
My point, if I haven't made it clear yet, is that forcing customers to
migrate between incompatible platforms hasn't proven to be the best
strategy.
Seems it's likely to lose a company as many or more customers than it
gains; and customers seem to lose trust in such a company.
Richard Brodie
"Michael Kraemer" <M.Kr...@gsi.de> wrote in message
news:f5ujc1$tfa$02$1...@news.t-online.com...
> The only customers who got away w/o a major migration
> in the past 15..20 years are those of IBM and Sun, IIRC.
Motorola->Sparc, SunOS->Solaris.
Scratch Sun from the list.
Bob Koehler
>
> Within the language or as compiler options?
For C, both. The compiler option is /member_alignment and the
language option is #pragma member_alignment .
I haven't used Pascal enough to know, but since every other compiler
I've used on VMS allows me to control alignment either through
compiler options or language statements I'd be real suprised if the
Pascal compiler doesn't do both.
Bob Koehler
>
> Why, to think that any computer company would expect their customers
> to convert and rebuild all of their software, or buy all new app's
> just to upgrade to newer hardware is ridiculous. I mean, what company
> in their right mind would do that? Twice?
Right, most of them expect their customers to throw out the baby
with the bath water and invest in all new applications.
In today's IT environment you buy COTS apps and put them on the
platform. Platform change time includes buying a new copy of the
same COTS app, or a replacement.
I'm glad I don't work in that environment, but I think it has a much
bigger piece of the market. I'll be happy as long as I can count on
VMS to do what I need, no matter the size of its market share.
Michael Kraemer
koe...@eisner.nospam.encompasserve.org (Bob Koehler) writes:
> In article <5efqh9F...@mid.individual.net>, bi...@cs.uofs.edu (Bill
> Gunshannon) writes:
> >
> > Within the language or as compiler options?
>
> For C, both. The compiler option is /member_alignment and the
> language option is #pragma member_alignment .
>
#pragma isn't part of the language, it depends on the compiler,
so it's just another way to apply a compiler flag.
If it would be a standardized attribute like "volatile" or "static",
that would be "part of the language".
Bill Gunshannon
But my point was that this is not the language. This is a proprietary
extension to a language. In PL/I I am fairly certain it was a part of
the language specifically. Right, Tom?
John Reagan
> In article <vBl8T4...@eisner.encompasserve.org>,
> koe...@eisner.nospam.encompasserve.org (Bob Koehler) writes:
>
>>In article <5efqh9F...@mid.individual.net>, bi...@cs.uofs.edu (Bill Gunshannon) writes:
>>
>>>Within the language or as compiler options?
>>
>> For C, both. The compiler option is /member_alignment and the
>> language option is #pragma member_alignment .
>>
>> I haven't used Pascal enough to know, but since every other compiler
>> I've used on VMS allows me to control alignment either through
>> compiler options or language statements I'd be real suprised if the
>> Pascal compiler doesn't do both.
>
>
> But my point was that this is not the language. This is a proprietary
> extension to a language. In PL/I I am fairly certain it was a part of
> the language specifically. Right, Tom?
>
> bill
>
>
For C, the standard for struct just says that fields have addresses that
increase in the order in which they are declared. The standard also
says that there may be unnamed padding within (including the end) of a
structure object,just no padding at the beginning.
For Pascal, the standard doesn't try to specify a physical form at all.
It is just says that the various fields have a list of known states
(ie, all integer values plus the undefined state) and that the fields
have one and only one state associated with them. A compiler can
allocate them in any order it felt like or even not bother allocating
fields that are unused. OpenVMS Pascal has several language extensions
to control the alignment, position, and allocation of fields in a record.
Beyond the various language standards, the OpenVMS Calling Standard does
require additional record layout behavior of compilers.
--
John Reagan
OpenVMS Pascal/Macro-32/COBOL Project Leader
Hewlett-Packard Company
Tom Linden
wrote:
> In article <vBl8T4...@eisner.encompasserve.org>,
> koe...@eisner.nospam.encompasserve.org (Bob Koehler) writes:
>> In article <5efqh9F...@mid.individual.net>, bi...@cs.uofs.edu (Bill
>> Gunshannon) writes:
>>>
>>> Within the language or as compiler options?
>>
>> For C, both. The compiler option is /member_alignment and the
>> language option is #pragma member_alignment .
>>
>> I haven't used Pascal enough to know, but since every other compiler
>> I've used on VMS allows me to control alignment either through
>> compiler options or language statements I'd be real suprised if the
>> Pascal compiler doesn't do both.
>
> But my point was that this is not the language. This is a proprietary
> extension to a language. In PL/I I am fairly certain it was a part of
> the language specifically. Right, Tom?
Correct.
>
> bill
Bill Gunshannon
Thank you. Nice to know I haven't lost too many of the little grey cells
just yet.
Bob Koehler
>
> #pragma isn't part of the language,
#pragma is a defined prepocessor keyword in the ANSI C language
standard.
Bob Koehler
> In article <vBl8T4...@eisner.encompasserve.org>,
> koe...@eisner.nospam.encompasserve.org (Bob Koehler) writes:
>> In article <5efqh9F...@mid.individual.net>, bi...@cs.uofs.edu (Bill Gunshannon) writes:
>>>
>>> Within the language or as compiler options?
>>
>> For C, both. The compiler option is /member_alignment and the
>> language option is #pragma member_alignment .
>>
>> I haven't used Pascal enough to know, but since every other compiler
>> I've used on VMS allows me to control alignment either through
>> compiler options or language statements I'd be real suprised if the
>> Pascal compiler doesn't do both.
>
> But my point was that this is not the language. This is a proprietary
> extension to a language. In PL/I I am fairly certain it was a part of
> the language specifically. Right, Tom?
The provision for #pragma is a part of the language standard.
Bill Gunshannon
koe...@eisner.nospam.encompasserve.org (Bob Koehler) writes:
> In article <5ehv74F...@mid.individual.net>, bi...@cs.uofs.edu (Bill Gunshannon) writes:
>> In article <vBl8T4...@eisner.encompasserve.org>,
>> koe...@eisner.nospam.encompasserve.org (Bob Koehler) writes:
>>> In article <5efqh9F...@mid.individual.net>, bi...@cs.uofs.edu (Bill Gunshannon) writes:
>>>>
>>>> Within the language or as compiler options?
>>>
>>> For C, both. The compiler option is /member_alignment and the
>>> language option is #pragma member_alignment .
>>>
>>> I haven't used Pascal enough to know, but since every other compiler
>>> I've used on VMS allows me to control alignment either through
>>> compiler options or language statements I'd be real suprised if the
>>> Pascal compiler doesn't do both.
>>
>> But my point was that this is not the language. This is a proprietary
>> extension to a language. In PL/I I am fairly certain it was a part of
>> the language specifically. Right, Tom?
>
> The provision for #pragma is a part of the language standard.
So is #define but what comes next is totally up to the user and not
part of the language spec. You can't reasonably expect any #pragma
to exist in any other compiler so it anything that follows it is a
proprietary extension. The original point was that the PLI/I language
took alignment into consideration and included alignment instructions
as a part of the language which means it is reasonable to expect a
program using it to be somewhat portable. C and Pascal do not share
this feature. Any concern for alignment is little more than an after-
thought.
John Reagan
> proprietary extension. The original point was that the PLI/I language
> took alignment into consideration and included alignment instructions
> as a part of the language which means it is reasonable to expect a
> program using it to be somewhat portable. C and Pascal do not share
> this feature. Any concern for alignment is little more than an after-
> thought.
>
Well, so does COBOL (see the SYNCRONIZED clause).
As a member of a standards committee, it is hard for a language to
predict behavior & performance of future hardware. COBOL and PL/I
provide some sort of "using my crystal ball, I think I need some
alignment help right here". What if my crystal ball is broken and I
actually needed alignment help somewhere else?
Allowing/requiring the human to "help" the compiler produce code on some
current hardware (much less future hardware) seems silly to me. Kinda
like the "register" declaration in C. Not really part of any language
design, but just a crutch for a brain-dead compiler.
Computer languages allow you to represent some algorithm to the
underlying hardware. How does alignment help you do that?
The mindset of the Pascal committee (the only one I can speak for
personally) is that alignment (or any physical implementation detail)
doesn't add to the ability to describe more algorithmic detail. We left
things vague for things like interpreters, p-code, ASCII or EBCID, etc.
to all live within the standard.
bri...@encompasserve.org
Maybe I'm missing the point.
If all we're worrying about is a performance hint for ephemeral
storage layout within a single compilation unit then I certainly agree
that embedding a hint mechanism into the standard is silly.
But if we're worrying about the ability to specify field layout
for purposes of cross-language programming or for data storage on
a file or data transport across the network then the ability of
the programmer to apply some sort of "representation specification"
is surely important. Encoding the relevant syntax and semantics
into the standard is surely a reasonable way to ensure both
portability and interoperability.
I may even want to know that the record data I pass in main line
code compiled with version 3.1 will be properly understood in
a subroutine compiled under version 3.2. [On the other hand,
I'd rather expect my vendor to keep things consistent enough
that I don't have to overspecify things just to get alignment behavior
that is consistent between minor releases of the same compiler]
Bill Gunshannon
>
>> proprietary extension. The original point was that the PLI/I language
>> took alignment into consideration and included alignment instructions
>> as a part of the language which means it is reasonable to expect a
>> program using it to be somewhat portable. C and Pascal do not share
>> this feature. Any concern for alignment is little more than an after-
>> thought.
>>
>
> Well, so does COBOL (see the SYNCRONIZED clause).
No one said it didn't. The languages referenced were Pascal and C.
COBOL does a lot of stuff that other languages don't, kind of liker
PL/I. :-)
>
> As a member of a standards committee, it is hard for a language to
> predict behavior & performance of future hardware. COBOL and PL/I
> provide some sort of "using my crystal ball, I think I need some
> alignment help right here". What if my crystal ball is broken and I
> actually needed alignment help somewhere else?
Yeah, but it seems to work, at least most of the time.
>
> Allowing/requiring the human to "help" the compiler produce code on some
> current hardware (much less future hardware) seems silly to me.
I guess I'm just old school. I hate languages that hogtie me and then
expect me to get some work done. I still think tyhe programmer should
actually know what the comouter is doing and should have a say in how
best to do it.
> Kinda
> like the "register" declaration in C. Not really part of any language
> design, but just a crutch for a brain-dead compiler.
Sorry, I don't agree but then, I don;t know how any of the compilers
you get to work on do it. My experience has always been that the
compiler is free to use registers for normal variables if it wants,
but need not. The "register" directive was provided to allow the
programmer the ability to take some variable that he thinks will
benefit from being kept in a register and force it to be there.
Of course, over use on a machine with insufficient resources is just
another sign of an incompetent programmer.
>
> Computer languages allow you to represent some algorithm to the
> underlying hardware. How does alignment help you do that?
I don't understand the question. Not everything a computer does
is mathematical algorithms. One needs to store information in
memory. Some machines have a serious efficiency penalty when this
results in unaligned access. Some way to fix this is needed. Even
Macro-11 had one.
>
> The mindset of the Pascal committee (the only one I can speak for
> personally) is that alignment (or any physical implementation detail)
> doesn't add to the ability to describe more algorithmic detail. We left
> things vague for things like interpreters, p-code, ASCII or EBCID, etc.
> to all live within the standard.
No problem with that. It was a conscious decision made by the standards
committee after due consideration and that is fine. COBOL originally did
the same thing over things like Range Checking. But, programmers should
know this and take it into consideration when writing code. Some people
might think this makes PL/I a superior language. :-) Others are just
as happy with Pascal and C. And, yes, I still like COBOL. But I know
not to trust the compiler to keep all my arrays within bounds.