Re: In the Shallow End

Dan Johnson wrote:

"GreyCloud" <mist@xxxxxxxxxxx> wrote in message news:yImdnTZfVdzT2kPZnZ2dnUVZ_uudnZ2d@xxxxxxxxxxxxxx

Dan Johnson wrote:

What are these wonderful DEC development tools you
refer to? When I used VMS- many years ago in my misspent
youth- the development tools were not impressive at all.

Their languages were complete and well documented. Their libraries were complete and well documented without leaving anything out that one might want.


MS usually manages this.

No, they haven't.

I don't recall VMS docs being
especially good, though. Better than Unix man pages, sure,
but what isn't?

The DEC docs are legendary. Maybe it is because you were given a simple account and weren't given the huge set of docs to read.
The DEC docs were very accurate, while the MS ones were not. Plus there were many needed simple functions that just didn't exist with MS.



Ex: you can open a file as indexed and key on anything you want to do.


Yes, I know about RMS. It has no relevance to my
question that I can see.


Hah. MS never included it in their file system structure. I found it to be a very big plus using VMS. Built right in.
There were so many conveniences in VMS that were outright non-existent in the MS o/s that it is amazing why anyone would want to use any MS products or even UNIX products.


Back then MS never had anything like it. You had to roll your own custom database back then.


MS offers several database products today for your databasing
pleasure. Modern relational databases kick the tar out of the old
indexed filesystems, like what VMS had.


Back then, MS didn't have a thing close to it. And they still don't.
Why else do the real pros on Wallstreet use VMS as the main transaction engines then and not MS?


I could also mix language very easily under VMS. All I had to do was link all the object modules together to make a complete program from using the various strengths of each language. Their stack frame was standardized, unlike what MS had pushed out where some languages you had to futz with the stack first.


Ya, this was a nice thing about VMS, but .NET goes
beyond it. It provides seamless interoperability for
objects, not just procedures.

..NET is a mess according to others accounts and appears to be a dying concept.



What was the deal in MS Fortran where you had to use [HUGE] or [LARGE] or [SMALL] and other non standard labels? Then they had to preface a call saying that the subroutine was in pascal. I never had to do this in vms. DEC made it dirt easy.


That stuff about HUGE, LARGE and SMALL was there
to support the 80286, and its craptacular segmented memory
model. They had to do extensions like that for all their
languages back then.


That wasn't the only kluge that MS had to bolt on.

But today, you don't need them and they compile away to
nothing.

Way back in the 16-bit days, MS called their standard calling
convention (they did have one!) the "pascal convention", which
was confusing. With Win32 they renamed it "stdcall", which is
unpronouncible. Pick yer poison.


Why bother, when VMS had their stuff straight from the git-go?


Anyway, it still seems a non-sequitur. MS has dropped
a very old language- Fortran- for a shiny new one- C#. This
seems to suggest forward motion on their part.

Intel just wrote Fortran 2003.


Presumably "just wrote" means "three years ago", right? :D



The very latest standard. But it appears you aren't that well informed are you.

There is a very good reason to use fortran where numeric accuracy has to be kept and flagged when you lose that accuracy. That was the purpose of many languages on one main frame in the older days. Each language had its own strengths and weaknesses.


Yes. FORTRAN still has the most manic optimizers,
for numerical computing.

Guffaw!! Scientific computing appears to not be your forte.


[snip]

I do not think MS can be blamed for Apple mistakes,
nor those of the Unix vendors. If Apple did not produce
great products sooner, that's *Apple's* fault. They wasted
a lot of time and effort producing not much in the mid
nineties.

Let me put it a different way. Back before IBM made the PC, there were many micro-computer makers that were rolling out some nice machines that could do multi-user/multi-tasking at a low cost. Then along came the IBM PC.


I take a less rosy view of this period.

I don't. MS tactics actually stalled the marketplace there.
The only real computing was still on minis and mains.


Back then, innovations could only occur at the
level of the whole computer. The Amiga had
nifty graphics accelerator hardware. The Mac
had a nifty GUI.

The Amiga was the only one out there that offered multi-tasking and good color graphics at that time. The death of the Amiga was caused by poor management and their lack of understanding of NOT to enter competition with MS DOS by producing a PC clone. Had they not, they might still be around... but they are not.


But to deliver this, Commodore had to build
the Amiga; and Apple the Mac.

Today an innovation can be deployed without
shipping a whole new computer to host it. This
is a great advance. And it's Windows that makes
it possible.

Windows is too steeped in backward compatibility to allow it to be trully innovative.


(OK, the open architecture of the PC deserves
some credit too. But Windows was the missing
link to make it actually work.)


Everybody jumped over there due to Big Blues influence. Everybody was hoping for some of that mainframe technology to filter down to the PC... it never happened.


Sure it did. It just took awhile.


Yes, over a decade... then win3.1 finally came out and bits and pieces of the main frame environment started to show up about the time NT was named win2000. But even these are not true multi-user environments, excluding the win2003 server of course.
Now it is steeped in buzz words that have no real meaning or connection for anyone these days. In this, MS has muddied the waters so bad that people are now losing interest.


Every one waited a decade for MS to come out with a multi-tasking o/s that never showed up.


NT is here now. Hell, they even shipped the OS/2 that
they promised, though that didn't pan out too well.


It wasn't what people wanted, and were too late in delivering it.
Multi-tasking executives were available, yet MS never delivered one in the mid 80s. Even the el-cheapo Coco from Tandy had OS9 for it.
MS should have delivered multi-tasking in their 3.0 version of MS DOS.
It would have been the next logical step.


By then most of the other vendors died out leaving the public with MS-DOS. That's what I mean by stagnation.
The industry was pushed back a decade and shouldn't have happened.


It's only pushed back if you *desired* the mess that
was the computer industry in the 80s. I don't think
many people would desire that.


There were no mess... what you saw was true competition. What you desire is a one vendor market... bad idea. Of course so did Bill.

[snip]

This is an excellent development. Microsoft is at its best
when competing with other vendors. During the period that
MS was unchallenged, they spent their time animating ordinary
office products with a perverse unlife.

In this case, MS went out of their way to derail Java by incorporating their own changes to Java... which led to a lawsuit.


Just to quibble: they were trying to derail Sun's
plans. They had other plans for Java, and didn't (then)
intend to kill it.

No, they didn't want to kill it... they wanted it for themselves... you know... steal by stealth.


Of course, if you care about politics more than technology,
that will seem to be a distinction without a difference.


I have no way to change what had happened to the computer industry, but only observe why it went the way it did. One problem was the good ol-boy club of executives like those at DEC, expecting huge sums of money in an obviously changing marketplace. That's why they failed, not from being very innovative, but failing to see where things were going and failing to see what MS was doing to the marketplace.
Once the impetus of the Gates empire put PCs in almost everyones homes, the necessary force needed to change this trend will not exist. What will happen is when empire rot occurs on the inside of MS, which I see happening now. It is like a big oil tanker... takes time to turn and time to stop. All it really is ... inertia. Steve Jobs knows this so is doing the next best thing... build a better product for the masses and make it easier to use than the competitors product.



--
Where are we going?
And why am I in this handbasket?
.

Relevant Pages

  • Re: COMMON structures and text parsing
    ... The VMS operating system, designed in conjunction with VAX hardware, had ... a novel concept - a common calling standard which supported all ... programming languages. ...
    (comp.lang.fortran)
  • Re: Microsoft pro says OpenVMS best OS ever invented!
    ... He actually knows VMS. ... to mix multiple languages in your app" ... the language interoperability ... ... Solak points out that architecture is something which VMS had from day 1, ...
    (comp.os.vms)
  • Re: OpenVMS security?
    ... OpenVMS is mostly written in a variety of type-safe languages, ... The main VMS implementation languages are VAX macro, Bliss, and C. ... In one description of very early VMS security I read, ... >to the damned Opteron? ...
    (comp.os.vms)
  • Re: Loose Cannon-dian
    ... "security" was the underlying principle in everything VMS did any more ... descriptors for varying-length items encourages careful programming ... Languages deficient in such constructs ... OS's like Multics, Primos, VOS, ...
    (comp.os.vms)
  • Re: COBOL Transactions?
    ... You've got two months of VMS experience, ... 27 years of experience with COBOL now, in a very wide variety of environments. ... something that is NOT true of other languages. ...
    (comp.os.vms)