31 Dec
2017
31 Dec
'17
5:20 a.m.
A bit off the PDPs, but to do a minor correction on mail below
The commercial version of 'UNIX' on Alpha was maybe first called
Digital Unix OSF/1, but quickly changed to Digital Unix at least with
v3 and v4.0 (A - G). From there we had a 'break' which only in part
was due to take over by Compaq and we had Tru64 UNIX v5.1A and V5.1B.
The V5.1B saw updates till B-6.
As for the Digital C compiler, I'm still using
DTCCMPLR650 installed Compaq C Version 6.5 for Compaq Tru64 UNIX Systems
When I get some old source (some even developed on SCO UNIX 3.2V4.2) I
like to run it through all compiler /OS-es I got handy. With the
Compaq C compiler and HP-UX ANSI C I mostly get pages of warning and a
few errors. By the time I 'corrected' what I think is relevant some
nasty coredumps tend to disappear :-)
Compile for a better 2018,
uncle rubl
Date: Fri, 29 Dec 2017 21:30:11 -0500.
From: Paul Winalski <paul.winalski(a)gmail.com>
To: Ron Natalie <ron(a)ronnatalie.com>
Cc: TUHS main list <tuhs(a)minnie.tuhs.org>
Subject: Re: [TUHS] Why did PDPs become so popular?
Message-ID: <CABH=_VRwNXUctFPav5rHX83wfUS0twMQuBhinRZ6QEY1cB3TNQ(a)mail.gmail.com>
Content-Type: text/plain; charset="UTF-8"
On 12/29/17, Ron Natalie <ron(a)ronnatalie.com> wrote:
The Alpha was hot
stuff for about nine months. Ran OSF/1 formerly DigitalUnix formerly
OSF/1.
Digital UNIX for the VAX was indeed derived from OSF/1.
The port to
Alpha was called Tru64 UNIX.
Tru64 UNIX was initially a pure 64-bit system, with no
provision for
building or running 32-bit program images. This turned out to be a
mistake . DEC found out that a lot of ISVs had code that implicitly
"knew" that sizeof() a pointer was the same as sizeof(int) was the
same as 4 bytes. Tru64 was forced to implement a 32-bit compatibility
mode.
There was also a problem with the C compiler initially
developed at
DECwest in Seattle. It supported ONLY ANSI standard C and issued
fatal errors for violations/extensions of the standard. We (DEC
mainstream compiler group) called it the Rush Limbaugh
compiler--extremely conservative, and you can't argue with it.
31 Dec
31 Dec
12:56 p.m.
New subject: Why did PDPs become so popular?
On Dec 31, 2017, at 12:20 AM, Rudi Blom
<rudi.j.blom(a)gmail.com> wrote:
.With the
Compaq C compiler and HP-UX ANSI C I mostly get pages of warning and a
few errors. By the time I 'corrected' what I think is relevant some
nasty coredumps tend to disappear :-)
I had to chuckle when we read these two lines. As Paul I’m sure also remembers and
mentioned about the 32/64 bit issues, when we released Alpha there we very few to zero
tools available to help find errors in ISV or customer created C code that was caused by
assumptions of size. As you point out with the new ‘Gem’ compilers a great deal of work
was spent by Paul and his brothers and sisters in the compiler group putting in just those
messages. For C++ it was worse than C because the language was so new at the time few
tools for it existed. So one of our engineers, Judy Ward, lead the effort to make the C++
front end be extremely useful and offered extremely good errors and warnings (she also had
worked on the C FE before that).
When people complained about the compiler before no so “noisy” my response was “Listen to
Judy’s advise. She’s seeing something you are not and she knows more about C and C++ that
all of the rest of us.”
We discovered a curious thing. The ISVs reported way fewer bugs after the Alpha port
because Judy’s messages forced them to clean up long standing issues that were silently
causing problem on other systems. I’ve always said Alpha C and C++ compilers was the
greatest gift to the Sparc ISVs there was.
Like you to this day, except for possibly Gimple’s Flexilint product, the old DEC
compilers are still the best tools I have ever used to clean up code.
I’ve reminded and thanked Judy for this many times when I seen her. That was a labor of
love by some folks that really cared to make a great product as useful as it could be.
Clem
3:03 p.m.
New subject: Why did PDPs become so popular?
The Plan9 c Alpha compiler also uses 64bit pointers and 32bit ints. For several
years Bell labs kept the port alive, checking builds on real Alpha hardware long
after it was no-longer competitive on the basis that “It keeps us honest”.
-Steve
PS: Gimple’s Llint is a truly wonderful tool, it has saved me from myself many times.
On 31 Dec 2017, at 12:56, Clement T. Cole
<clemc(a)ccc.com> wrote:
On Dec 31, 2017, at 12:20 AM, Rudi Blom
<rudi.j.blom(a)gmail.com> wrote:
.With the
Compaq C compiler and HP-UX ANSI C I mostly get pages of warning and a
few errors. By the time I 'corrected' what I think is relevant some
nasty coredumps tend to disappear :-)
I had to chuckle when we read these two lines. As Paul I’m sure also remembers and
mentioned about the 32/64 bit issues, when we released Alpha there we very few to zero
tools available to help find errors in ISV or customer created C code that was caused by
assumptions of size. As you point out with the new ‘Gem’ compilers a great deal of work
was spent by Paul and his brothers and sisters in the compiler group putting in just those
messages. For C++ it was worse than C because the language was so new at the time few
tools for it existed. So one of our engineers, Judy Ward, lead the effort to make the C++
front end be extremely useful and offered extremely good errors and warnings (she also had
worked on the C FE before that).
When people complained about the compiler before no so “noisy” my response was “Listen to
Judy’s advise. She’s seeing something you are not and she knows more about C and C++ that
all of the rest of us.”
We discovered a curious thing. The ISVs reported way fewer bugs after the Alpha port
because Judy’s messages forced them to clean up long standing issues that were silently
causing problem on other systems. I’ve always said Alpha C and C++ compilers was the
greatest gift to the Sparc ISVs there was.
Like you to this day, except for possibly Gimple’s Flexilint product, the old DEC
compilers are still the best tools I have ever used to clean up code.
I’ve reminded and thanked Judy for this many times when I seen her. That was a labor of
love by some folks that really cared to make a great product as useful as it could be.
Clem
2302
days inactive
2302
days old
2 comments
3 participants
participants (3)
-
Clement T. Cole -
Rudi Blom -
Steve Simon