Todays compiler webinar.

[Jan-Erik Söderholm]

Hi.

I missed todays compiler webinar. Anyone who attended and that
have some memorable points from the event?

[Phillip Helbig (undress to reply)]

In article <s69lfm$kuh$1...@dont-email.me>,
=?UTF-8?Q?Jan-Erik_S=c3=b6derholm?= <jan-erik....@telia.com>
writes:

> I missed todays compiler webinar. Anyone who attended and that
> have some memorable points from the event?

I think that the slides will be on YouTube soon if they aren't already.
Most of the presentation was going through the slides, but you could get
much of that by just reading the slides. If the YouTube stuff has the
voiceover, then all you miss is the opportunity to ask questions and
have them answered. Maybe the Q&A will be there as well.

Nice to hear that a native Fortran compiler is just around the corner,
and Fortran already running on x86. There will be cross-compilers for a
while, but that is not the long-term plan, with all native at some
point. Via Flang even very recent Fortran standards should be possibe
to use.

As always, for Rdb-specific stuff ask Oracle, but my impression is
confirmed that the two teams (VSI and Oracle Rdb) are in very close
contact.

[abrsvc]

The Webinar should be posted on the VSI site within 48 hours. You can see/hear it there soon.

Dan

[John Reagan]

Yes, I mostly read the slides along with some additional information and bad jokes.

The Fortran cross-compiler is in pretty good shape. I can cross-compile, link, and run Adventure
on my VirtualBox. Works fine.

The native compilers will be using a newer LLVM so there is always some risk of something coming
loose in the bootstrapping step but we'll sort it. And then we'll do a 2nd generation build by using the native
compilers to build themselves and compare the output. There will be lots of cross vs native testing
and comparisons.

[Rich Alderson]

John Reagan <xyzz...@gmail.com> writes:

> The native compilers will be using a newer LLVM so there is always some risk
> of something coming loose in the bootstrapping step but we'll sort it. And
> then we'll do a 2nd generation build by using the native compilers to build
> themselves and compare the output. There will be lots of cross vs native
> testing and comparisons.

The 2Gen build is where the snakes come out ot the rocks, of course.

--
Rich Alderson ne...@alderson.users.panix.com
Audendum est, et veritas investiganda; quam etiamsi non assequamur,
omnino tamen proprius, quam nunc sumus, ad eam perveniemus.
--Galen

[Robert A. Brooks]

On 4/27/2021 6:26 PM, Rich Alderson wrote:
> John Reagan <xyzz...@gmail.com> writes:
>
>> The native compilers will be using a newer LLVM so there is always some risk
>> of something coming loose in the bootstrapping step but we'll sort it. And
>> then we'll do a 2nd generation build by using the native compilers to build
>> themselves and compare the output. There will be lots of cross vs native
>> testing and comparisons.
>
> The 2Gen build is where the snakes come out ot the rocks, of course.

Ah, we've all been bitten and strangled by enough snakes already with the cross compilers.

However, John has become a very good snake charmer.

--

-- Rob

[Chris Townley]

You should try to turn them into Happy compilers :)

Chris

[Phillip Helbig (undress to reply)]

In article <7919d207-e540-4e80...@googlegroups.com>,

John Reagan <xyzz...@gmail.com> writes:

> Yes, I mostly read the slides along with some additional information
> and bad jokes.

I guess ADA goes into the bad-joke category. :-D

[Ian Miller]

I just had to ask about ADA :-)

[Arne Vajhøj]

It comes up frequently.

And maybe the software world would have been better if Ada had
become *the* native code language. But the fact is that most
of the world moved in different directions.

If we look at existing applications, then I am sure that VSI
would be happy if ACT announced VMS support, but when prioritizing
their own resources then Ada does not seem to be a top priority.
Which I suspect is a correct analysis of the market.

If we look at new applications, then support for Rust and
Go may be more relevant than Ada.

Arne

[Jon Schneider]

Very interesting but I found myself, not for the first time as the only
participant and could not see anything in the Q&A or Chat though it was
referred to.

I used the temporary (Windows) executable access method.

Jon

[Jan-Erik Söderholm]

Right, same thing last time. Seems as that the webinar is setup to
hide everyone else so you look as being alone there...

[Michael Moroney]

Yes, it looked like I was alone there, and I was certain I wasn't.

[Simon Clubley]

On 2021-04-27, John Reagan <xyzz...@gmail.com> wrote:
>
> The Fortran cross-compiler is in pretty good shape. I can cross-compile, link, and run Adventure
> on my VirtualBox. Works fine.
>

Good. It's nice to see you have the important stuff tested. :-)

Simon.

--
Simon Clubley, clubley@remove_me.eisner.decus.org-Earth.UFP
Walking destinations on a map are further away than they appear.

[Simon Clubley]

What was said about Ada ?

[Simon Clubley]

On 2021-04-28, Arne Vajhøj <ar...@vajhoej.dk> wrote:
> On 4/28/2021 9:27 AM, Ian Miller wrote:
>> On Wednesday, April 28, 2021 at 7:53:48 AM UTC+1, Phillip Helbig (undress to reply) wrote:
>>> In article <7919d207-e540-4e80...@googlegroups.com>,
>>> John Reagan <xyzz...@gmail.com> writes:
>>>> Yes, I mostly read the slides along with some additional information
>>>> and bad jokes.
>>> I guess ADA goes into the bad-joke category. :-D
>>
>> I just had to ask about ADA :-)
>
> It comes up frequently.
>
> And maybe the software world would have been better if Ada had
> become *the* native code language. But the fact is that most
> of the world moved in different directions.
>

Ada would have been very nice, I would have still been happy with
something that was not as strict as Ada, but was still well along the
Wirth language line as the would have still been an improvement.

There's a reason why I keep wishing that DEC had continued with Pillar.

[Jon Schneider]

On 28/04/2021 18:32, Simon Clubley wrote:
> On 2021-04-28, Phillip Helbig (undress to reply) <hel...@asclothestro.multivax.de> wrote:
>> In article <7919d207-e540-4e80...@googlegroups.com>,
>> John Reagan <xyzz...@gmail.com> writes:
>>
>>> Yes, I mostly read the slides along with some additional information
>>> and bad jokes.
>>
>> I guess ADA goes into the bad-joke category. :-D
>>
>
> What was said about Ada ?

It was taken as seriously as PL/I.

Jon

[John Reagan]

That isn't true at all.

For Ada, what I said was: "VSI understands that Ada is very important to small set of customers. We are still investigating how we can provide a solution."

For PL/I, I said that "VSI has no ownership in the old PL/I compiler. That ended up with Tom Linden at Kednos and that Tom passed away in 2019 I believe".

In my personal opinion, I think we should provide/support/encourage an Ada product for OpenVMS. I'm a fan of strong typing. That's why I'm also a Rust fan. It was also mentioned and I think I replied that I've been in touch with the Rust developers.

[Dave Froble]

I prefer Basic. It does what I tell it to do, not what it thinks I
should do.

:-)

--
David Froble Tel: 724-529-0450
Dave Froble Enterprises, Inc. E-Mail: da...@tsoft-inc.com
DFE Ultralights, Inc.
170 Grimplin Road
Vanderbilt, PA 15486

[Arne Vajhøj]

On 4/28/2021 5:12 PM, Dave Froble wrote:
> On 4/28/2021 4:14 PM, John Reagan wrote:
>> For Ada, what I said was: "VSI understands that Ada is very important
>> to small set of customers.  We are still investigating how we can
>> provide a solution."
>>
>> For PL/I, I said that "VSI has no ownership in the old PL/I compiler.
>> That ended up with Tom Linden at Kednos and that Tom passed away in
>> 2019 I believe".
>>
>> In my personal opinion, I think we should provide/support/encourage an
>> Ada product for OpenVMS.  I'm a fan of strong typing.  That's why I'm
>> also a Rust fan.  It was also mentioned and I think I replied that
>> I've been in touch with the Rust developers.
>
> I prefer Basic.  It does what I tell it to do, not what it thinks I
> should do.

I don't see Rust and VMS Basic as alternatives for each other.

Rust will compete with what C, C++, Ada etc. for metal near code.

VMS Basic is for writing business code and alternatives are
more Pascal, Cobol, PL/I and the new stuff (JVM languages,
Python, PHP etc.).

Well - C/C++ has also been used for business code, but
"misused" may be more accurate.

Arne

[Tim Sneddon]

John Reagan <xyzz...@gmail.com> wrote:
> On Wednesday, April 28, 2021 at 1:43:10 PM UTC-4, Jon Schneider wrote:
>> On 28/04/2021 18:32, Simon Clubley wrote:
>> > On 2021-04-28, Phillip Helbig (undress to reply) <hel...@asclothestro.multivax.de> wrote:
>> >> In article <7919d207-e540-4e80...@googlegroups.com>,
>> >> John Reagan <xyzz...@gmail.com> writes:
>> >>
>> >>> Yes, I mostly read the slides along with some additional information
>> >>> and bad jokes.
>> >>
>> >> I guess ADA goes into the bad-joke category. :-D
>> >>
>> >
>> > What was said about Ada ?
>> It was taken as seriously as PL/I.
>>
>> Jon
> That isn't true at all.
>
> For Ada, what I said was: "VSI understands that Ada is very
> important to small set of customers. We are still
> investigating how we can provide a solution."
>
> For PL/I, I said that "VSI has no ownership in the old PL/I
> compiler. That ended up with Tom Linden at Kednos and that
> Tom passed away in 2019 I believe".

Hopefully that will become a solved problem in the not too distant
future.

Regards, Tim.

[Phillip Helbig (undress to reply)]

In article <s6cr4s$1t35$1...@gioia.aioe.org>, =?UTF-8?Q?Arne_Vajhøj?=

<ar...@vajhoej.dk> writes:

> >> In my personal opinion, I think we should provide/support/encourage an
> >> Ada product for OpenVMS. I'm a fan of strong typing. That's why I'm
> >> also a Rust fan. It was also mentioned and I think I replied that
> >> I've been in touch with the Rust developers.
> >
> > I prefer Basic. It does what I tell it to do, not what it thinks I
> > should do.
>

> I don't see Rust and VMS Basic as alternatives for each other.
>
> Rust will compete with what C, C++, Ada etc. for metal near code.
>
> VMS Basic is for writing business code and alternatives are
> more Pascal, Cobol, PL/I and the new stuff (JVM languages,
> Python, PHP etc.).

And, of course, anything can be written in Fortran. :-)

And a good Fortran programmer can write Fortran in any language.

[Simon Clubley]

On 2021-04-29, Phillip Helbig (undress to reply) <hel...@asclothestro.multivax.de> wrote:
>
> And a good Fortran programmer can write Fortran in any language.
>

I would question the use of the word "good" in the above statement. :-)

[Dave Froble]

Must be interesting to see the definition of "business code", vs "????".

I'm thinking the following header from a tool might not fit any such
description?

!
********************************************************************
!
! Program: RMS_LOCKS.BAS
! Function: Watch RMS Locks
! Version: 1.00
! Created: 10-May-2001
! Author(s): David Froble
! Dave Froble Enterprises, Inc.
!
! Purpose/description:
!
! This program will scan the lock database selecting
! RMS created locks. It will build data structures
! for filenames, opens, locks, and blocked locks.
!
! A user can select to see the entire data structure,
! files, opens, and locks, or just blocked locks.
!
!********************************************************************

But, what do I know ????

How about a messaging utility?


!********************************************************************
!
! Program: TQUMAN.BAS
! Function: TQUMAN - Message Queue Manager
! Version: 2.00
! Created: 09-Nov-88
! Author(s): Dave Froble
!
! Purpose/description:
!
! This program will run detached to manage the
! system message queue file.
!

!********************************************************************

Then again, I guess just about anything could be considered "business".

[Arne Vajhøj]

So if I take this Fortran (that I have shown before):

PROGRAM GOTOFUN
INTEGER*4 LBL
INTEGER*4 I
731 GOTO 113
ASSIGN 399 TO LBL
113 I=0
WRITE(*,*) I
247 ASSIGN 601 TO LBL
IF(I) 588,922,399
588 I=I+1
WRITE(*,*) I
GOTO 399
761 GOTO LBL
922 I=I+1
WRITE(*,*) I
GOTO(588,399,922) I
399 GOTO 761
601 CONTINUE
END

and convert it to "Pascal":

PROGRAM GOTOFUN(INPUT,OUTPUT);
LABEL L731,L113,L247,L588,L761,L922,L399,L601;
VAR LBL:INTEGER;
VAR I:INTEGER;
BEGIN
L731: GOTO L113;
LBL:=399;
L113: I:=0;
WRITELN(I);
L247: LBL:=601;
IF(I<0)THEN GOTO L588 ELSE IF(I=0)THEN GOTO L922 ELSE GOTO L399;
L588: I:=I+1;
WRITELN(I);
GOTO L399;
L761: CASE(LBL)OF 399: GOTO L399; 601: GOTO L601; END;
L922: I:=I+1;
WRITELN(I);
CASE(I)OF 1: GOTO L588; 2: GOTO L399; 3: GOTO L922; END;
L399: GOTO L761;
L601: END.

then ...

:-) :-) :-)

Arne

[Arne Vajhøj]

On 4/28/2021 11:35 PM, Tim Sneddon wrote:

> John Reagan <xyzz...@gmail.com> wrote:
>> For Ada, what I said was: "VSI understands that Ada is very
>> important to small set of customers. We are still
>> investigating how we can provide a solution."
>>
>> For PL/I, I said that "VSI has no ownership in the old PL/I
>> compiler. That ended up with Tom Linden at Kednos and that
>> Tom passed away in 2019 I believe".
>
> Hopefully that will become a solved problem in the not too distant
> future.

How many PL/I users are left on VMS?

I still think that you should have made a version targeting JVM.
Then porting would not be your problem. And given that RainCode
made a PL/I compiler targeting .NET then it should be possible.

Arne

[Arne Vajhøj]

On 4/29/2021 12:38 PM, Dave Froble wrote:
> On 4/28/2021 7:29 PM, Arne Vajhøj wrote:
>> On 4/28/2021 5:12 PM, Dave Froble wrote:
>> I prefer Basic.  It does what I tell it to do, not what it thinks I
>>> should do.
>>
>> I don't see Rust and VMS Basic as alternatives for each other.
>>
>> Rust will compete with what C, C++, Ada etc. for metal near code.
>>
>> VMS Basic is for writing business code and alternatives are
>> more Pascal, Cobol, PL/I and the new stuff (JVM languages,
>> Python, PHP etc.).
>>
>> Well - C/C++ has also been used for business code, but
>> "misused" may be more accurate.
>

> Must be interesting to see the definition of "business code", vs "????".

There is a traditional distinction between business code, scientific
code and system code.

Definitions are probably not super clear.

> I'm thinking the following header from a tool might not fit any such
> description?

>        !       Purpose/description:

>        !
>        !               This program will scan the lock database selecting
>        !               RMS created locks.  It will build data structures
>        !               for filenames, opens, locks, and blocked locks.
>        !
>        !               A user can select to see the entire data structure,
>        !               files, opens, and locks, or just blocked locks.

This must be system code.

>        !       Purpose/description:
>        !
>        !               This program will run detached to manage the
>        !               system message queue file.

This could be system code or business code depending on what type
of messages it is and how it manage.

Is it some data on disk / in memory and CRUD operations? Or
is it messages from customers that needs to be responded to in a timely
manner.

Arne