Sheffield Pascal

[Dennis Boone]

J R Gilbert of U Sheffield wrote a document on the Sheffield Pascal
compiler in 1987. It appears he's still at U Sheffield:

R.Gi...@sheffield.ac.uk

Has anyone here talked to him to see if he might still have this
software in rescuable form?

De

[marlow...@googlemail.com]

On Jun 23, 5:07 pm, d...@ihatespam.msu.edu (Dennis Boone) wrote:
> J R Gilbert of U Sheffield wrote a document on the Sheffield Pascal
> compiler in 1987. It appears he's still at U Sheffield:
>

> R.Gilb...@sheffield.ac.uk

>
> Has anyone here talked to him to see if he might still have this
> software in rescuable form?
>
> De

I am suprised to hear of a Pascal compiler for the Prime other than
the Prime one and the one produced by the University of Hull. I'll
always remember the Hull one because the command to invoke it was HVP.
It stood for 'Hull V-mode Pascal'.

How does the Sheffield one compare with the one from Hull?

-Andrew Marlow

[Dennis Boone]

> I am suprised to hear of a Pascal compiler for the Prime other than
> the Prime one and the one produced by the University of Hull. I'll
> always remember the Hull one because the command to invoke it was HVP.
> It stood for 'Hull V-mode Pascal'.

> How does the Sheffield one compare with the one from Hull?

Seems like we had a discussion here recently about the Prime one having
been written by someone else. Memory fails as usual, no recollection
of who the "else" was. Did they buy the Hull compiler?

The reason I know about the Sheffield compiler is that the Sheffield
Editor was built using it. Since the source for the editor is
available, it'd be nice to obtain the compiler too.

I've not actually _used_ any of the pascal compilers on a Prime. We
never had access to them. The Prime one is installed on the 19.2
emulator though, so one could e.g. look at expanded listings to see if
the format is familiar.

De

[marlow...@googlemail.com]

On Jun 24, 2:52 pm, d...@ihatespam.msu.edu (Dennis Boone) wrote:
> Seems like we had a discussion here recently about the Prime one having
> been written by someone else. Memory fails as usual, no recollection
> of who the "else" was. Did they buy the Hull compiler?

I'm pretty sure the HVP was not used as the basis of the Prime Pascal
compiler. The latter had this really irritating error message 'source
skipped untl this point' which HVP never had. The 'source skipped'
error usually meant it skipping over some random number of tokens that
meant it would not be able to interpret the next bit either. Errors
cascaded rapidly.

[Jim Wilcoxson]

I'm 99% sure the Prime Pascal compiler was based on the TSI
(Translator Systems Inc.) compiler technology, also used by the F77,
PL/I-G, full PL/I, CBL, VRPG, and SPL compilers. If you look in
SYSOVL on the 19.2 emulator, there is a PASCALDATA file just like for
the other TSI compilers. I don't know whether TSI wrote and supported
all of these compilers, or whether Prime eventually brought them in-
house and did all the development and support themselves. TSI became
LPI (Language Processors Inc.), or was bought by them, then was bought
by Ryan-McFarland, the guys that did PC COBOL compilers, and then
became Liant, Inc. Liant is still in business and selling compiler
technology.

I've seen source for these compilers, but no longer have it. Lots of
FIND_NODE calls all over the place as I recall!

Jim

[teache...@gmail.com]

I studied Computational Science at Hull University at the time that HVP was being developed. It was written by Barry Corneius and two other lecturers whose first names were Ian and Dave. There was also a Pascal compiler for the PR1ME computer written by the University of Braunschweig which wasn't as well used. From what I can recall, HVP was developed using Nicklaus Wirth's compiler. Barry, Ian and Dave stated development around 1979.

For more information, my email is groun...@hotmail.co.uk