Some N00B Medley Interlisp questions

[Alexander Shendi]

Dear list,

I have successfully built the emulator (maiko) and started the run-medley script (under OpenBSD 7.0-beta/amd64). Now I need to find my way round the system.
I have a few questions:
* The graphics subsystem is monochrome only, like Smalltalk 80?
* Is there a bitblt function?
* Is there a plot function that e.g. takes two sequences of the same length and plots the corresponding graph in a new window?
* How do I find out the answers to the above? RTFM, but how? DInfo?

Many thanks in advance for helping me out.

Have a nice day or night or whatever :)

Alexander
--
Diese Nachricht wurde von meinem Android-Gerät mit K-9 Mail gesendet.

[Nick Briggs]

> On Aug 31, 2021, at 1:51 PM, Alexander Shendi <Alexande...@web.de> wrote:
>
> Dear list,
>
> I have successfully built the emulator (maiko) and started the run-medley script (under OpenBSD 7.0-beta/amd64). Now I need to find my way round the system.
> I have a few questions:
> * The graphics subsystem is monochrome only, like Smalltalk 80?

Yes.

> * Is there a bitblt function?

Yes, it's called BITBLT if you're in the Interlisp name space or IL:BITBLT if your default package is something else (e.g., from the XCL Exec)

> * Is there a plot function that e.g. takes two sequences of the same length and plots the corresponding graph in a new window?

If I had to guess, I'd say yes, and probably in a Lispusers package

> * How do I find out the answers to the above? RTFM, but how? DInfo?

Since I think we haven't got all the PDFs of the Lispusers documentation (from the Postscript generated from the TEdit documentation files) up yet, i'd start by looking at the filenames in medley/lispusers for a likely name (e.g., PLOT...) and then from a running system (TEDIT (FINDFILE 'PLOT.TEDIT)) -- if you started it with "run-medley" it should have set up the DIRECTORIES variable to include all the likely places for interesting files so that FINDFILE will find it without you having to "cd" to the right place or speak the full filename.

>
> Many thanks in advance for helping me out.
>
> Have a nice day or night or whatever :)
>
> Alexander
> --
> Diese Nachricht wurde von meinem Android-Gerät mit K-9 Mail gesendet.
>

> --
> https://Interlisp.org for more details
> ---
> You received this message because you are subscribed to the Google Groups "Medley Interlisp" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to interlisp+...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/interlisp/4942D327-7AB7-47BB-8398-8B25E26DC7E9%40web.de.

[Nick Briggs]

Also... look in the medley/docs directory -- where there are a bunch of files that we have already made PDFs of so they're easily browsable... there's quite a bit of introductory material there.

> On Aug 31, 2021, at 1:51 PM, Alexander Shendi <Alexande...@web.de> wrote:
>

[Larry Masinter]

I'm hoping we can put together a tour of the system and its documentation.

https://github.com/Interlisp/medley/wiki lists lots of document; I've been using the home of medley/wiki to point to all the documents.


https://interlisp.org/docs/Medley-Primer.pdf
looks like you might find it helpful (and please let us know if not)



--
https://LarryMasinter.net https://interlisp.org

[Alexander Shendi]

Dear list,

Thanks for your replies so far. They have been very helpful. Now I have a bunch more.

1. lde (or ldex, I don't know the difference) is started with the argument "-m 256" by the run-medley script. Is this somehow a hard limit or could I increase it to 1024 or 2048?
2. I want to port a Scheme program to Interlisp. I have managed to open the Scheme file in TEDIT via the Interlisp EXEC. Embarassingly enough I can't seem to figure out how to scroll in the TEDIT window.
3. Is there a way to write my own docs. I am assuming that I probably should use TEDIT and then convert to PDF. Is there a way to render math equations?
4. Would anyone be intrested, if I made a video of the process?

Again thanks in advance for your help and have a nice day.

/Alexander

[Nick Briggs]

> On Sep 1, 2021, at 1:18 PM, Alexander Shendi <Alexande...@web.de> wrote:
>
> Dear list,
>
> Thanks for your replies so far. They have been very helpful. Now I have a bunch more.
>
> 1. lde (or ldex, I don't know the difference) is started with the argument "-m 256" by the run-medley script. Is this somehow a hard limit or could I increase it to 1024 or 2048?

256 (MB) is as big as the VM can be.

> 2. I want to port a Scheme program to Interlisp. I have managed to open the Scheme file in TEDIT via the Interlisp EXEC. Embarassingly enough I can't seem to figure out how to scroll in the TEDIT window.

Move the mouse out of the left edge of the TEdit window, so that the scroll bar pops up, then left (mouse) button scrolls up, right button down (an amount proportional to how far from the top/bottom the cursor is) and middle button "thumb scrolls" -- the scroll bar represents the whole document, where you let go of the middle button is where in the document you are.

You're probably not going to want to edit Lisp s-expressions in TEdit in the longer term (or even particularly short term!) -- you should get it into Lisp and use the structure editor, SEDIT, on the definitions.

> 3. Is there a way to write my own docs. I am assuming that I probably should use TEDIT and then convert to PDF. Is there a way to render math equations?

I don't think I've ever seen anyone try to do any complicated math expressions in TEdit. There are some math fonts around (math 8, 10, 12 point) but there's not a lot of support in TEdit to help you use them. Remember, you're in a timewarp back to 1989 or so.

> 4. Would anyone be intrested, if I made a video of the process?

I'm sure someone would be.

> To view this discussion on the web visit https://groups.google.com/d/msgid/interlisp/3546748B-B89F-4520-BBC0-6201EFD26D06%40web.de.

[Alexander Shendi]

Hi Nick,

Thanks for your answers. I wasn't really going to write Interlisp code in TEDIT. I just wanted to view the original Scheme file there. I would use DF to define new Interlisp functions via SEDIT.

All the best.

/Alexander

[Larry Masinter]

There is a LispUsers packages EQUATIONS for a Equation editor, and a related MATHTONS (Math to NS aka XCCS) for printing to Interpress printers which have the appropriate fonts. But wiring up Postscript / PDF is an open issue. Perhaps someone has an Interpress to PDF converter?

 

 

 

 

 

[Al Kossow]

On 9/2/21 11:29 AM, Larry Masinter wrote:
> There is a LispUsers packages EQUATIONS for a Equation editor, and a related MATHTONS (Math to NS aka XCCS) for printing to Interpress
> printers which have the appropriate fonts. But wiring up Postscript / PDF is an open issue. Perhaps someone has an Interpress to PDF converter?
>

I think Paul McJones has been working on that.

[Paul McJones]

No, I only have coverage for Bravo, Press, and Tioga (the latter was supplied by Josh Dersch, and is based on a C library written in 1990 by David Nichols at PARC). Doing an Interpress-to-PostScript converter would probably be relatively straightforward if documentation were available. BitSavers has:

http://www.bitsavers.org/pdf/xerox/xns/standards/XSIG_038306_Introduction_to_Interpress_Jun1983.pdf

and

http://www.bitsavers.org/pdf/xerox/xns/standards/XSIG_018404_Interpress_82_Readers_Guide_Apr1984.pdf

but not

lnterpress Electronic Printing Standard. XSIS 048306

I don’t even see any references to it, except for from the manuals in Bitsavers and archive.org’s mirror.

[Al Kossow]

On 9/2/21 12:09 PM, Paul McJones wrote:

> but not
>
> lnterpress Electronic Printing Standard. XSIS 048306
>
>

> I don’t even see any references to it, except for from the manuals in Bitsavers and archive.org <http://archive.org>’s mirror.

I have a paper copy, somewhere... in the box that has all of my other Xerox standards. I think I roughly know where that box is
since I've seen it in the past year.

[William Fisher]

There was and interpress to postscript translator written by Jerry  Mendelson who work in el Segundo.

I looked for the code in the Xerox software archive I have but did not find it. Since it was

used by the high end printing division in el Segundo, who knows where the code is!

Jerry is retired and still lives in Los Angeles.

Bill

--

https://Interlisp.org for more details
---
You received this message because you are subscribed to the Google Groups "Medley Interlisp" group.
To unsubscribe from this group and stop receiving emails from it, send an email to interlisp+...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/interlisp/A255D605-A062-4357-A45E-345F3A4D5D72%40mcjones.org.

[Nick Briggs]

There's also the DocuPrint NPS code that used the full Interpress implementation written in Cedar -- also distributed by PSD. 

I don't know where the archives of that code went, but I know we made CDs of every release.   If anyone can find those it would be handy (I have an old Solaris box that it runs on, but I've got a file system issue with the boot disk)

To view this discussion on the web visit https://groups.google.com/d/msgid/interlisp/CA%2B1K3up6BKT2AzW5qfvrUa9wr58-oBQ9yDDKqdqsJYP8pdPWGA%40mail.gmail.com.

[William Fisher]

There is a cedar source code archive at Parc and a few others have private copies that could be searched.

Bill

[Michele Denber]

On 09-02-2021 7:38 PM, William Fisher wrote:
>
> There is a cedar source code archive at Parc and a few others have
> private copies that could be searched.

Well while we're talking about lost documentation - I'm in great need of
the schematics for a 3Mb. Ethernet board for the Dolphin, the later
D0EN2 version, not the 1979 D0EN that's available on bitsavers.  The
board on my D0 has given up transmitting and the two versions are very
different  And a D0 extender board too, if anyone has a lead on one of
those rare items.  I'm looking everywhere I can possibly think of.

            - Michele

[William Fisher]

If the guys at digibarn do not have some if the old D0’s laying around, I have no idea. Alan freier and I had about 15 dandelions but we donated them and sold them to alumni.

Bill

--
https://Interlisp.org for more details
---
You received this message because you are subscribed to the Google Groups "Medley Interlisp" group.
To unsubscribe from this group and stop receiving emails from it, send an email to interlisp+...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/interlisp/acac9bfe-f5de-a9c9-b42c-88d772e7b129%40gmx.com.

[Paul McJones]

There was a CD-ROM that accompanied a CSL technical report:

Using Threads in Interactive Systems: A Case Study

Technical Report CSL 93-16

Computer Science Laboratory

Xerox Palo Alto Research Center

3333 Coyote Hill Road

Palo Alto, CA 94304

Copyright (c) Xerox Corporation 1993.  All rights reserved.

It contains a directory cedar/release/interpresstopostscript/ with source code in Cedar Mesa and also C (generated via a Cedar-to-C tool). I think it would take quite a bit of work to get the C to compile on modern machines (word size, endianness, etc.).

On Sep 2, 2021, at 6:38 PM, Nick Briggs <nicholas...@gmail.com> wrote:

On Sep 2, 2021, at 4:38 PM, William Fisher <zzyz...@gmail.com> wrote:

There is a cedar source code archive at Parc and a few others have private copies that could be searched.

Yeah, but since it was a product we didn't archive it in the regular Cedar archives, and I don't know that anyone at PARC was sensible enough to save all the DocuPrint NPS source code after we had all left. 

[Nick Briggs]

I'd forgotten that that part of the Interpress code went out.   It assumes as 32-bit big-endian system for the generated C code.

[Al Kossow]

Fisher may be interested in this.

https://github.com/devhawala/dodo/issues/4

https://github.com/devhawala/dodo

that I stumbled upon looking for the csl-93-16 cd

[John Cowan]

If Mimosa, the Cedar-to-C compiler, is still around somewhere (described at <http://www.bitsavers.org/pdf/xerox/parc/techReports/CSL-89-8_Experiences_Creating_a_Portable_Cedar.pdf>), it can be tuned for among other things big-vs-little-endian compilation.

To view this discussion on the web visit https://groups.google.com/d/msgid/interlisp/7E6EAEF0-187F-4222-9ABF-69D4FD5A5372%40gmail.com.

[Nick Briggs]

Russ Atkinson, who was in large part responsible for Mimosa said it was going to be WAY too much work to convert to little-endian.   We spent many years tossing this around when big-endian machines started to fade away at PARC.  We were building commercial products using Cedar and ended up sticking with SPARC systems because that was still easier than reworking Mimosa for x86.

-- Nick

[Larry Masinter]

It’s 30 years later. Don’t convert – emulate (QEMU) a sparc.

Index of /~aurel32/qemu/sparc (debian.org)

 

--

https://LarryMasinter.net https://interlisp.org

 

[William Fisher]

I always thought the whole Mimosa effort was not the right approach when I worked at SDD.

The whole idea on compiling Mesa to C and using GDB tools to debug the nightwear

of generated C code was too wired to be r4eal. You left the Mesa world with it's

integrated symbolic debugger to dive into GNU gdb and experiencing a complete

nightmare of regression.

Since PARC was the land of VM's, the Mesa Princops software emulator Xerox wrote

for small x86 machines seemed to be a more reasonable approach. The emulators

that exist today for the Mesa PrincOps are 100 times better than that x86 assembly

code Xerox initially wrote. I have all the code and getting it to work under

Windows 3.1 must have been the project from hell. They did device emulation

uner windows 3.1 aka extended DOS, for all the Dandelion devices and the amount

of code and it's complexity is numbing. Virtually unreadable today IMHO.

Bill

[Larry Masinter]

I started to reply here, but I added it to a Medley Issue Want FAQ

 

https://github.com/Interlisp/medley/issues/452#issuecomment-913038211

How do you debug in Medley?

And

What are the tradeoffs in performance/portability/debugability between emulation of virtual machine vs. compile time optimization.