FW: Putting together a Notecards demo

78 views
Skip to first unread message

Larry Masinter

unread,
May 5, 2021, 5:55:54 PM5/5/21
to fr...@halasz.org, Interlisp core

I’ve been thinking about Notecards as a cloud service.

You go to notecards.interlisp.org and log in, it spins up your vmem, and there is your environment. We’re close to being able to build on a minimal linux a Docker container.

 

Anyway the schedule is ‘as soon as we can’.

 

In addition to the Notecards in the Envos release, there are additional snapshots of software that come with a different license from PARC to the Computer History Museum.

Those haven’t been made public yet, but Paul McJones might help you get access.

 

--

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

 

From: Frank Halasz <fr...@halasz.org>
Sent: Wednesday, May 5, 2021 2:06 PM
To: Larry Masinter <L...@acm.org>
Subject: Re: Putting together a Notecards demo

 

Hi Larry,

 

Do you have a timeframe for when you want to do an Interlisp/NoteCards demo?  And is it a one-time event or somethings that sits there for a while to be played with?

 

I have Interlisp up and running on a Raspberry Pi from your github site..  Pretty cool and reasonably performant.

 

From what's on github, it looks like no one has tried to get NoteCards running on the current Interlisp platform.  I found multiple copies of the NoteCards sources in the envos repository and am rummaging around trying to figure out what the various copies are about.  

 

It's been almost 30 years since I last worked on Interlisp and it appears that I've forgotten quite a bit of how to get around the environment.  So I will spend a bit of time coming back up to speed and then see if I can get one of the NoteCards copies in the envos repository up and running again.  I will also look around to see if I can find some old NoteCards databases that might be worth showing off.

 

-- Frank

 

 

 

On Fri, Apr 23, 2021 at 12:18 PM Larry Masinter <L...@acm.org> wrote:

Hi Frank,

We’ve been working on a project https://interlisp.org to “restore” a working Interlisp-D / Medley.

Venue (John Sybalsky) had not only Interlisp/Medley but 3 other products: Notecards, Loops, Rooms.

 

I’ve talked to the Computer History Museum about demoing Interlisp (they’ve done Cedar and Smalltalk)

but I was arguing for a separate demo of Notecards as an important step in the evolution
of Hypertext, for example.  

 

Besides, it’s fun to see things running 1000 times faster than 30 years ago.

 

Check it out https://interlisp.org

 

--

https://LarryMasinter.net

 

Arun Welch

unread,
May 5, 2021, 10:44:39 PM5/5/21
to fr...@halasz.org, Larry Masinter, Interlisp core
Frank,
 I’ve been able to get Notecards to compile in the existing environment, but it was failing on loading some of my old Notefiles and I didn’t get to digging into what it was attempting to do. I’m happy to schedule some time with someone who knows more about the underlying code to tell me where to look.

Arun

-- 
You received this message because you are subscribed to the Google Groups "Medley Interlisp core" group.
To unsubscribe from this group and stop receiving emails from it, send an email to lispcore+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/lispcore/02a001d741f9%246ac3adb0%24404b0910%24%40acm.org.

Frank Halasz

unread,
May 11, 2021, 9:02:01 PM5/11/21
to Arun Welch, Larry Masinter, Interlisp core
There were a couple releases of NoteCards that involved changes to the NoteFile database format.  I can't recall the details, but I am trying to refamiliarize myself with the code so I can figure it out. 

What NoteCards source code did you use to recompile the system under the current environment?  

I have been trying to compile using the source code stored in the envos repository on github.com/interlisp.   None of the three versions included there will compile cleanly due to corrupted files.  I tried to trade source files between versions but there is (at least) one file (system/NCLINKS) that is corrupt and is identical (and identically corrupted) in all three versions.   Looks like the NCLINKS file(s) got overwritten in a couple places in the middle by blocks taken from NCCONVERTVERSION2TO3 and NCCONVERTVERSION2TO3.LCOM.

Similar corruption also appears in other files - e.g., envos/RELEASE/notecards/1.2/src/system/NCDECLS.  The same corruption also appears in ...2.0/src/system/NCDECLS.  But not in envos/LISPCORE/notecards/system/NCDECLS.

I will be able to progress much faster if I could get clean copies of the NoteCards sources (for all versions).

-- Frank




Arun Welch

unread,
May 11, 2021, 9:09:12 PM5/11/21
to Frank Halasz, Larry Masinter, Interlisp core
Hmm. I’m not sure which set of files I used, and I’m nervous about putting yet another possibly-not-working set in the Github. Let me put together a tarfile of what I have and share it with you.

Larry Masinter

unread,
May 12, 2021, 1:04:13 AM5/12/21
to Arun Welch, Frank Halasz, Interlisp core

When you said you were looking through the Envos files, I should have noted (I’ll add it to the README in that repo and the copy in the History folder)

That the files as we received them seemed to have gone through a tr ‘/n/r’ ‘/r/n’ pipeline, switching CR 13 with LF 10.

Not just for text files. The symptom was TEdit files breaking because they couldn’t find a font size 13.

greg christie

unread,
Jun 3, 2021, 6:46:12 PM6/3/21
to Medley Interlisp core
Hello all,

This thread really caught my eye - I'd love to run Notecards on your modern Interlisp implementation.

I don't know if this could possibly help, but last night i was able to stitch together the Xerox 1108 Medley Notecards lisp.sysout archived on Bitsavers and export it to a UNIX filesystem. I was also able to verify it ran by doing a SysIn and Boot using Josh Dersch's Darkstar 1108 /Star emulator.

It seems to run and work fine.

Please let me know if this could help your current efforts.

Greg

Nick Briggs

unread,
Jun 3, 2021, 6:47:23 PM6/3/21
to greg christie, Medley Interlisp core
Really a Medley one?   If you can make it available somewhere, I'll check it on the current Medley emulator.

-- Nick

greg christie

unread,
Jun 3, 2021, 7:32:27 PM6/3/21
to Medley Interlisp core
Hi, yes.

It is the version for the "1186" version of the Xerox 6085/Daybreak machine. Archived here: http://bitsavers.org/bits/Xerox/1186/Medley_Notecards_1.1.zip  The tar'ed and zipped lisp.sysout is only 3mb - I can email it to whoever...

I have tried running the latest release on github of your Medley with the sysout and get the following error:

sysout_loader:IFPAGE says sysout size is 0
But, sysout size from UNIX is 14396

I get similar errors when I use the old DOS or linux Medley emulators.

I'm guessing there is a significant format difference from this ancient Xerox/Envos sysout to your emulation target? Perhaps
conversion is possible? I'm ignorant of pretty much all of this .

G

Larry Masinter

unread,
Jun 3, 2021, 8:01:30 PM6/3/21
to greg christie, Medley Interlisp core, Frank Halasz

Frank, any luck with finding a good version?

 

Interlisp/envos: Repository for holding files migrated from 1993 Envos (github.com) .

There’s also a copy in https://github.com/interlisp/history under 1990s/1993-envos RELEASE/notecards and LISPCOE/notecards.

 

 

The compiled files are for Medley 1.2 and 2.0; while we the files in github/interlisp/medley github/interlisp/maiko are for Medley 3.5, with 4 more bits of address space, with the cost that you need to recompile and reload the sysouts; there isn’t a conversion.

 

It might be possible to make an emulator for medley 2.0 but I don’t think we’ve tried.

Nick Briggs

unread,
Jun 3, 2021, 8:05:02 PM6/3/21
to greg christie, Medley Interlisp core
On Jun 3, 2021, at 4:32 PM, greg christie <californi...@gmail.com> wrote:

Hi, yes.

It is the version for the "1186" version of the Xerox 6085/Daybreak machine. Archived here: http://bitsavers.org/bits/Xerox/1186/Medley_Notecards_1.1.zip  The tar'ed and zipped lisp.sysout is only 3mb - I can email it to whoever...

I've recreated it from the DMKs in the zip, so no need to mail it.


I have tried running the latest release on github of your Medley with the sysout and get the following error:

sysout_loader:IFPAGE says sysout size is 0
But, sysout size from UNIX is 14396

I think I can bypass that check, and it might be OK.


I get similar errors when I use the old DOS or linux Medley emulators.

We're an updated revision of the DOS/Linux Medley emulator that you've tried --most of the sources going back to 1988 or so, but things have been cleaned up, hand optimizations removed, support for some older OSs removed, etc.


I'm guessing there is a significant format difference from this ancient Xerox/Envos sysout to your emulation target? Perhaps
conversion is possible? I'm ignorant of pretty much all of this

No format difference.  We run the same sysouts as the D-machines did -- though Larry recently removed the code that placed the Dandelion/Daybreak I/O processor code into the image when you bootstrap a new sysout.

greg christie

unread,
Jun 3, 2021, 8:19:47 PM6/3/21
to Medley Interlisp core

Interlisp/envos: Repository for holding files migrated from 1993 Envos (github.com) .

There’s also a copy in https://github.com/interlisp/history under 1990s/1993-envos RELEASE/notecards and LISPCOE/notecards.

 


I've tried those sysout.Z files. Invariably  I get a corrupt input error when using UNIX uncompress on those. Earlier you mentioned all of that stuff when through a tr pipeline, presumably to "correct" linefeeds? I triued a couple of simple 0A/0D translations and always ended up with a corrupt .Z file.

thanks,

Greg

Nick Briggs

unread,
Jun 3, 2021, 8:24:46 PM6/3/21
to greg christie, Medley Interlisp core
These old ones require the medley 2.0 emulator rather than the 3.5.1 version that we're running.   As Larry mentioned, we haven't built a 2.0 version in the current round of hacking.  The NoteCards sysout in the RELEASE/notecards reports

Lisp VM is too old for this emulator.
(version is 4928, must be at least 35010.)

-- Nick

--
You received this message because you are subscribed to the Google Groups "Medley Interlisp core" group.
To unsubscribe from this group and stop receiving emails from it, send an email to lispcore+u...@googlegroups.com.

greg christie

unread,
Jun 3, 2021, 8:26:44 PM6/3/21
to Medley Interlisp core
On Thursday, June 3, 2021 at 5:05:02 PM UTC-7 nicholas...@gmail.com wrote:


I've recreated it from the DMKs in the zip, so no need to mail it.

Okay, just let me know if you need it to compare. FWIW,  I did repeated floppy read lisp.sysout operations in XDE for each floppy, pushed it to my XNS file service, exported it to the UNIX host from there, and then from Darkstar booted the supplied medley disk image, exited to the system tool and did a Sysin and Boot! from the XNS file service it al worked and i snapshotted the disk image right after boot.
 


I have tried running the latest release on github of your Medley with the sysout and get the following error:

sysout_loader:IFPAGE says sysout size is 0
But, sysout size from UNIX is 14396

I think I can bypass that check, and it might be OK.

That would be awesome.

best,

Greg

Nick Briggs

unread,
Jun 3, 2021, 8:47:03 PM6/3/21
to greg christie, Medley Interlisp core
On Jun 3, 2021, at 5:26 PM, greg christie <californi...@gmail.com> wrote:



On Thursday, June 3, 2021 at 5:05:02 PM UTC-7 nicholas...@gmail.com wrote:


I've recreated it from the DMKs in the zip, so no need to mail it.

Okay, just let me know if you need it to compare. FWIW,  I did repeated floppy read lisp.sysout operations in XDE for each floppy, pushed it to my XNS file service, exported it to the UNIX host from there, and then from Darkstar booted the supplied medley disk image, exited to the system tool and did a Sysin and Boot! from the XNS file service it al worked and i snapshotted the disk image right after boot.


OK then!  I used the dmklib-0.31 dmk2raw to extract each raw blob, cat'ed them together and used Darrel VanBuer's extractXeroxFloppy to pull the "lisp.sysout" out of the image.

Disk Lisp Sysout #1
 Length Created               Filename
7370752 Nov 19 08:50:50 1990  lisp.sysout
2 files

% shasum lisp.sysout 
e6facf729c9498d02f9e5ae6c243585d51c45fbe  lisp.sysout

Do we match (that's a SHA1 hash).
 


I have tried running the latest release on github of your Medley with the sysout and get the following error:

sysout_loader:IFPAGE says sysout size is 0
But, sysout size from UNIX is 14396

I think I can bypass that check, and it might be OK.

That would be awesome.

I've built a 1.1.5 version emulator and a 2.0 emulator, but neither is happy in its current state with that sysout.   I'm going to be away from my keyboard for about a week, but I'll add it to the list of things I want to look at.

I'm pretty sure that we *do* have a path to an uncorrupted version of the notecards sources, so it should be possible to get it running on the current 3.5.1 emulator.


best,

Greg


--
You received this message because you are subscribed to the Google Groups "Medley Interlisp core" group.
To unsubscribe from this group and stop receiving emails from it, send an email to lispcore+u...@googlegroups.com.

Nick Briggs

unread,
Jun 3, 2021, 9:00:36 PM6/3/21
to greg christie, Medley Interlisp core
On Jun 3, 2021, at 5:47 PM, Nick Briggs <nicholas...@gmail.com> wrote:


On Jun 3, 2021, at 5:26 PM, greg christie <californi...@gmail.com> wrote:



On Thursday, June 3, 2021 at 5:05:02 PM UTC-7 nicholas...@gmail.com wrote:


I've recreated it from the DMKs in the zip, so no need to mail it.

Okay, just let me know if you need it to compare. FWIW,  I did repeated floppy read lisp.sysout operations in XDE for each floppy, pushed it to my XNS file service, exported it to the UNIX host from there, and then from Darkstar booted the supplied medley disk image, exited to the system tool and did a Sysin and Boot! from the XNS file service it al worked and i snapshotted the disk image right after boot.


OK then!  I used the dmklib-0.31 dmk2raw to extract each raw blob, cat'ed them together and used Darrel VanBuer's extractXeroxFloppy to pull the "lisp.sysout" out of the image.

Disk Lisp Sysout #1
 Length Created               Filename
7370752 Nov 19 08:50:50 1990  lisp.sysout
2 files

% shasum lisp.sysout 
e6facf729c9498d02f9e5ae6c243585d51c45fbe  lisp.sysout

Do we match (that's a SHA1 hash).

Duh, the extractXeroxFloppy doesn't work on the concatenated files -- did it again properly and now its:

a6d95bb5b3ef480afefcaf7c411ac8c762aae340  lisp.sysout

So a 1.1.5 emulator starts up and gets as far as *almost* getting the screen image up:



ldex /tmp/Medley_Notecards_1.1/lisp.sysout 
Entering FindUnixPipes
NPROCS is 256; leaving FindUnixPipes
Failed to find UNIXCOMM file handles; no processes
VM allocated = 0x4000000 at 0x4b1000
Native Load Address = 0x0
FPTOVP Location 1023 
sysout size / 2 = 0x383c
vmem size = 0x383c
FBIOGTYPE w x h = 1440 x 856
FBIOGTYPE w x h = 1440 x 856
Display address: 0x6f1000
        length : 0x26000
        pg size: 0x1000
after clear_display()
exiting init_display
sysout file is read completely.
Stackspace = 0x4d1000.
AtomHT = 0x591000.

[... X keyboard setup junk ...]

Timer interval set to 25000 usec
I/O interrupts enabled
File timeout interrupts enabled
Panic interrupts enabled
contextsw:ESTKP set LADDR : 0x10600 (67072)
context:after:nopush 
Subr call to subr 0103.
contextsw:ESTKP set LADDR : 0x176fe (95998)
context:after:nopush 
Subr call to subr 0103.
Subr call to subr 061.
Subr call to subr 0103.
Subr call to subr 0122.
CELL 0x2f16e6 has non-NIL car = 0x3cf5ee 
*Error* QUIT from N_OP_cons
Enter the URaid
CL:NIL

and then we're dead.


greg christie

unread,
Jun 3, 2021, 9:21:14 PM6/3/21
to Medley Interlisp core
Yes, your second checksum matches my file. Your dmk-foo is better than mine.... I can't remember whether it was dmk2raw or imd2raw with extractxeroxfloppy but I had real trouble dropping the second side of some Xerox floppy images in the past for another project...

So interestingly enough on first boot of the Darkstar emulator with this sysout, the initial screen image was mixed up as in your screen shot, then it drew everything normally. As long as i save the image and logout, I can go back into it. Flushing the image renders the Darkstar disk image unbootable.

Again. I'm ignorant of how these sysouts work (my interest in xerox hardware & software revolves around XDE, ViewPoint/GlobalView and Smalltalk) but i notice that the framebuffer dimensions are in your printout. Could the mismatch of the framebuffer dimensions of your emu and that of an 1108 or 6085/1186 interfere with proper loading?

Nick Briggs

unread,
Jun 3, 2021, 9:33:56 PM6/3/21
to greg christie, Medley Interlisp core
On Jun 3, 2021, at 6:21 PM, greg christie <californi...@gmail.com> wrote:

Yes, your second checksum matches my file. Your dmk-foo is better than mine.... I can't remember whether it was dmk2raw or imd2raw with extractxeroxfloppy but I had real trouble dropping the second side of some Xerox floppy images in the past for another project...

Hmm.. don't know that I've ever had to deal with something that I knew was a 2-sided floppy image, in this case it just worked for me.

So interestingly enough on first boot of the Darkstar emulator with this sysout, the initial screen image was mixed up as in your screen shot, then it drew everything normally. As long as i save the image and logout, I can go back into it. Flushing the image renders the Darkstar disk image unbootable.

Again. I'm ignorant of how these sysouts work (my interest in xerox hardware & software revolves around XDE, ViewPoint/GlobalView and Smalltalk) but i notice that the framebuffer dimensions are in your printout. Could the mismatch of the framebuffer dimensions of your emu and that of an 1108 or 6085/1186 interfere with proper loading?

The sysout has the address space allocated for the largest possible display area, and the emulator just loads all the memory that's there (and zeroes otherwise) into the emulated display without regard for how the old memory maps to the new dimensions (it certainly could be smarter about that! but it would have to inspect the values of the atoms where it was stored) and then it forces a redraw so everything settles in to the new screen geometry.   You can get the same effect on a current working sysout by telling the emulator to change the emulated screen geometry.

What we're seeing here, I think, is an error out of the code that handled the *not* NEWCDRCODING case in the "allocate a new cons page" -- a code path which hasn't been executed in a good, oh, 30 years?

I think it would be useful to have the old versions of the emulator work, so this will remain on my list.

Larry Masinter

unread,
Jun 3, 2021, 11:35:51 PM6/3/21
to greg christie, Medley Interlisp core

Did you try

   tr ‘\r\n’ ‘\n\r’

 

by any chance?

From: lisp...@googlegroups.com <lisp...@googlegroups.com> On Behalf Of greg christie
Sent: Thursday, June 3, 2021 5:20 PM
To: Medley Interlisp core <lisp...@googlegroups.com>
Subject: Re: Putting together a Notecards demo

 

--

You received this message because you are subscribed to the Google Groups "Medley Interlisp core" group.
To unsubscribe from this group and stop receiving emails from it, send an email to lispcore+u...@googlegroups.com.

greg christie

unread,
Jun 4, 2021, 12:12:55 AM6/4/21
to Medley Interlisp core
Hi,

You know, i thought I tried:

   tr ‘\r\n’ ‘\n\r’

and still had corrupted archives but I just tried it on one of them and it worked.

of course I still get the

sysout_loader:IFPAGE says sysout size is 0
But, sysout size from UNIX is 14396

error trying to start the  new medley with that sysout, same error as the sysout I constructed.

Oh well, worth a try...

Greg

greg christie

unread,
Jun 4, 2021, 12:20:14 AM6/4/21
to Medley Interlisp core
Okay thanks, I'll keep an eye on this thread - at least you have a confirmed checksum for that sysout and it DOES work on a D-machine...

best,

Greg

Larry Masinter

unread,
Jun 4, 2021, 6:13:03 PM6/4/21
to greg christie, Medley Interlisp core, Guenther Goerz, Josh Dersch, Frank Halasz

I put the agenda topic “Notecards demo”

Meeting Agendas and Notes · Interlisp/medley Wiki (github.com)

 

At the start 10:30 am pacific time Monday June 7

 

 

https://zoom.us/j/4083324290?pwd=RXlFeFJ3NHRRQmZvQXFkMzltTlJMQT09

 

 

Nick Briggs

unread,
Jun 5, 2021, 1:19:08 AM6/5/21
to Larry Masinter, greg christie, Medley Interlisp core, Guenther Goerz, Josh Dersch, Frank Halasz
I'm afraid I will be out of town this Monday.   Hope the discussion goes well, though.

-- Nick

-- 
You received this message because you are subscribed to the Google Groups "Medley Interlisp core" group.
To unsubscribe from this group and stop receiving emails from it, send an email to lispcore+u...@googlegroups.com.

Ron Kaplan

unread,
Jun 5, 2021, 1:49:44 AM6/5/21
to Nick Briggs, Larry Masinter, greg christie, Medley Interlisp core, Guenther Goerz, Josh Dersch, Frank Halasz
I also might not make it this week—I’m giving a colloquium at Oxford and the end-time is a bit uncertain.

greg christie

unread,
Jun 5, 2021, 5:41:12 AM6/5/21
to Medley Interlisp core
Hi,

Just to close the circle a bit on some of this for sake of future searchers...

If you apply the tr ‘\r\n’ ‘\n\r’ transformation as Larry specifies on the sysouts in those two envos archive repositories you end up with de-corrupted files, able to be uncompressed without error. While they won't load into the current work or the old Linux medley emu, the 2.0 notecards sysouts WILL load and function fine on the old DOS emu, as long as MEDLEY.EXE is started with the -noeurokbd switch. And all of this runs perfectly fine under DosBox.

Greg

Frank Halasz

unread,
Jun 6, 2021, 10:04:02 PM6/6/21
to Larry Masinter, greg christie, Medley Interlisp core, Guenther Goerz, Josh Dersch
I will attend.

-- Frank

Frank Halasz

unread,
Jun 8, 2021, 1:05:39 AM6/8/21
to Larry Masinter, greg christie, Medley Interlisp core
I am able to cleanly compile the version that Arun Welch provided to me.  But I have not been able to get anything substantial to run.   Its a lot of little issues one after the other.  Not sure that Arun's version is fully Medley compatible and I am not convinced yet that all of the things NoteCrads depends on are there.   I haven't had a lot of time to work on this, but I have been solving the issues one by one as they come up.  Will try to put more time into this in the coming weeks.

-- Frank

Paul McJones

unread,
Jun 8, 2021, 6:47:55 PM6/8/21
to Frank Halasz, Larry Masinter, greg christie, Medley Interlisp core
Off the subject but of potential interest: while going through an old
file cabinet, I ran across my notes from an 1988 talk by Randy Trigg on
Notecards (see attached PDF). I was at DEC SRC at the time -- I don't
remember if Randy gave the talk there, or at some other venue.


Paul

Trigg-Notecards-1988.pdf

Larry Masinter

unread,
Jul 13, 2021, 8:38:53 PM7/13/21
to Paul McJones, Frank Halasz, greg christie, Medley Interlisp core
I'm not sure where this would belong;
In Zotero's Interlisp bibliography under "Systems Implemented in Interlisp"? As a Note or an associated file?

Let me know if you need write access.

I'm less sure about the structure of the Zotero library. Zotero supports individual and group "Libraries".
I have a Library (and if you make a Zotero account, you'll have one too.)
I made a Group called Interlisp and invited lispcore members to join. Please let me know if you need another invite.
I made a Interlisp Group Library.
In each Library there are items. In each Library there are Collections and SubCollections.
Interlisp has 6 explicit collections, one of which is "Interlisp.org Bibliogrphy".
Under that, there are 3 sub-collections "About Interlisp-D", "About Interlisp", "Applications built in Interlisp"
Which seemed like the three categories which we might want to feature.
The other categories are the collections "About Lisp" "Common Lisp", "mentions/discusses Interlisp", "Not relevant to Medley Interlisp", SK

I think the "Not relevant" collection are "Not sure how these got in"
> -----Original Message-----
> From: lisp...@googlegroups.com <lisp...@googlegroups.com> On Behalf
> Of Paul McJones
> Sent: Tuesday, June 8, 2021 3:48 PM
> To: Frank Halasz <fr...@halasz.org>; Larry Masinter <L...@acm.org>
> Cc: greg christie <californi...@gmail.com>; Medley Interlisp core
> <lisp...@googlegroups.com>
> Subject: Re: Putting together a Notecards demo
>
> --
> You received this message because you are subscribed to the Google Groups
> "Medley Interlisp core" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to lispcore+u...@googlegroups.com.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/lispcore/494dba49-aaf1-bc77-8f01-
> 2d87003483fc%40mcjones.org.

Reply all
Reply to author
Forward
0 new messages