What is lispusers/TEDITTALK supposed to do?

[rmkaplan]

There is no documentation, and it seems to muck around with Tedit in funky ways. It advises a bunch of Tedit functions to do some sort of printout of internal structures some of which don't make sense or no longer exist.

It may be that that advice is just for debugging purposes, and can simply be stripped out. But unless this is providing a useful capability, the best thing might be to move it to obsolete.

[nbriggs]

It's documented in lispusers>TALK.TEDIT as part of the TALK facility

[rmkaplan]

Ahh. Maybe all the lispusers/*TALK* files should be moved into a lispusers/talk/ subdirectory (with some renaming according to our hyphen convention: TEDITTALK --> TALK-TEDIT). They would all be together with the documentation and the main driver.

…

On Apr 5, 2025, at 11:19 AM, Nick Briggs ***@***.***> wrote: nbriggs left a comment (Interlisp/medley#2095) It's documented in lispusers>TALK.TEDIT as part of the TALK facility — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread. <#2095 (comment)> <https://github.com/notifications/unsubscribe-auth/AQSTUJPUZ7BQSSXCX27W56T2YANEJAVCNFSM6AAAAAB2QXXLF6VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDOOBRGAZDQMRVGE> nbriggs left a comment (Interlisp/medley#2095) <#2095 (comment)> It's documented in lispusers>TALK.TEDIT as part of the TALK facility — Reply to this email directly, view it on GitHub <#2095 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AQSTUJPUZ7BQSSXCX27W56T2YANEJAVCNFSM6AAAAAB2QXXLF6VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDOOBRGAZDQMRVGE>. You are receiving this because you authored the thread.

[nbriggs]

Perhaps -- and IPTALK and NSTALK and NSTALKGAP should be brought back from obsolete/lispusers/ -- but we should check which is more recent between obsolete/lispusers/NSTALKGAP and TALKGAP

[rmkaplan]

TALK actually doesn't compile. There are a number of functions with comments of the form (* DECLARATIONS%: [some record declaration]), and those record declarations are not recognized. I don't remember ever trying to define a lexically scoped record declaration. Is that something that used to work--or should work now?

…

On Apr 5, 2025, at 1:00 PM, Nick Briggs ***@***.***> wrote: nbriggs left a comment (Interlisp/medley#2095) Perhaps -- and IPTALK and NSTALK and NSTALKGAP should be brought back from obsolete/lispusers/ -- but we should check which is more recent between obsolete/lispusers/NSTALKGAP and TALKGAP — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread. <#2095 (comment)> <https://github.com/notifications/unsubscribe-auth/AQSTUJIM6ENU7LDYGOL3KTL2YAY6FAVCNFSM6AAAAAB2QXXLF6VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDOOBRGA3DINZUGU> nbriggs left a comment (Interlisp/medley#2095) <#2095 (comment)> Perhaps -- and IPTALK and NSTALK and NSTALKGAP should be brought back from obsolete/lispusers/ -- but we should check which is more recent between obsolete/lispusers/NSTALKGAP and TALKGAP — Reply to this email directly, view it on GitHub <#2095 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AQSTUJIM6ENU7LDYGOL3KTL2YAY6FAVCNFSM6AAAAAB2QXXLF6VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDOOBRGA3DINZUGU>. You are receiving this because you authored the thread.

[nbriggs]

There are quite a few files with "DECLARATIONS%:" in them, include "DECL" which you were working on in 1983. I don't know anything about it, though.

[rmkaplan]

Maybe something that used to work and was deprecated. But looking at little bit more at the TALK-related files: TALKGAP doesn't load because loading calls the undefined function DEFINE.GAP.SERVER. I can probably get TALK to compile by moving the delcarations out of the functions and into the file. But it might still be tricky to get it to work. As for TEDITTALK, it is advising a number of Tedit functions (mostly public) so that they call a TALK function before they execute. The TALK function encodes the command (insert, move selection, copy...) and sends it with a function-character over the channel so that the remote Tedit can then execute the same command in the same way. On the other end, there would be a charbinding that causes the command to be decoded and executed. The effect is that both sides see the changes as they are happening. It's a clever but heavy-handed way of getting this behavior. One of the pieces of advice is to CHANGENAME OPENFILE in TEDIT.INCLUDE...but OPENFILE has been banished from the system. I could set it up so that it doesn't look for any user-defined functions (even the ones that are currently built in) unless a trigger bit is set on the text object. So it wouldn't scan for special behavior on every operation for ordinary streams. But if the rest of the scheme can't be made to work, or is not worth the effort, then I think we should wrap all the talk files into a subdirectory, and make them all obsolete.

…

On Apr 6, 2025, at 10:08 AM, Nick Briggs ***@***.***> wrote: nbriggs left a comment (Interlisp/medley#2095) There are quite a few files with "DECLARATIONS%:" in them, include "DECL" which you were working on in 1983. I don't know anything about it, though. — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread. <#2095 (comment)> <https://github.com/notifications/unsubscribe-auth/AQSTUJINITNRLZN66LLJ6332YFNP5AVCNFSM6AAAAAB2QXXLF6VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDOOBRGUYTKMZZGY> nbriggs left a comment (Interlisp/medley#2095) <#2095 (comment)> There are quite a few files with "DECLARATIONS%:" in them, include "DECL" which you were working on in 1983. I don't know anything about it, though. — Reply to this email directly, view it on GitHub <#2095 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AQSTUJINITNRLZN66LLJ6332YFNP5AVCNFSM6AAAAAB2QXXLF6VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDOOBRGUYTKMZZGY>. You are receiving this because you authored the thread.

[nbriggs]

Shared document editing long before Google docs... but only for a pair collaborators.

[MattHeffron]

Resolved by PRs #2147 & #2218