home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.lang.tcl
- Path: sparky!uunet!think.com!linus!linus.mitre.org!news!mbunix!blk
- From: blk@mitre.org (Brian L. Kahn)
- Subject: Re: Tcl with emacs or epoch
- In-Reply-To: blk@mitre.org's message of Sun, 8 Nov 1992 21:30:51 GMT
- Message-ID: <BLK.92Nov10114311@vanity.mitre.org>
- Sender: news@linus.mitre.org (News Service)
- Nntp-Posting-Host: vanity.mitre.org
- Organization: The MITRE Corp. Bedford, MA
- References: <BLK.92Nov8163051@vanity.mitre.org>
- Date: Tue, 10 Nov 1992 16:43:11 GMT
- Lines: 50
-
-
- In article <BLK.92Nov8163051@vanity.mitre.org> blk@mitre.org (Brian L. Kahn) writes:
-
- Has anyone produced patches to incorporate tk/tcl into emacs or epoch?
- I looked in the FAQ and docs, but couldn't find a reference.
-
-
-
- Here's some email that I thought might be of general interest:
-
- To: lvirden@cas.org (Larry W. Virden)
- Subject: Re: Tcl with emacs or epoch
- In-reply-to: Your message of "Mon, 09 Nov 92 07:41:30 EST."
- <9211091242.AA06062@cas.org>
- Date: Tue, 10 Nov 92 11:16:40 EST
- From: blk
-
-
- >> there is a tk widget which uses emacs bindings, but so far, few folk
- >> have contributed ANY existing programs modified to incorporate tk/tcl/tclx.
-
- I don't quite get what you're saying here, do you mean there is a set
- of emacs lisp bindings for all the tk functions? I don't think that's
- what you mean.
-
- >> Emacs would particularly be a bear since it has its own interpreter so
- >> firmly embedded.
-
- Two approaches are possible: the Tk library can be bound to elisp
- functions, or tcl can be a subordinate to the elisp interpretter.
-
- I thought it would be fairly easy to do the following:
- add C code to init a tcl interpretter
- add an emacs function (in C) that gives a string to the interp
-
- The async input generated by the tcl interp is more of a problem. The
- only async mechanisms I know in emacs are process sentinals, which I
- suppose could be cobbled up to this internal, async process. I use
- epoch, so I would much prefer to get my input from property changes on
- the relevant windows. This wouldn't be the default mechanism, but
- rather something that is settable via tcl.
-
- This approach drops tcl out of the main loop, which probably makes the
- "send" function inconsistent. Can be handled other ways, I suppose.
- Sends can still work to the epoch implementation by setting properties
- on the emacs windows instead of on the root window, which will be
- executed by the tcl interp.
- --
- B< Brian Kahn blk@mitre.org "may the farce be with you"
-
-