2:28 p.m.
(copying wmperry who is a xemacs hacker)
Hi Steve --
We met a few years ago in Mountain View along with Bob
Weiner and co.
I just pulled and installed xemacs 21.2.27 --along with
xemacs-sump.tar.gz --and got emacspeak 11.0 built against
it.
Reason I'm writing is that for the next emacspeak release
(May 2000) I'd like to make emacspeak into a xemacs package.
There are a few significant differences between FSF and
Xemacs that I need to handle in emacspeak --which is why I'm
writing to you.
Incidentally, emacspeak now works using the software speech
synthesizer from IBM (Viavoice Outloud)--this is currently
free to download off the WWW --so you or other xemacs
hackers should now be able to experience emacspeak
first-hand on Linux.
Here are some of the questions I have on making everything
work in xemacs --
1) what happened to the *Messages* buffer in xemacs?
2) inhibit-point-motion-hooks appears to be missing --what
is the xemacs equivalent?
Going forward, do you think you'll have time to play with
emacspeak on xemacs --alternatively know anyone in the
xemacs hacker community who might be interested?
--
Best Regards,
--raman
Email: raman(a)cs.cornell.edu
WWW: http://cs.cornell.edu/home/raman/
PGP: http://cs.cornell.edu/home/raman/raman.asc
10:13 p.m.
>>>> "TVR" == T V Raman
<ramantv(a)earthlink.net> writes:
TVR> (copying wmperry who is a xemacs hacker)
TVR> Hi Steve --
TVR> We met a few years ago in Mountain View along with Bob
TVR> Weiner and co.
TVR> I just pulled and installed xemacs 21.2.27 --along with
TVR> xemacs-sump.tar.gz --and got emacspeak 11.0 built against
TVR> it.
TVR> Reason I'm writing is that for the next emacspeak release
TVR> (May 2000) I'd like to make emacspeak into a xemacs package.
Yes! It should be a package within the package system. Raman, I'd
like to give you CVS commit access to at least the part of the CVS
tree that is to contain Emacspeak - more if you think you will be
contributing to XEmacs. Please follow the instructions at
http://cvs.xemacs.org/#committers. Then you can maintain your own
package as you wish.
TVR> There are a few significant differences between FSF and
TVR> Xemacs that I need to handle in emacspeak --which is why I'm
TVR> writing to you.
TVR> Incidentally, emacspeak now works using the software speech
TVR> synthesizer from IBM (Viavoice Outloud)--this is currently
TVR> free to download off the WWW --so you or other xemacs
TVR> hackers should now be able to experience emacspeak
TVR> first-hand on Linux.
This is something I would like to do.
TVR> Here are some of the questions I have on making everything
TVR> work in xemacs --
TVR> 1) what happened to the *Messages* buffer in xemacs?
The *Messages* buffer is hidden by being named " *Message-Log*". I'm
not sure if that was a mistake.
The user interface is C-h l .
That displays recent keystrokes as well, which I do consider a user
interface bug. There should be two entries in help-map for the two
different kinds of information.
Martin
2:56 a.m.
Martin --for now I probably wont need the cvs commit access
--but thanks for offering it.
I may well take you up on it down the road.
Drop me a line soon as you have the outloud drop
working. You may need an emacspeak snapshot from me to run
under xemacs since I had to make a few mods.
--
Best Regards,
--raman
Email: raman(a)cs.cornell.edu
WWW: http://cs.cornell.edu/home/raman/
PGP: http://cs.cornell.edu/home/raman/raman.asc
4:06 p.m.
Martin --
1) Yes, I think it would be a good idea to separate the
lossage and messages buffers and provide a keystrok for
popping up the messages buffer.
2) The xemacs compiler appears to complain incorrectly
sometimes about special variables --this is with
(require 'cl) and
(declare (special ...))
inside function bodies.
It also appears to have problems the cl (declaim ...)
construct --it continues to produce the warning that that
declaration is supposed to suppress.
--need to create a minimal .el file that demonstrates this.
3) I guess inhibit-point-motion-hooks got dropped in xemacs.
Not that I care --emacspeak doesn't use this -however in the
case of FSF emacs
emacspeak carefully sets this variable at specific times in
order to avoid emacs doing unexpected things.
I've worked around it for now by introducing an inline
function that I call rather than directly referring to
variable inhibit-point-motion-hooks
--is there more standard fsf-compat way of doing this?
4) Bill (wmperry)-- emacspeak-w3 throws a strange warning
when loaded
--the error is being thrown by the bytecompiler saying
invalid bytecode --this is after comiling all emacspeak
files with xemacs.
--the problem appears to be in the section of the code where
we insert the emacspeak help into the widgets via symbol
emacspeak-help
--
Best Regards,
--raman
Email: raman(a)cs.cornell.edu
WWW: http://cs.cornell.edu/home/raman/
PGP: http://cs.cornell.edu/home/raman/raman.asc
7:54 p.m.
>>>> "TVR" == T V Raman
<ramantv(a)earthlink.net> writes:
TVR> 2) The xemacs compiler appears to complain incorrectly
TVR> sometimes about special variables --this is with
TVR> (require 'cl) and
TVR> (declare (special ...))
TVR> inside function bodies.
I guess I'm the byte compiler maintainer, so I should add these
things. They would be useful to have anyways, for other packages.
What we do in XEmacs is simply use (defvar special-var). That form
doesn't have to be at top-level. Emacs Lisp doesn't really have
non-special vars anyways.
TVR> It also appears to have problems the cl (declaim ...)
TVR> construct --it continues to produce the warning that that
TVR> declaration is supposed to suppress.
TVR> --need to create a minimal .el file that demonstrates this.
Maybe I should try to byte-compile emacspeak. I should get your
latest development version.
TVR> 3) I guess inhibit-point-motion-hooks got dropped in xemacs.
TVR> Not that I care --emacspeak doesn't use this -however in the
TVR> case of FSF emacs
TVR> emacspeak carefully sets this variable at specific times in
TVR> order to avoid emacs doing unexpected things.
TVR> I've worked around it for now by introducing an inline
TVR> function that I call rather than directly referring to
TVR> variable inhibit-point-motion-hooks
TVR> --is there more standard fsf-compat way of doing this?
TVR> 4) Bill (wmperry)-- emacspeak-w3 throws a strange warning
TVR> when loaded
TVR> --the error is being thrown by the bytecompiler saying
TVR> invalid bytecode --this is after comiling all emacspeak
TVR> files with xemacs.
TVR> --the problem appears to be in the section of the code where
TVR> we insert the emacspeak help into the widgets via symbol
TVR> emacspeak-help
I've rewritten the bytecode loader, and there are now more run-time
sanity checks on bytecode. There might be a bug in my code, or in
your code. One way to get such an error is to do:
(setq t 1)
Unfortunately, my time is currently overcommitted to actually do much
work on emacspeak itself. But I'll help to make emacspeak work with
XEmacs.
Martin
9054
days inactive
9056
days old
4 comments
2 participants
participants (2)
-
Martin Buchholz -
T. V. Raman