August 1998 - XEmacs-Beta - XEmacs Mailing List Archives

Re: Macros: Scheme vs Common Lisp 25 years, 10 months

Clemens Heitzinger

sperber(a)informatik.uni-tuebingen.de (Michael Sperber [Mr. Preprocessor]) writes: > Clemens> What do you mean, "syntactic sugar"? Are macros syntactic sugar? > > No, but you're creating syntactic sugar around the explicit use of > MAKE-SYMBOL. You're not describing a new approach, but rather new > notation for an old approach. My intention was neither to describe a new approach, nor to write macros which don't use make-symbol, but to write macros which make it easy to use hygienic variables in CL. > >> Hygiene, however, means that variables in macros obey the same scoping > >> rules as the rest of the language. WITH-HYGIENIC-VARIABLES gives you > >> no help whatsoever referring to lexical bindings of the macro. CL > > Clemens> What help do you mean? > > >> macros are not hygienic, period. > > Clemens> Nobody claimed that. I wanted to make the point that it is quite easy > Clemens> to define hygienic macros in CL. > > And I made the point that what you're talking about is not hygiene. > You still need explicit annotations to prevent variable capture, at > definitions *and* uses. This is tedious and error-prone. You're > doing nothing to respect the binding rules of the underlying language. (I don't understand your last sentence.) Where do I need explicit annotations? At definition time, I say which hygienic variables I want. This is an explicit annotation. On the other hand, I don't think that in Scheme you can say you want this and that variable *not* to be hygienic. In CL you can have both behaviours, in Scheme only one. The usage is certainly the same: (foo0 1) (foo1 1) (foo2 1) Yours, -- Clemens Heitzinger Student der Techn. Mathematik, Technische Universität Wien, Österreich Vienna University of Technology, Austria

1 participants
0 comments

Re: [Richard Stallman <rms@santafe.edu>] Emacs 20.3 25 years, 10 months

Didier Verna

SL Baur <steve(a)xemacs.org> writes: > They won't have all of the features of XEmacs until they have real > scrollbars and toolbars Yeah, they'll be far behind us until they can crash their editor by clicking on a Motif scrollbar !! :-) [ Ben sez we could be very well surprised by their new redisplay code ] So it will soon be time to open the XEmacs 2000 developpement ...

1 participants
0 comments

Re: Macros: Scheme vs Common Lisp 25 years, 10 months

Michael Sperber

>>>>> "Rick" == Rick Campbell <campbell(a)cyberpass.net> writes: Rick> (setq bar 23) Rick> Of course, many consider this sloppy. CMUCL issues a warning at this Rick> point. And so it should. Thanks for pointing this out! Rick> What I want is: Rick> (setq 23) Rick> (defmacro foo () 'bar) Rick> (foo) => 23 Rick> (let ((bar 42)) (foo)) => 23 [note: CL actually yields 42] Rick> (setq bar 57) Rick> (foo) => 57 Rick> (let ((bar 42)) (foo)) => 57 [note: CL actually yields 42] Rick> For what it's worth, CMUCL gives the values that you want, assuming Rick> that you meant (setq bar 23) as the first form rather than (setq 23). But that's probably because CMUCL makes BAR special. CLisp doesn't. So you cannot rely on this behavior. -- Cheers =8-} Chipsy Friede, Völkerverständigung und überhaupt blabla

1 participants
0 comments

Re: Macros: Scheme vs Common Lisp 25 years, 10 months

Michael Sperber

>>>>> "Holger" == Holger Schauer <schauer(a)coling.uni-freiburg.de> writes: >>>>> "MS" == Michael Sperber schrieb am 24 Aug 1998 11:14:43 +0200: MS> What I want is: MS> (setq 23) MS> (defmacro foo () 'bar) MS> (foo) => 23 MS> (let ((bar 42)) (foo)) => 23 [note: CL actually yields 42] MS> (setq bar 57) MS> (foo) => 57 MS> (let ((bar 42)) (foo)) => 57 [note: CL actually yields 42] Holger> But exactly this behaviour is what you get when you use Holger> (defmacro foo () bar) instead of (defmacro foo () 'bar): Ah, now I see what you meant. Unfortunately, this approach works only with self-quoting values for BAR: (setq bar #'(lambda (x) x)) (defmacro foo () bar) (let ((bar #'(lambda (x) (+ x 1)))) (funcall (foo) 23)) => illegal form #<CLOSURE :LAMBDA (X) X> Consequently, what you're doing is really an accidental hack. -- Cheers =8-} Chipsy Friede, Völkerverständigung und überhaupt blabla

1 participants
0 comments

21.2 interacts badly with WM 25 years, 10 months

giacomo boffi

------------------------------------------------test with 20.4 step 1a: % xemacs-20.4 -vanilla & i get one frame step 2a: restart my wm (it is gwm: it undecorates all windows, then reloads init files, finally decorates all windows) step 3a: i have one xemacs frame, the original frame ------------------------------------------------test with 21.2b1 step 1b: % xemacs-21.2b1 -vanilla & i get one frame step 2b: as step 2a above step 3b: i have one xemacs frame, the original frame, AND a second, void frame, taking the whole of my display. when i iconify this larger frame, all of my xemacs frames are iconified, when i kill it, my xemacs dies here i put windows' descriptions for my mail frame and the spurious one Window: Emacs.emacs.emacs:_mail_to_?＠boffi95: Geometry=593x626+180+70 Window: Emacs.xemacs-21_2-b1.xemacs-21_2-b1＠boffi95: Geometry=1002x770+0+0 note that my screen is 1000x750, probably the extra space is courtesy of gwm session management? 21.2 has session management, i cannot remember about 20.4...

1 participants
0 comments

Re: Some of Will Clinger's Publications 25 years, 10 months

Karl M. Hegbloom

>>>>> "Holger" == Holger Schauer <schauer(a)coling.uni-freiburg.de> writes: >>>>> "KMH" == Karl M Hegbloom schrieb am 21 Aug 1998 10:31:55 >>>>> -0700: >>> http://www.ccs.neu.edu/home/will/papers.html KMH> There's a good paper about hygenic macros available here. KMH> It's in .tex... the "acmconf" file that it needs to be KMH> latexed was easy to find using altavista. :-) Holger> Could you provide a PostScript version somewhere so that Holger> not all of us have to find the style-file as well ? I'm not sure that's ok, I don't want to break any copyright laws, or especially, anger Mr. Clinger. I will mail you a copy in private email, rather than publish it for download. I hope that's ok.

1 participants
0 comments

Re: [Richard Stallman <rms@santafe.edu>] Emacs 20.3 25 years, 10 months

SL Baur

They won't have all of the features of XEmacs until they have real scrollbars and toolbars and they can open simultaneous ttys and X windows. Ben, you probably know better than anyone else how long that will take to implement. Martin, your speedup work is important but you're not benchmarking right. You should be benchmarking against FSF Emacs as well as 21.0. If you can get me something in the next two or three weeks, I'll make another 21.2 beta for it.

1 participants
0 comments

Re: [for 20.2] ~user completion 25 years, 10 months

SL Baur

[xemacs-beta cc restored] This is all new ground to me. Please correct me if I got any of this wrong. Greg Klanderman <greg(a)alphatech.com> writes: > Hi Steve, > I'd like to send an update to the patch referred to below reflecting > discussion with Ben and adding documentation. In your 21.0-pre10 > announcement, you mention: >> Note also that the 21.2 CVS tree was synched with these changes and is >> now at 21.2-beta2-pre3. I just performed the checkout command: $ cvs -d :pserver:xemacs＠cvs.xemacs.org:/usr/CVSroot checkout \ -rrelease-21-2 -d XEmacs-21.2 xemacs-20 and got the latest version with the src/dired.c boo-boo. I believe the trick is to use the branch tag `release-21-2' instead of a specific beta tag. Once a tree is checked out, `cvs update -d' should always get the latest and greatest revision on that branch. > and you also say below that my patch is applied. I'm not sure what > tag however is the current 21.2. Doing `cvs log src/dired.c' gives > the following symbolic names containing 21-2: > r21-2b1: 1.18 > release-21-2: 1.18.0.2 > My understanding was that release-21-2 was the split and that r21-2bN > would be betas from there. That is correct. > (Bear with me I've not used branches yet). No problem. I am not comfortable with branches yet either. > Anyway, I would have expected a r21-2b2, but it is not there. I haven't made a 21.2-beta2 yet. The r21-2b2 tag won't be created until the beta is formally made. This is consistent with how I've managed the 21.0 beta series. > Furthermore, r21-2b1 is version 1.18, same as the current 21.0, and > thus does not have my patch applied. That is correct. > Only release-21-2 has the patch applied. Am I just confused, or is > there a problem here? Or have you moved the release-21-2 tag forward > as you synch with 21.0? Hmm maybe that's it. That appears to be a feature. It is consistent. If you don't specify a release tag, you get the latest revision from the default branch. If you specify a branch tag, you get the latest revision from that branch.

4 participants
3 comments

Re: 28 bit ought to be enough for anyone. 25 years, 10 months

SL Baur

Simon Josefsson <jas(a)pdc.kth.se> writes in xemacs-beta(a)xemacs.org: > My point was having a native GSSAPI/Kerberos5 implementation written > in Elisp. I will write a API for the bignum stuff so I can switch > between Calc/GMP/externalprogram/whatever without breaking anything > else, so if calc turns out to be too slow even for developing I could > do something like this. > (Yeah, I know, I will never be finished with this until Kerberos 10 is > out, but it's great hack value and I'll probably learn a lot) Go for it. Crypto stuffs are considered munitions by current USG laws, but if the President doesn't have to abide by the law, neither do we. Let's get this stuff in ASAP and make the world a better place. You have a special O.K. for adding this to what will otherwise be a bugfix 21.1 if you can get it ready fast enough.

1 participants
0 comments

Re: local and networked X connections 25 years, 10 months

Didier Verna

Hrvoje Niksic <hniksic(a)srce.hr> writes: > Didier Verna <verna(a)inf.enst.fr> writes: > > > Chuck or Ben, could you explain to me exactly what represents a > > console (as opposed to a device) under X ? > > I am not sure whether Ben reads thist list (Chuck does not), but I'll > try to answer. If I understand correctly, console is the place where > input is drawn from, which may encompass several devices. For > example, if you use a multi-headed X display, you may have more than > one device under just one console. Yes, that's what I'd understood too, and that's the reason why I don't understand how (find-console ":0") can return nil when I have a connection on "metheny:0". Even if one connection is on a unix socket, and the other on a TCP one, I thought both would share the same console. > However, in 99% cases, one console corresponds to one device. This is > why I think the introduction of consoles was a mistake, because it > totally muddled the otherwise clear and well-thought-out concept of > devices. -- / / _ _ Didier Verna http://www.inf.enst.fr/~verna/ - / / - / / /_/ / E.N.S.T. INF C201.1 mailto:verna＠inf.enst.fr /_/ / /_/ / /__ / 46 rue Barrault Tel. (33) 01 45 81 80 72 75634 Paris cedex 13 Fax. (33) 01 45 81 31 19

1 participants
0 comments

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

2001

2000

1999

1998

XEmacs-Beta August 1998