March 2001 - XEmacs-Beta - XEmacs Mailing List Archives

large rearrangements of code that handles textual representation 23 years, 8 months

Ben Wing

i'm starting to work on reintegrating my Mule workspace back into the development workspace. since the development workspace has just started a new cycle and development on it is fairly slow or perhaps even nonexistent, this seems a good time to act. i'm generally going to be working from large changes to small ones. the first thing i'm going to be doing is what might be the "largest" change from the perspective of diffs, although there isn't much actual code logic change going on. rather, i cut out all the code that's scattered over numerous modules that knows about and deals with the exact representation of internal text in buffers, strings, and characters, and put it into a single file, text.c, along with a header text.h. this makes a lot of sense, because it isolates the code that knows the particulars about the representation into one place, and we can put a "firewall" of sorts around it -- all code outside knows only some very general characteristics, that we've specified. it must do all its work without knowledge of these particulars. in essence, this is just standard data-hiding and data-abstraction, so familiar in the OO world. the reason for this is because the internal structure is going to change at some point -- e.g. UTF2000, fixed 8-, 16-, and 32- bit buffers, fixed 16-bit Unicode buffers, etc. etc. the practical ramifications of this is that i'm going to be cutting and pasting a lot of code, which will mess people up if they've been working in these areas. luckily the code that works with the textual representations is pretty stable and seen almost no work on it in quite awhile, so i don't think the problems will be too much. but i'm going ahead and posting to warn people -- they have to be extra careful when they commit these changes. in particular, if they've made any changes in any of the areas where the major code movement will happen, their best bet would be to make a cvs diff of their whole workspace before updating after i've committed my changes, and back up the workspace. then do the update, and if there are any really huge conflicts [which would happen if you changed a line in the middle of some big piece of text that's getting moved], resolve the conflicts in favor of my changes [or just erase the file entirely and retrieve it again from cvs, if that's possible], then reapply your changes using patch or whatever in the place where the text was moved to. i'd like to hear from anyone who thinks they might run into problems from this change, so that i can coordinate when i'll do it. sound good? let me know. -- ben I'm sometimes slow in getting around to reading my mail, so if you want to reach me faster, call 520-661-6661. See http://www.666.com/ben/chronic-pain/ for the hell I've been through.

1 participants
0 comments

Eicq causing more problems than it's worth. 23 years, 8 months

Steve Youngs

It seams that Eicq is causing more problems than it is worth because of it's incompatibility with anything other than GNU/Linux. For that reason, I'm going to remove Eicq from the XEmacs Packages. Does anyone have any objections to this? -- |---<Steve Youngs>---------------<GnuPG KeyID: 9E7E2820>---| | XEmacs - It's not just an editor. | | It's a way of life. | |------------------------------------<youngs(a)xemacs.org>---|

3 participants
2 comments

skeleton.el, 21.1.14, is xemacs-base package always in every release? 23 years, 8 months

Michael McNamara

I've had a number of complaints from people that my verilog-mode.el, which requires skeleton.el, 'doesn't work with 21.1.14' I have that release running on my laptop, and I run 'locate-library skeleton.el', and see that for me it is in /usr/lib/xemacs/xemacs-packages/xemacs-base/skeleton.el However, for me the file is dated Jan 14, 1998. Q1: What should I tell these people? It is possible that they do not have xemacs-base installed on their machine? (How could anything work?) //' Michael McNamara <mac(a)verisity.com> _ // Sr VP Technology 650-934-6888 \ // Verisity Design 650-934-6801 FAX \// <http://www.verisity.com> 408-930-6875 Cell -------------------------------------------------------------- Get my verilog emacs mode from <http://www.verilog.com>

2 participants
1 comments

Updated Packages in Pre-Releases - Mar 3 23 years, 8 months

Steve Youngs

I've just uploaded the following packages to the Pre-Releases directory: cc-mode-1.24-pkg.tar.gz edit-utils-1.57-pkg.tar.gz eicq-1.04-pkg.tar.gz ilisp-1.16-pkg.tar.gz jde-1.25-pkg.tar.gz pcl-cvs-1.52-pkg.tar.gz psgml-1.23-pkg.tar.gz semantic-1.05-pkg.tar.gz speedbar-1.19-pkg.tar.gz vm-6.91-pkg.tar.gz xslt-process-1.01-pkg.tar.gz Installing them: =============== 21.1 1) Options -> Manage Packages -> Add Download Site -> Pre-Releases 2) Options -> Manage Packages -> List and Install 3) Select the packages you wish to install/upgrade (instructions at the end of Packages buffer) 4) Package -> Install/Remove Selected 5) Re-start XEmacs 21.2 1) Tools -> Packages -> Add Download Site -> Pre-Releases 2) Tools -> Packages -> List and Install 3 - 5) As per 21.1 As usual, please let me know if anything goes belly up. -- |---<Steve Youngs>---------------<GnuPG KeyID: 9E7E2820>---| | XEmacs - It's not just an editor. | | It's a way of life. | |------------------------------------<youngs(a)xemacs.org>---|

2 participants
2 comments

make bindist failure on eicq 23 years, 8 months

Vin Shelton

I had a few spare minutes on my hands at work this afternoon, so I decided to fire up a build of the XEmacs packages from the latest CVS sources on a SunOS-5.5 box. I build the packages regularly on my Linux box at home, but I haven't recently built the packages under SunOS. I'm getting the following failure in comm/eicq: make[2]: Entering directory `/u/shelton/new/build/SunOS-5.5/xemacs-packages-2001 -03-04/comm/eicq' g++ -o udp2tcp udp2tcp.cc -O2 -Wall -static udp2tcp.cc: In function `int main(int, char **)': udp2tcp.cc:341: passing `size_t *' as argument 3 of `accept(int, sockaddr *, int *)' changes signedness make[2]: *** [udp2tcp] Error 1 make[2]: Leaving directory `/u/shelton/new/build/SunOS-5.5/xemacs-packages-2001-03-04/comm/eicq' make[1]: *** [eicq/bindist.target] Error 2 make[1]: Leaving directory `/u/shelton/new/build/SunOS-5.5/xemacs-packages-2001- 03-04/comm' make: *** [comm/bindist.target] Error 2 Holy signed/unsigned conversion, Batman! After hacking that in the sources, I had to make the following changes to the link command: I had to add '-lsocket -lnsl -lresolv -ldl'. I had to remove -static, because resolv and dl are not available in static form. This enabled me to proceed. It seems that it will be necessary to have a configure script in the comm/eicq directory in order to build udp2tcp. - vin -- In a minute there is time For decisions and revisions which a minute will reverse. T.S. Eliot [URL: http://www.cs.amherst.edu/~ccm/prufrock.html]

1 participants
0 comments

Re: W3 problems 23 years, 8 months

William M. Perry

Norbert Koch <nk(a)LF.net> writes: > Hi! > > I've destroyed part of my installation. Whenever I try to run w3, I > get the following bunch of warnings: Just run make install (or reinstall the package) again, and it will reinstall the css stylesheet. -bp

2 participants
1 comments

GNUS vs. Gnus 23 years, 8 months

Robin S. Socha

Not really a bug, but: |In any case, after spending all that energy on coming up with a new |and spunky name, we decided that the name was _too_ spunky, so we |renamed it back again to "Gnus". But in mixed case. "Gnus" vs. |"GNUS". New vs. old. Now, at least lisp/menubar-items.el lisp/simple.el still carry GNUS ;-)

1 participants
0 comments

Important: GTK builds and configure options; NT v. GTK 23 years, 9 months

Stephen J. Turnbull

>>>>> "Gordon" == Gordon Sadler <gbsadler1(a)lcisp.com> writes: Gordon> I did not specify --with-gtk=no, as I'm not even sure if Gordon> gtk will run under openbsd, probably does, but I don't Gordon> have it there. Please use explicit --with-gtk flags when building in the GTK tree. The configure default is set to --with-gtk=no. The difference between defaulting this option and using --with-gtk=no is purely for the same of my statistical scripts: a successful configure using this flag must be using GTK sources. If you default it, I can only guess by looking for references to gtk in the text. I simply use /gtk/i; I have false negatives (people who simply use the default "xemacs" root for the checkout) and false positives (people who write "my gtk report is in a separate message"). Gordon> Are you still interested in native NT builds as well? Yes! NT (all platforms) is extremely important, because it's not one that Bill is terribly familar with AFAIK, and therefore most at risk. NT native is surely the most risky, since we can't even count on a Unix-like build environment (eg, no autoconf). -- University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN Institute of Policy and Planning Sciences Tel/fax: +81 (298) 53-5091 _________________ _________________ _________________ _________________ What are those straight lines for? "XEmacs rules."

1 participants
1 comments

Re: build reports on gtk-xemacs 23 years, 9 months

William M. Perry

Peter Brown <rendhalver(a)users.sourceforge.net> writes: > hi all > was trying to send a build report on gtk-xemacs > the M-X build-report command doesnt seem to work as it does with > xemacs > > is there a tricky way to send them from gtk-xemacs ?? Mine worked here (I sent off about 125 of them off to stephen the other day). Does it signal an error? If it does, try setting debug-on-error and sending the backtrace. -bp

3 participants
3 comments

Re: gtk-xemacs 23 years, 9 months

William M. Perry

Andy Piper <andyp(a)bea.com> writes: > I needed the following patch to get it to build under mswindows: I'll commit this patch when I get home tomorrow, unless you want to do it before then. Any chance you could look at the reported failure in the pdump code under cygwin? I'm not sure if I know where to start tracking that thing down. -bp

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 March 2001