home *** CD-ROM | disk | FTP | other *** search
- Xref: sparky gnu.emacs.sources:962 comp.lang.c++:19844 gnu.g++.help:1700 comp.emacs:4058 alt.lucid-emacs.help:939
- Newsgroups: gnu.emacs.sources,comp.lang.c++,gnu.g++.help,comp.emacs,alt.lucid-emacs.help
- Path: sparky!uunet!haven.umd.edu!darwin.sura.net!lhc!lhc!warsaw
- From: warsaw@nlm.nih.gov (Barry A. Warsaw)
- Subject: Announcing c++-mode.el version 2.261
- Message-ID: <WARSAW.93Jan24182134@anthem.nlm.nih.gov>
- Lines: 85
- Sender: news@nlm.nih.gov
- Reply-To: c++-mode-help@anthem.nlm.nih.gov (Barry A. Warsaw)
- Organization: Century Computing, Inc.
- Date: 24 Jan 93 23:21:34 GMT
-
-
- [[I don't think this made it out the first time. Sorry for any
- duplication... baw]]
-
- This message is to announce the release of c++-mode.el version 2.261
- for GNU emacs and derivatives (Epoch, Lucid emacs). This version
- should replace all previous versions, including the last public
- release 2.195. Differences between 2.195 and 2.261 are briefly
- outlined below. c++-mode.el can be used to edit C++, K&R C, and ANSI C
- code.
-
- As usual, I'm not posting the lisp file since it should be pretty easy
- for just about everyone to get it via anonymous ftp or mail-server.
- 2.261 should be available soon on the elisp archive, but you can also
- aftp it from ftp.cme.nist.gov:pub/gnu/c++-mode.el.Z. For those of you
- without aftp, send the following message to library@cme.nist.gov:
-
- send pub/gnu/c++-mode.el.Z
-
- Be sure that you include a valid reply path. When you receive the
- response, you need to remove the mail headers and other cruft,
- uudecode, then uncompress. For more information about the
- mail-server, send the message `help' (sans quotes) to the library
- address.
-
- Read the comments at the top of the file for information about getting
- on the beta testers list, and submitting bug reports.
-
- Performance
- ===========
- There have been lots of complaints about performance with 2.195. These
- problems are related to bugs in emacs which prohibit more than 1 style
- of comments per buffer. C++, of course, requires 2 styles of comments.
- c++-mode employs lisp workarounds so that indentation is accurate, but
- lisp is slower than using the C built-ins. I've tried to speed up the
- lisp routines as much as possible, but you'll still notice c++-mode
- getting slow inside big routines, or in a file of preprocessor
- commands. The solution is to patch emacs...
-
- ...which has been done for you as of the 19.4 Lucid Emacs release!
- Tons 'o thanks to Jamie Zawinski for working with me on these. Lemacs
- 19.4 also comes with 2.261 so lemacsers win big. c++-mode can
- recognize a patched emacs and will automatically use the faster
- built-in routines. Also, backslashing certain characters (e.g. single
- quote) inside comment regions are unnecessary when the patches are
- installed and c++-mode will automatically set the value of the
- c++-untame-characters variable to nil in a patched emacs.
-
- Jack Repenning has ported the lemacs patches back into GNU emacs 18.59
- and this will be submitted to the gnu.emacs.sources newsgroup in a
- companion article (thanks Jack!). The patches will also be made
- available on ftp.cme.nist.gov for access via aftp and mail-server,
- under the file name pub/gnu/syntax-18.59.Z. Of course, these are *not*
- official GNU patches. Volunteers to port the patches to Epoch are
- welcome!
-
- Other Changes Since 2.195
- =========================
-
- * LOTS OF BUG FIXES. I won't bore you with the details, but almost all
- reported bugs have been fixed. There are still some problems with
- indentation of continued statements at the top level (ie not inside
- functions), and re-indenting continued statements with
- c++-indent-exp (ESC C-q).
-
- * function c++-fill-C-comment has been removed.
-
- * variable c++-class-member-indent has been removed. c-indent-level is
- now used to determine indentation inside class definitions.
-
- * column-zero comment only lines need not be anchored to column-zero.
- See the docstring for variable c++-comment-only-line-offset for more
- information.
-
- * new variable c++-backscan-limit to improve performance when *not*
- running a patched emacs. This limits how far back lisp back-scanning
- will look at the expense of occasionally (hopefully, very rarely)
- being wrong. This feature is not used in patched emacses.
-
- * new functions c++-backward-sexp and c++-forward-sexp are bound to
- M-C-b and M-C-f respectively. These work better than backward-sexp
- and forward-sexp for c++-mode.
-
- Enjoy,
- -Barry
-