was causing problems with a few ports determining that .c was the suffix
for executables. The fix is to expand it to '*.c | *.C' so that m4(1)
doesn't try to outsmart us.
Noticed by: George W. Dinolt <gdinolt@pacbell.net>
be used inside Makefiles for testing.
Submitted by: sobomax
(2) Use newly added /etc/mtree/BSD.x11-4.dist when XFREE86_VERSION=4.
Requested by: taguchi@tohoku.iij.ad.jp
(3) Re-enable MASTER_SITE_OVERRIDE and MASTER_SITE_BACKUP
functionalities by making MASTER_SORT ignore them (i.e., leave
*_OVERRIDE at the beginning and *_BACKUP at the end). As a side
effect, there are new targets "master-sites" and "patch-sites"
that will print out the correctly sorted versions of said lists.
Reviewed by: the ports list
(3') Sort patch sites too, not only master sites.
Reviewed by: the ports list
(4) New target "deinstall-depends" which will do a "make deinstall" in
all ports this one depends on.
(4') Only run pkg_delete from the deinstall target when package
exists, to avoid "no such package installed" errors.
URGENT UPGRADE REQUIRED: The previous version had a security hole when
it is used as a CGI program -- which allows HTTP remote users to view
arbitrary files on the web server.
2000-08-25 18:01 knu
* cvsweb.cgi: Get side-by-side diff working when $showfunc is true,
noting that `-p' option of diff(1) can only be used with context
and unified diffs.
Submitted by: Roger Hardiman <roger@cs.strath.ac.uk>
Allow to specify CVS tags or branch names (including `.' and
`HEAD') on cvs checkout.
e.g. http://foo/cvsweb.cgi/bar/dood.c?rev=.
http://foo/cvsweb.cgi/bar/dood.c?rev=RELENG_4http://foo/cvsweb.cgi/bar/dood.c?rev=RELENG_4_1_0_RELEASE
Not for diff currently, as rcsdiff does not grok CVS tags.
Inspired by: Folks at the FreeBSD cvs-all list.
Specify `-R' (turn on read-only repository mode) and `-l' (do not
log in the command history) on cvs checkout, so that one does not
need a write permission with the repository.
2000-08-25 00:53 knu
* cvsweb.cgi: MFZ 1.94. (Fix a typo in JavaScript code)
- New chipset supported: S3 Savage3D (I have not tested it personally);
- Improved support for already supported chipsets;
- Change port versioning scheme to PORTVERSION=YYYYMMDD, because despite the
promise, the project had not released its magic 1.0 version (promised to be
about 3 months ago);
- Catch change of the Mesa3 version.
Please note, that it's a work in progress, so some features doesn't work (for
example mouse support). However, the functionality already present is
sufficient to run some of the popular apps (quakeforge for example).
I tested it on ATI Mach64 and C&T55554 adapters.
system, since the code has bit-rotted and does not grok the new bonobo
Problem reports by: Archie Cobbs <archie@whistle.com>
Sledgehammer patch by: ade
my dual PIII-700 trying to figure out the sizes of the
primary and secondary caches. The sizes turned out to be
rather wrong (definetly true for the L2 cache, may be L1
is indeed 32Kb). Their algorithm seems bogus to me, but
most importantly, the obtained values are NOT USED
anywhere in the GhostScript code... One of the submitted
patches rips the "cache size" section from the genarch
entirely.
The attached modification and new patches also:
. replace the use of mktemp/fopen with mkstemp/fdopen
. modify the unix-gcc.mak so that there is no need for
GNU-make
PR: 20038
Submitted by: Mikhail Teterin <mi@aldan.algebra.com>
