Bug#205325: glibc: please add strnatcmp / strnatcasecmp
On 2 Oct 2003, Anand Kumria [EMAIL PROTECTED] wrote: On Thu, Aug 14, 2003 at 02:53:37PM +1000, Martin Pool wrote: On 14 Aug 2003, Anand Kumria [EMAIL PROTECTED] wrote: Package: glibc Version: unavailable; reported 2003-08-14 Severity: wishlist Hi there, I'm surprised there isn't already something similiar to strnatcmp in glibc URL: http://sourcefrog.net/projects/natsort/ but both functions would certainly be usable. The license appear to allow for the inclusion of both functions, so hopefully this isn't a huge to deal to get included. glibc has strverscmp(). strnatcmp() is much nicer IMO. Far more descriptive. Yes, the semantics are a little different. If you think you can persuade the glibc maintainers to take it then I'd be happy to support it. The strnatcmp licence was necessary to merge it into Apache, which is not possible with the GNU implementation. Does that mean you would be unwilling to relicense it - should it be required - to get it into GNU libc? At the moment it is under zlib, which is (I think) strictly weaker than GNU GPL. So they can either use it as is, or I can relicense it. -- Martin
Bug#205325: glibc: please add strnatcmp / strnatcasecmp
On 14 Aug 2003, Anand Kumria [EMAIL PROTECTED] wrote: Package: glibc Version: unavailable; reported 2003-08-14 Severity: wishlist Hi there, I'm surprised there isn't already something similiar to strnatcmp in glibc URL: http://sourcefrog.net/projects/natsort/ but both functions would certainly be usable. The license appear to allow for the inclusion of both functions, so hopefully this isn't a huge to deal to get included. glibc has strverscmp(). The algorithm is similar but not quite the same. You might still argue for including strnatcmp but it's probably not worthwhile. The glibc maintainers might (?) want it to be internationalized/wcharized if it was going to be merged. The strnatcmp licence was necessary to merge it into Apache, which is not possible with the GNU implementation. Regards, -- Martin -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#192876: libc6: getaddrinfo() does not detect invalid numeric services
I suppose it might be argued that it is not getaddrinfo()'s job to detect illegal invalid ports. I'm not sure this stands up: the function is documented as being a single step to get from a host and/or service name to a (set of) addresses that can be used to connect. If this checking is not done here, then where should it be done? And what possible purpose could there be for allowing an invalid port into the result? -- Martin
Bug#192876: libc6: getaddrinfo() does not detect invalid numeric services
Package: libc6 Version: 2.3.1-16 Severity: normal By my reading of the getaddrinfo() manual (and also the Solaris manpage) getaddrinfo() ought to return EAI_SERVICE when the requested service is not available. However, it does not seem to do this when the service is a number, but out of the valid range 1-65535 for TCP ports. Here is a reproduction case: -- #include string.h #include stdio.h #include stdlib.h #include sys/types.h #include sys/socket.h #include netdb.h /* try getaddrinfo */ int main(void) { int err; struct addrinfo *res; struct addrinfo hints; bzero(hints, sizeof hints); hints.ai_family = AF_INET; hints.ai_socktype = SOCK_STREAM; hints.ai_flags = AI_PASSIVE; err = getaddrinfo(NULL, 8, hints, res); printf(err=%d, %s\n, err, gai_strerror(err)); return 0; } -- On my machine this returns 0, rather than the expected -8 servname not supported. If I change 8 to an invalid non-numeric string then the error is flagged as expected. -- System Information: Debian Release: testing/unstable Architecture: i386 Kernel: Linux wistful 2.4.20 #24 Mon Dec 30 19:32:14 EST 2002 i686 Locale: LANG=C, LC_CTYPE=C Versions of packages libc6 depends on: ii libdb1-compat 2.1.3-7The Berkeley database routines [gl -- no debconf information -- Martin
Bug#57584: libc6: problem with devpts during configuration
Package: libc6 Version: 2.1.3-1 Severity: important Doing an apt-get dist-upgrade to libc6-2.1.3-1 from potato, configuration stops with these messages: --- Setting up libc6 (2.1.3-1) ... Current default timezone: 'Australia/Canberra'. Local time is now: Wed Feb 9 10:30:57 EST 2000. Universal Time is now: Tue Feb 8 23:30:57 UTC 2000. Run 'tzconfig' if you wish to change it. mount: devpts already mounted or /dev/pts busy mount: according to mtab, none is already mounted on /dev/pts dpkg: error processing libc6 (--configure): subprocess post-installation script returned error exit status 32 Errors were encountered while processing: libc6 --- $ mount /dev/hda6 on / type ext2 (rw,errors=remount-ro,errors=remount-ro) proc on /proc type proc (rw) /dev/hda7 on /usr type ext2 (rw) /dev/hda10 on /var type ext2 (rw) /dev/hda8 on /home type ext2 (rw) /dev/hda1 on /rh-root type ext2 (rw) /dev/hda11 on /snd type ext2 (rw) none on /dev/pts type devpts (rw,gid=5,mode=620) $ grep pts /etc/fstab none/dev/ptsdevpts gid=5,mode=620 0 0 Thanks, -- Martin -- System Information Debian Release: 2.2 Kernel Version: Linux sanguine 2.2.13 #3 Thu Feb 3 19:33:37 EST 2000 i686 unknown Versions of the packages libc6 depends on: ii ldso 1.9.11-5 The Linux dynamic linker, library and utilities.