Garrick Staples wrote: > On Fri, Aug 04, 2006 at 03:18:30PM +0200, Axel Thimm alleged: > >>On Fri, Aug 04, 2006 at 02:40:11PM +0300, Andy Shevchenko wrote: >> >>>Michael Schwendt wrote: >>> >>>>>Tonight I've received the following bug >>>>>http://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=201279 >>>>>I'm packager of the qstat tool. But the torque (in torque-client) >>>>>already consists the /usr/bin/qstat. >>> >>>>Why isn't this program called "torque-qstat"? You could patch it and >>>>rename it like that. What is the sense in creating such a short and >>>>generic name like "qstat"? Is the program useful outside of "torque"? >>>>Could it be placed in /usr/libexec/torque/qstat instead if not renamed? >>> >>>Thanks for opinoin, but I think some miscommunication is present here. >>>I'm not a packager of torque. As I understand you correctly the issue >>>inhere torque than qstat. Anyway I've filled bug to torque with >>>references to this maillist discussion already. I'm wating for torque >>>packager response. >> >>The bugzilla URL would be nice. >> >>On the use of qstat: qstat/qsub and friends have been around in >>Unix/Linux queuing systems for some decades now, and it is expected by >>users of these systems to find the tools under qstat and not >>dqs-qstat, nqs-qstat or torque-qstat. > > > The name "qstat" is actually a requirement in POSIX 1003.2d, and as Axel > stated, it has been implemented by many batch queueing systems over the > decades. > > If we had more than 1 batch queueing system, than setting up an > alternatives system would work. The 'qstat' package that Andy maintains is not part of a batch queuing system, however. It's a quake/halflife server query tool: http://www.qstat.org Perhaps the game-server qstat package should rename its binary to 'quakestat' to avoid the conflict? --Mike
Attachment:
smime.p7s
Description: S/MIME Cryptographic Signature
-- fedora-extras-list mailing list fedora-extras-list@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/fedora-extras-list
