Re: [Pgsqlrpms-hackers] Antw: Re: 8.3beta2 fails to compile on Centos5 x86_64
От | Tom Lane |
---|---|
Тема | Re: [Pgsqlrpms-hackers] Antw: Re: 8.3beta2 fails to compile on Centos5 x86_64 |
Дата | |
Msg-id | 6619.1194818113@sss.pgh.pa.us обсуждение исходный текст |
Ответ на | Re: [Pgsqlrpms-hackers] Antw: Re: 8.3beta2 fails to compile on Centos5 x86_64 (Devrim GÜNDÜZ <devrim@CommandPrompt.com>) |
Список | pgsql-ports |
Devrim =?ISO-8859-1?Q?G=DCND=DCZ?= <devrim@CommandPrompt.com> writes: > On Sun, 2007-11-11 at 10:18 -0500, Tom Lane wrote: >> redhat-rpm-config, however, might well be relevant. I remember Devrim >> running into an odd build failure on a machine where it wasn't >> installed. > Yes. All Fedora clones that are < 7 (CentOS 4,5 , RHEL 4,5 and Fedora > <=6) have this problem. > I have filed a bug report against this, and here is the reply (and bug > was closed) > https://bugzilla.redhat.com/show_bug.cgi?id=355551 Hmmm ... Panu has a point, I guess. His reasoning seems to be that redhat-rpm-config is required for sane behavior on RHEL/Fedora platforms, but it might not be required --- or even available --- on other RPM-using platforms, therefore it's not appropriate for either rpm-build or individual SRPMs to Require: it. It would seem to me, though, that that makes it a distribution bug: Red Hat distros should ensure that redhat-rpm-config is always installed. Otherwise it's possible to mis-build on Red Hat platforms, which is exactly what all the package-require infrastructure is supposed to prevent. *Somebody* other than the end user ought to be taking care of this. I've added a comment about this to the above BZ entry. regards, tom lane
В списке pgsql-ports по дате отправления: