Bug 191594
Summary: | Review Request: gtkglextmm | ||
---|---|---|---|
Product: | [Fedora] Fedora | Reporter: | Gilles Gagniard <gilles.gagniard> |
Component: | Package Review | Assignee: | Ralf Corsepius <rc040203> |
Status: | CLOSED NEXTRELEASE | QA Contact: | Fedora Package Reviews List <fedora-package-review> |
Severity: | medium | Docs Contact: | |
Priority: | medium | ||
Version: | rawhide | CC: | karlthered |
Target Milestone: | --- | ||
Target Release: | --- | ||
Hardware: | All | ||
OS: | Linux | ||
Whiteboard: | |||
Fixed In Version: | Doc Type: | Bug Fix | |
Doc Text: | Story Points: | --- | |
Clone Of: | Environment: | ||
Last Closed: | 2006-05-22 22:48:00 UTC | Type: | --- |
Regression: | --- | Mount Type: | --- |
Documentation: | --- | CRM: | |
Verified Versions: | Category: | --- | |
oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |
Cloudforms Team: | --- | Target Upstream Version: | |
Embargoed: | |||
Bug Depends On: | |||
Bug Blocks: | 163779 |
Description
Gilles Gagniard
2006-05-13 16:43:28 UTC
Some comments: * The spec "BuildRequires: gtkext >= 1.2.0". AFAICT, 1.0.0 should be sufficient. At least, I am not aware about any API changes between GtkGlExt-1.0.0 and 1.2.0 making this requirement necessary. A fact confirming this, is gtkglextmm's configure script to only check for gtkglext >= 1.0.0. * The spec explicitly Requires: gtkglext Requires: gtkmm24 This shouldn't be necessary. * Please explain /usr/lib/gtkglextmm-1.2/proc/m4/* I don't know what these files are (Look like some m4 macros to help converting some types), how they are supposed to be used and why they need to be shipped. AFAIS, they don't they seem to be used by anything in gtkglextmm. Alright, I reuploaded a new version of the spec. > * The spec "BuildRequires: gtkext >= 1.2.0". > AFAICT, 1.0.0 should be sufficient. At least, I am not aware about any API > changes between GtkGlExt-1.0.0 and 1.2.0 making this requirement necessary. > A fact confirming this, is gtkglextmm's configure script to only check for > gtkglext >= 1.0.0. Yup, you're right. As both of these packages have been released almost together, I thought that it was necessary ;) > * The spec explicitly > Requires: gtkglext > Requires: gtkmm24 > This shouldn't be necessary. Ok, I removed them. But for my personal education, could you explain me why ? It is the first rpm spec I write, I only have a previous experience with gentoo ebuilds ... Is it because rpm added automatically a dependency on the *librairies* in the gtkglext package by checking undefined symbols ? > * Please explain /usr/lib/gtkglextmm-1.2/proc/m4/* > I don't know what these files are (Look like some m4 macros to help converting > some types), how they are supposed to be used and why they need to be shipped. > AFAIS, they don't they seem to be used by anything in gtkglextmm. Well, I use this library as an app developper, and I don't need these files either. However, directly from the README file in the source package in tools/m4 : "This directory contains additional type conversions for gtkglextmm. The convert.m4 file overrides the file of the same name in gtkmm. Like the gtkmm m4 conversion files, these files are also installed, for use by other libraries." So I guess some people have a use for it ... (In reply to comment #2) > Alright, I reuploaded a new version of the spec. Please increment the release-tag. I am going to continue the review once you do. > > * The spec explicitly > > Requires: gtkglext > > Requires: gtkmm24 > > > This shouldn't be necessary. > > Ok, I removed them. But for my personal education, could you explain me why ? The main package is a pure run-time library package. Rpms of applications using these libraries will automatically be added dependencies on these shared libraries and their dependencies, when *building* rpms of these applications. (i.e. these Requires are necessary in *-devel rpms, but not in runtime rpms). > is the first rpm spec I write, I only have a previous experience with gentoo > ebuilds ... Is it because rpm added automatically a dependency on the > *librairies* in the gtkglext package by checking undefined symbols ? Almost. Rpm adds dependencies on the libraries (It adds "Requires: libfoo.so.1"), not on the package ("Requires: foo"). Depsolvers/Installers such as yum/apt etc. will translate "Requires: libfoo.so.1" into packages. > > * Please explain /usr/lib/gtkglextmm-1.2/proc/m4/* > > I don't know what these files are (Look like some m4 macros to help converting > > some types), how they are supposed to be used and why they need to be shipped. > > > AFAIS, they don't they seem to be used by anything in gtkglextmm. > > Well, I use this library as an app developper, I am heavily using GtkGLExt with C++ and have never found gtkmm/gtkglextmm to be attractive ;) > and I don't need these files > either. However, directly from the README file in the source package in tools/m4 : .. > So I guess some people have a use for it ... Hmm, seems like an historic artefact (== upstream bug) or a packaging bug in FE gtkmm to me. AFAICT, older versions of gtkmm seem to have shipped scripts using the macros, newer versions don't seem to do so. Anyway, this is not a blocker for this review. The release tag is incremented. Here are now the new URLs : Spec URL: http://jaile.ath.cx/gilles/fedora_core_5/gtkglextmm.spec SRPM URL: http://jaile.ath.cx/gilles/fedora_core_5/gtkglextmm-1.2.0-2.src.rpm Package looks fine, except one issue remaining: You are shipping libtool archives (*.la). The PackageGuideLines Gods want you to remove them. I for one consider this part of the package guidelines as in error, and therefore will not force anybody to remove *.la, but will leave such a decision to the packager. I.e. decide on yourself if you want to ship them or not. APPROVED. To get sponsored, please proceed with "Get a Fedora Account" on http://fedoraproject.org/wiki/Extras/Contributors Gilles, 1. In CVS, you are using the same Release:-tag for devel and FC-5. In future, please make sure these are different and that the Release tag of devel is greater than that of FC-5, otherwise upgrades from one FC-release to another one are likely to fail. I recommend using %{?dist}. As this is your first package, I've applied corresponding changes to devel and FC-5, and triggered a rebuild for devel. (ATM, there is no need to rebuild FC-5) 2. You should have closed this PR, when the packages had been released. 3. Please remove FE-NEEDSPONSOR from this PR and all others you might have added it to. *** Bug 193107 has been marked as a duplicate of this bug. *** |