Please do not reply directly to this email. All additional comments should be made in the comments box of this bug report. Summary: Merge Review: w3m https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=226535 pnemade@xxxxxxxxxx changed: What |Removed |Added ---------------------------------------------------------------------------- Severity|normal |urgent ------- Additional Comments From pnemade@xxxxxxxxxx 2007-03-27 00:33 EST ------- (In reply to comment #3) > Created an attachment (id=150834) --> (https://bugzilla.redhat.com/bugzilla/attachment.cgi?id=150834&action=view) [edit] > w3m.spec with some fixes (0.5.1-18) > > Well, for 0.5.1-17.fc7: > > * Source0 > - Please check: > http://fedoraproject.org/wiki/Packaging/SourceURL > (A) It seems that newest gc is 6.8. > > * CFlags > (Please see the attached mock build log) > - Fedora specific compilation flags are not passed > (for gc.a). > For gc.a, it seems okay when ABI_FLAG is set as > RPM_OPT_FLAGS > And.. actually this changes debuginfo contents. Agree. > (B) NOTE: > Usually gc should be seperated from this package (w3m) > and other packages for gc should be created. > In this case, gc package should provide shared library > (not static archive). > However is this too late for F7? Don't think so. Yes static library should be removed. > > * Version provides > ------------------------------------------------ > Provides: webclient = 0.5.1 > ------------------------------------------------ > - Any reason to provide version-dependent virtual dependency? > I don't see the reason, and other package which provide > "webclient" virtual dependency does not specify version. Ok. > > * Requires > - For main package: > ------------------------------------------------ > Requires: perl, openssl > ------------------------------------------------ > These should be removed. rpmbuild automatically > finds these dependencies. yes this should be removed. > > * spec file description > ------------------------------------------------ > [ -n "$RPM_BUILD_ROOT" -a "$RPM_BUILD_ROOT" != / ] && rm -rf $RPM_BUILD_ROOT > ------------------------------------------------ > - The part [ ...... ] is redundant and should be removed. > RPM_BUILD_ROOT _MUST_ not be empty or / . Ok. > > * Timestamps > ------------------------------------------------ > install -m 644 %{SOURCE10} $RPM_BUILD_ROOT%{_sysconfdir}/w3m/config > ------------------------------------------------ > - Please keep timestamp. i.e. use "install -p". Yes. > > * Misc > * w3m.lang usage > ------------------------------------------------ > find $RPM_BUILD_ROOT%{_libexecdir} -type f -print | grep -v w3mimgdisplay | sed > -e "s,$RPM_BUILD_ROOT,," >> w3m.lang > ........ > %files > ........ > %exclude %{_libexecdir}/w3m/w3mimgdisplay > ------------------------------------------------ > - At the first description, w3mimgdisplay are already excluded. > Anyway, these can be unified. Simply, > ------------------------------------------------ > %files > ....... > %{_mandir}/man1/w3mman.1* > %{_libexecdir}/%{name}/ > %exclude %{_libexecdir}/w3m/w3mimgdisplay > ------------------------------------------------ > should be okay. Agree. > > * Conditional dependency > - Well, I found that when I rebuild w3m locally, inline image handler > is enabled for x11 and fb, > while for mockbuild only x11 image handler is enabled > > This is because configure reads: > ------------------------------------------------ > 5785 enable_image=x11 > 5786 case "`uname -s`" in > 5787 Linux|linux|LINUX) > 5788 if test -c /dev/fb0; then > 5789 enable_image=x11,fb > 5790 fi;; > 5791 esac > 5792 fi > 5793 save_ifs="$IFS"; IFS=","; > 5794 for img in $enable_image; do > ------------------------------------------------ > However, on mockbuild /dev/fb0 is not created, so fb image handler > will not be enabled. > To fix this, configure option must handle this explicitly. > > * lang > - for Japanese documents, these should be treated as %lang(ja). yes. > > * Documentation > (C) doc*/w3m.1 seems unneeded, as they are already included as > man files. However, I leave this as how you judge. > Ok. will remove from doc. > (D) Question: > - Should all the documents in doc-jp/ files should be > converted from EUC-JP to UTF-8? > I think so, however, if you want to do so, please keep > timestamps on these files even after encodings are converted, > as these documents are 3-6 years old. Agree. > > ------------------------------------------------- > My attached spec file should fix all the issues above > expect (A)-(D). > Please check my spec file and comment on (A)-(D) > > * NOTE > I have not checked yet > * what documentation should be added to this rpm > * whether license is correct and has no problem > * some other issues may exist -- Configure bugmail: https://bugzilla.redhat.com/bugzilla/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are the QA contact for the bug, or are watching the QA contact. _______________________________________________ Fedora-package-review mailing list Fedora-package-review@xxxxxxxxxx http://www.redhat.com/mailman/listinfo/fedora-package-review