https://bugzilla.redhat.com/show_bug.cgi?id=1557371 --- Comment #10 from Jie Kang <jkang@xxxxxxxxxx> --- (In reply to jiri vanek from comment #6) > [] Package must own all directories that it creates. > will be fixed. Only /usr/lib/jvm should not be owned. it is owned by > japackage tools. Also I think it is mentioned somewher ein guidelines, but > have not found. Are you sure with generic /usr/share/ stuff? > Also I belive I own /usr/lib/jvm/java-10-openjdk-10.0.0.46-3.fc27.x86_64 > .... at least it is listed for me... Right, [ ] means to manually check. I guess it wasn't able to automatically see? I would be suprised about /usr/lib/jvm/java-10-openjdk-10.0.0.46-3.fc27.x86_64 too but anyways, I'm going to see how it looks on my system. > [] %config files are marked noreplace or the reason is justified. > Note: No (noreplace) in %config > > /etc/java/java-10-openjdk/java-10-openjdk-10.0.0.46-3.fc27.x86_64/conf/ > management/jmxremote.password.template > This one realy do not deserves to be noreplace. It is config tempalte > > %config > > /etc/java/java-10-openjdk/java-10-openjdk-10.0.0.46-3.fc27.x86_64/conf/ > accessibility.properties > I had not marked this one, as it contains generated path and shold be > fully contorlled by packager. Unluvkily, the only palce where jdk founds it > is conf. Okay. As it asks for justification, can comments be added to the spec for each of them? > > Same for debug variants. > > [ ]: gtk-update-icon-cache is invoked in %postun and %posttrans if package > contains icons. > Note: icons in java-openjdk, java-openjdk-debug > > It is, inst it? Checked and you're right, it is invoked. > > > [!]: Uses parallel make %{?_smp_mflags} macro. > I do not dare to use this flag. The build of OJDK is to complex to set > anything on top. Okay. This isn't MUST or even SHOULD but EXTRA so can be ignored; thanks for justification. > > [!]: Spec use %global instead of %define unless justified. > This is major thing, and cant be fixed. The Usage of %global on top of > %define is wrongly interpreted in fedora. If you change any of highlighted > define to global, the build will misbehave for you. This rpmlint was > invalidated in f27 and RPM 4.14. I will try to find the chnageset and > bugzilla which made me to change those selected globals back to defines. Okay, your justification sounds good. Is there anyway to add information to the spec on this? Any point even? I'll post an updated review (still two things incomplete) but I will try to check file ownerships next morning. -- You are receiving this mail because: You are on the CC list for the bug. You are always notified about changes to this product and component _______________________________________________ package-review mailing list -- package-review@xxxxxxxxxxxxxxxxxxxxxxx To unsubscribe send an email to package-review-leave@xxxxxxxxxxxxxxxxxxxxxxx
