https://bugzilla.redhat.com/show_bug.cgi?id=1344276
Sachidananda Urs surs@redhat.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |sabose@redhat.com
--- Comment #32 from Sachidananda Urs surs@redhat.com --- (In reply to Paulo Andrade from comment #31)
Links are confusing. The https://download.gluster.org/pub/gluster/gdeploy/gdeploy.spec file is very outdated. The actual srpm is https://download.gluster.org/pub/gluster/gdeploy/gdeploy-2.0.2-13.src.rpm
Sorry about that.
SPEC URL: http://thegaul.org/gdeploy/gdeploy.spec SRPM URL: http://thegaul.org/gdeploy/gdeploy-2.0.2-13.src.rpm
I'll keep them consistent.
Spec from link differs from spec in srpm. Changelog is missing from -5 to -13 in the srpm. From the link jumps to -4 to -13.
Fixed.
Tested downloading the src.rpm and running: $ fedora-review -r -n gdeploy
I suggest you to install the fedora-review package, and test yourself running: $ fedora-review -b 1344276
fedora-review does not report anything alarming. Except few complaints on missing shebang which I think is fine.
This way you can see most if not all details a reviewer will talk about :)
- Please update the bug report with matching SRPM URL and SPEC URL. Also make sure src.rpm spec matches spec in SPEC url.
Done. They match now.
- License does not match. There are several GPLv3+ licensed files in the tarball. Apparently, should use "License: GPLv2+ and GPLv3+" File gdeployrest/gdeployapi.py is GPLv3+ with incorrect FSF address.
Fixed.
- Documentation should be built with sphinx. Not install sources, e.g. in build have: pushd doc make html popd
Done. This is a good idea, thanks.
- Documentation should be in a separate -doc package. It is already large in source format, and will be larger in html format.
gdeploy is now split into gdeploy and gdepoy-doc packages.
- Upstream source is not available:
https://github.com/gluster/gdeploy/archive/v2.0.2.tar.gz#/gdeploy-2.0.2.tar. gz Latest from github is 2.0.1.
Fixed this.
- Please either add a %check section and run the script in the tests directory, or give a good reason for not to. Usually, just loading the python files is enough to detect issues on other architectures, or other "random" issues.
The tests are not complete, they are outdated and not maintained. Plan to fix this in future releases. Request an exception for this.