Bug#681020: RFS: radeontop/0.5.4+1-gf2c42f4-1
Hi Jonathan,
On Tue, Jul 10, 2012 at 03:33:44PM +0100, Jonathan McCrohan wrote:
> Hi John,
Adrian is actually my primary name ;).
> I am also interested in radeontop, and was going to file an ITP, but it
> seems you got there before me. :-)
Hehe, I saw it on Phoronix and wanted to package it right away!
> I've had a look at your package, and I can see a number of
> issues/niggles:
>
Ah, thanks for the heads-up!
> general: * FTBFS because of missing Build-Depends: on
> libncurses5-dev and pkg-config.
> * Are you building the package in a clean chroot?
> (pbuilder, cowbuilder sbuild etc?)
Argh, no. I actually just used debuild this time. Was way too focused on
getting it lintian-clean. Will fix that!
>
> d/control: * Remove Vcs-Browser if you don't plan on using it.
Will do.
> * "It comes with a manpage" in the package description
> isn't very useful. It is expected that programs come
> with manpages. :-)
Will do.
> * Multi-Arch: foreign -> radeontop is not a -dev or
> support so this is not needed.
Well, I thought it's good to keep everything cross-installable. As far
as I understood the Multi-Arch policy every package except the
essential ones should be Multi-Arch-capable.
> d/README.Debian:* This states that this tarball was packaged from
> commit gf2c42f4, but I can't find any such commit
> upstream.
> * If this tarball is actually 0.5.4.1, it should be
> packaged from the tagged commit 2ebfb800e2 instead.
> * If you do switch to using upstream tarballs, drop
> README.Debian totally.
Ah, yes. I initially used a git snapshot and therefore put this
information into the README.Debian. I convinced the upstream author to
release a tarball which I then used for my package. I will just remove
the README.Debian.
> d/copyright: * The GPLv3 license is indented by too many spaces.
I will look into that.
> d/rules: * Remove the dh_make boilerplate
Will do.
> * Is export debug=1 required for the build? If not,
> remove it.
I discussed that in #debian-devel and I got the recommendation to set
debug=1 such that the Makefile of radeontop does not set CFLAGS=-s to
strip the code. dh_strip is supposed to do that.
Thank you very much for your input!
Adrian
Reply to: