Re: RFS: mustang, btk-core
On 09/01/2008, at 1.45, Charles Plessy wrote:
For the copyright files, you may be interested by the proposed
machine-parsable format described in the following link. Although no
parser has been written yet, it could be useful to start to use it:
http://wiki.debian.org/Proposals/CopyrightFormat
A very interesting proposal! I will certainly start to use that format.
About the manpages, many thanks for writing them. Have you considered
submitting them upstream ?
Yes, I will send them. I wrote the man pages because I first included
the "example" programs in the package as binaries. However, I later
decided that they were not really robust enough to justify their
presence as "real" applications. So I kept the manpages, but put them
with the examples.
I have a few comments specific to btk-core:
- why providing libbtk-core-dev but not libbtk-core ?
Upstream does not build shared libraries, so there is no need for it.
I have chosen not to build shared libraries, because I would have to
choose a soname, and if upstream at some point decides to offer a
shared library, we could be in trouble.
- libbtk-core-dev should probably be in the libdevel section.
Yes it could. Upstream defines the intended audience as "Developers,
Science/Research". I assumed the package would appeal more to
scientists than "ordinary" developers, so I chose the "science"
category. I have no strong opinions on the matter, however.
- in your changelog, a colon is missing: (Closes: #459753)
Fixed.
- how about packaging the docs as well ?
I assume you mean the doxygen documentation. I seem to remember I had
problems generating it, so I let it be, but I can take another look
at it. Of course the documentation should be available to the
programmer!
Cheers,
Morten
PS: I am quite often on IRC. If anyone wants to chat, I will be in
#debian-med and #debian-mentors whenever possible.
Reply to: