[Please respect the CoC[0] and avoid sending me copies of the message]
[0]http://www.debian.org/MailingLists/#codeofconduct
Charliej wrote:
Raphael Geissert wrote:
Charliej wrote:
With that said this is a quote from my upstream:
"Whee, is this something you could take care of at install, delete and
then create a symlink. Or would this still violate this policy"
Would this be a viable solution? (I told upstream I would ask)
I don't think this is a viable solution because the lintian error would
still remain.
You should symlink the file provided by the other package, if lintian
still complains it is a bug in lintian then.
I think one of the reasons he is hesitant is that the removal of
"prototype" from the .tar.gz that he hosts would severely impact his M$
Windows users (he has a rather large M$ Windows community). Actually I
could care less about M$ Windows users, but he does.
As I see it to comply with Policy 4.13 "prototype" will have to be
removed from upstreams .tar.gz.. Am I correct in this assumption?
No, he doesn't need to remove his copy of prototype, it is _you_ who
needs to prevent it from being installed in the package you build.
This I can do, but to clarify your above statement does that mean that I
remove it from the .orig.tar.gz or do I adjust the rules and or postinst
to insure that it does not get installed?
The latter :)
hmm I think I may have answered my own question. Upstreams .tar.gz and
the .orig.tar.gz must have the same md5sum correct? if so then that
s/must/should, there are situations where one must repackage the tarball,
but I don't believe this is the case.
means the rules and or postinst have to be modified. Is my thinking
here correct?
Besides what I just clarified, yes.
This brings me to the next point what if upstream refuses to remove
"prototype" from the .tar.gz?
Thx
Charlie
Cheers,
Cheers,