[Bug 616193] Review Request: freerdp - X Remote Desktop Protocol Client

bugzilla at redhat.com bugzilla at redhat.com
Thu Jul 29 00:40:47 UTC 2010


Please do not reply directly to this email. All additional
comments should be made in the comments box of this bug.


https://bugzilla.redhat.com/show_bug.cgi?id=616193

--- Comment #5 from Mads Kiilerich <mads at kiilerich.com> 2010-07-28 20:40:46 EDT ---
(In reply to comment #4)

I'm trying to extract as much knowledge and opinion from your packaging as
possible ;-)

> I was under the impression that certain libs were BSD (only). As soon as
> something BSD licensed builds into a separate library or binary, BSD must be
> mentioned explicitly. AFAICS this is the case here.

Can you be more specific? Which library/binary, and where are the statements of
BSD license?

> > It is mostly the core freerdp lib, so shouldn't it be called something with lib
> > instead?
> 
> Not sure, it also includes the keyboard definitions.

Yes, but that is data for libfreerdpkbd and not intended to be used or seen in
other ways.

> > I also find it confusing that the -devel package is for a -common package.
> 
> Huh?

I would expect "foo-devel" to contain the headers etc for libraries in "foo".
The devel stuff for "foo-common" should be in "foo-common-devel"?

> > > sed -i 's|^hardcode_libdir_flag_spec=.*|hardcode_libdir_flag_spec=""|g' libtool
> > > sed -i 's|^runpath_var=LD_RUN_PATH|runpath_var=DIE_RPATH_DIE|g' libtool
> > 
> > Are they really needed? AFAICS my packages without this hack don't have any
> > issues.
> 
> I'm getting rpaths when I build this locally, this doesn't happen in koji or
> mock tough.

You don't know why?

I will investigate further.

> > > make install DESTDIR=$RPM_BUILD_ROOT INSTALL='install -p'
> > 
> > AFAIK -p isn't required by the guidelines. Just doing it on the packages where
> > the maintainer cares seems a bit odd.
> 
> The guidelines mention to preserve timestamps. Things like the keyboard
> defintions that are not getting compiled should IMHO have the upstream
> timestamp.    

Yes, "consider" preserving timestamps. Something less vague could be nice. Now
it is up to my preference to focus on simple spec and legibility or on
preserving timestamps - that is bad for consistency ...

-- 
Configure bugmail: https://bugzilla.redhat.com/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.



More information about the package-review mailing list