On Sunday 10 February 2008, Frans Pop wrote:
As you may know, dependencies on glibc udebs are the only category that
is still incorrect [1]. This is an attempt to fix that.
JFYI.
After some more testing and minor changes in the patch, I have just filed
the BR against glibc to request
Hi Raphael,
Thanks for the preemptive response :-)
On Wednesday 13 February 2008, Raphael Hertzog wrote:
On Mon, 11 Feb 2008, Frans Pop wrote:
BTW, has anyone looked at how library symbols files interact with
udeb: lines in shlibs files? The symbols file tend to override the
shlibs
On Mon, 11 Feb 2008, Frans Pop wrote:
BTW, has anyone looked at how library symbols files interact with udeb:
lines in shlibs files? The symbols file tend to override the shlibs
files, and I wonder if this won't result in bad dependencies in udebs as
the libraries they depend on are
On Wed, 13 Feb 2008, Frans Pop wrote:
Symbols files are not used for udebs. So they won't affect dependencies
of udebs.
Not that I don't trust you about this, but just to make sure...
Do symbol files affect the way shlibs files are generated, or will those
remain unchanged?
Those
On Monday 11 February 2008, Joey Hess wrote:
Since splitting udebs differently than the debs are split is a special
case that seems likely to affect other libraries than glibc, I'd prefer
to keep the special case in glibc, rather than in dehelper.
s/likely/unlikely/ I guess. Yes, I thought as
Frans Pop wrote:
As you may know, dependencies on glibc udebs are the only category that is
still incorrect [1]. This is an attempt to fix that.
Normally we would add an option '--udeb udeb name' to the dh_makeshlibs
call in debian/rules. However, in the case of glibc this does not work.
On Sunday 10 February 2008, Frans Pop wrote:
Solution I have come up with is to write a small script that will
generate the udeb: lines _after_ dh_makeshlibs has generated the basic
shlibs file.
One thing I'm unsure of is whether or not I was correct in excluding shlibs
files for library
As you may know, dependencies on glibc udebs are the only category that is
still incorrect [1]. This is an attempt to fix that.
Normally we would add an option '--udeb udeb name' to the dh_makeshlibs
call in debian/rules. However, in the case of glibc this does not work.
Reason is that two
8 matches
Mail list logo
