On Sat, Jul 14, 2007, Jeff Johnson wrote:
[...]
- add arbitrary tag strings to header metadata.
[...]
default:
- rpmError(RPMERR_INTERNAL, _(Internal error: Bogus tag %d\n), tag);
- return RPMERR_INTERNAL;
+ macro = 0;
+ (void) headerAddEntry(pkg-header, tag,
On Sat, Jul 14, 2007, Jeff Johnson wrote:
[...]
%_extra_tags Class:Foo:Bar:Quux:X-*
and only if this macro exists we allow the tags Class, Foo, Bar,
Quux, X-Whatever, X-Something, etc. For the crazy distributions
we could even support %_extra_tags *. But only if this tag is defined
we
I fixed a problem with file magic previously, now this
make tmacro
...
macro.c: In function ‘expandMacro’:
macro.c:1455: error: ‘rpmlua’ undeclared (first use in this function)
macro.c:1455: error: (Each undeclared identifier is reported only once
macro.c:1455: error: for each function it
Since 4.4.7 rpm checks (auto-generated) dirnames deps, but I haven't
found how to ask rpmdb for packages which requires particular directory,
i.e. something like:
$ rpm -q --what-requires-dir /etc
Is there any way to do this? I need it for poldek's depsolver.
On Jul 14, 2007, at 11:10 AM, Paweł A. Gajda wrote:
Since 4.4.7 rpm checks (auto-generated) dirnames deps, but I haven't
found how to ask rpmdb for packages which requires particular
directory,
i.e. something like:
$ rpm -q --what-requires-dir /etc
Is there any way to do this? I need it
On 7/14/07, Jeff Johnson [EMAIL PROTECTED] wrote:
On Jul 14, 2007, at 3:37 AM, Ralf S. Engelschall wrote:
For RPM itself it could make sense to ship with either #%_extra_tags
* (backward compatibility mode), %_extra_tags X-* (escaping mode)
or %_extra_tags * (convenience mode) in macros. I
On Jul 14, 2007, at 11:22 AM, Paweł A. Gajda wrote:
On 7/14/07, Jeff Johnson [EMAIL PROTECTED] wrote:
On Jul 14, 2007, at 3:37 AM, Ralf S. Engelschall wrote:
For RPM itself it could make sense to ship with either #%
_extra_tags
* (backward compatibility mode), %_extra_tags X-* (escaping
On 7/14/07, Jeff Johnson [EMAIL PROTECTED] wrote:
On Jul 14, 2007, at 11:10 AM, Paweł A. Gajda wrote:
Since 4.4.7 rpm checks (auto-generated) dirnames deps, but I haven't
found how to ask rpmdb for packages which requires particular
directory,
i.e. something like:
$ rpm -q
Jeff Johnson wrote:
This patch is pretty close but is not expanding the parameterized %patch
correctly after the rewrite.
Specifically -b .foo is not becoming -b --suffix .foo with
the rewrite %patchNN - %patch -P NN.
What should be responsible for the -b .foo - -b --suffix .foo?
I don't
Ralf S. Engelschall wrote:
On Sat, Jul 14, 2007, Jeff Johnson wrote:
[...]
I don't agree, but I do not wish to discuss further. The cost of
the discussion is already higher than the cost of the preemptive
implementation.
Lusers asked for the implementation, now deal with the consequences
On Jul 14, 2007, at 4:30 PM, Mark Hatle wrote:
Ralf S. Engelschall wrote:
On Sat, Jul 14, 2007, Jeff Johnson wrote:
[...]
I don't agree, but I do not wish to discuss further. The cost of
the discussion is already higher than the cost of the preemptive
implementation.
Lusers asked for the
Thank you.
A regression test of macro syntax is almost certainly gonna be
needed, sez'
the guy who just spent the day debugging a smallish %patch macro
expansion.
Recursive expansions are tough debugging.
I shudder when I think of the amount of time that I will have to
spend on IRC
dumpasn1 is carried for X.509 signatures somewhen.
On Jul 14, 2007, at 6:45 PM, Ralf S. Engelschall wrote:
RPM Package Manager, CVS Repository
http://rpm5.org/cvs/
__
__
Server: rpm5.org
Thanks for the %patchNN rewrite. Holler if %patch breaks.
On Jul 14, 2007, at 7:03 PM, Mark Hatle wrote:
RPM Package Manager, CVS Repository
http://rpm5.org/cvs/
__
__
Server: rpm5.org
Jeff Johnson wrote:
Thanks for the %patchNN rewrite. Holler if %patch breaks.
I'm about to run it through another build sanity check.. (thats how I
noticed the '\' was missing suddenly nothing was patched..)
--Mark
On Jul 14, 2007, at 7:03 PM, Mark Hatle wrote:
RPM Package Manager, CVS
On Jul 14, 2007, at 8:57 PM, Mark Hatle wrote:
In rewriting the %patch macro, I don't understand why the \\n%{nil}
was necessary. w/o it though there is no \n at the end of the
generated
line. Is this a parser bug/feature?
White space trimmed at EOL, the %{nil} stops the trim.
--Mark
16 matches
Mail list logo
