Re: Shipping .orig files

2014-04-05 Thread Orion Poplawski
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

On 04/04/2014 12:45 PM, Stephen Gallagher wrote:
 On 04/04/2014 11:44 AM, Oden Eriksson wrote:
 fredagen den 4 april 2014 09.32.06 skrev  Orion Poplawski:
 A number of rpms currently ship .orig files, which are
 presumably from patching the files in the spec.  Would it make
 sense to have rpm complain about this?
 
 No, if -b is used there won't be a .orig file, unless it's named 
 so.
 
 
 I'm not sure when this changed, but I just did the following 
 experiment on nodejs-less (one of the identified packages from the 
 original message):
 
 1) Extract tarball 2) Validate that bin/lessc is the only file in
 the bin/ directory 3) patch -p1 
 0001-Require-include-files-from-the-default-location.patch 4) Look
 at bin/lessc: it created a .orig file
 
 Digging much deeper, it appears that this is because my patch
 applied with an offset. The default behavior of patch on Fedora
 appears to be --backup-if-mismatch, hence the creation of the .orig
 file.
 
 I'd be inclined to suggest that we may want to file a bug against 
 redhat-rpm-macros and get the %patch macro updated to include 
 --no-backup-if-mismatch in rpmbuild.
 

Yeah, this is what I was seeing too.  It's been suggested before
apparently: https://bugzilla.redhat.com/show_bug.cgi?id=884755


- -- 
Orion Poplawski
Technical Manager 303-415-9701 x222
NWRA/CoRA DivisionFAX: 303-415-9702
3380 Mitchell Lane  or...@cora.nwra.com
Boulder, CO 80301  http://www.cora.nwra.com
-BEGIN PGP SIGNATURE-
Version: GnuPG v1
Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/

iEYEARECAAYFAlNAH0UACgkQORnzrtFC2/tV0wCgioVNySiBQ6o05x0KQ1KqIU0C
f1QAn2Xbzz/pu8ckfH1AKASrVoa2yd+a
=PAAC
-END PGP SIGNATURE-
-- 
devel mailing list
devel@lists.fedoraproject.org
https://admin.fedoraproject.org/mailman/listinfo/devel
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct

Shipping .orig files

2014-04-04 Thread Orion Poplawski
A number of rpms currently ship .orig files, which are presumably from 
patching the files in the spec.  Would it make sense to have rpm complain 
about this?


# repoquery --disablerepo=\* --enablerepo=rawhide --whatprovides '/*.orig' | 
sort -u

alliance-doc-0:5.0-35.20090901snap.fc18.x86_64
arduino-core-1:1.0.5-6.fc21.noarch
asterisk-0:11.8.1-1.fc21.i686
asterisk-0:11.8.1-1.fc21.x86_64
cgit-0:0.10.1-1.fc21.x86_64
cinnamon-0:2.0.14-16.fc21.x86_64
cmake-0:3.0.0-0.5.rc3.fc21.x86_64
conntrack-tools-0:1.4.2-3.fc21.x86_64
devassistant-0:0.8.0-2.fc21.noarch
dokuwiki-0:0-0.21.20131208.fc21.noarch
drupal6-0:6.30-1.fc21.noarch
drupal7-0:7.26-3.fc21.noarch
eclipse-jbosstools-as-0:4.1.1-2.fc21.noarch
eclipse-pydev-1:3.4.1-1.fc21.noarch
eclipse-swt-1:4.4.0-0.5.git201404011500.fc21.x86_64
erlang-tools-0:R16B-03.3.fc21.x86_64
eucalyptus-console-0:3.3.0-0.5.20130408git32052445.fc20.x86_64
faust-tools-0:0.9.46-3.fc20.noarch
freenx-server-0:0.7.3-32.fc20.x86_64
geda-docs-1:1.8.2-1.fc21.noarch
geda-gaf-1:1.8.2-1.fc21.x86_64
geda-gnetlist-1:1.8.2-1.fc21.x86_64
git-0:1.9.0-1.fc21.x86_64
gplcver-0:2.12a-7.fc20.x86_64
japanese-bitmap-fonts-0:0.20080710-14.fc20.noarch
jruby-0:1.7.2-5.fc21.noarch
latex2html-0:2012-4.fc20.noarch
marked-0:0.2.9-3.fc20.noarch
mingw32-qt-0:4.8.5-4.fc20.noarch
mingw32-qt5-qmake-0:5.2.1-1.fc21.x86_64
mingw64-qt-0:4.8.5-4.fc20.noarch
mingw64-qt5-qmake-0:5.2.1-1.fc21.x86_64
MySQL-zrm-0:3.0-2.fc21.noarch
node-gyp-0:0.10.6-2.fc20.noarch
nodejs-less-0:1.7.0-1.fc21.noarch
openstack-tripleo-0:0.0.2-4.20140220git.fc21.noarch
oscap-anaconda-addon-0:0.6-2.fc21.noarch
perl-ExtUtils-AutoInstall-0:0.63-22.fc20.noarch
perl-Frontier-RPC-0:0.07b4p1-19.fc20.noarch
perl-Graph-Easy-0:0.73-3.fc20.noarch
perl-SystemPerl-0:1.336-9.fc20.x86_64
perl-Verilog-CodeGen-0:0.9.4-14.fc20.noarch
phpesp-0:2.1.1-8.fc20.noarch
python3-crypto-0:2.6.1-1.fc21.x86_64
python-altgraph-0:0.10.1-3.fc20.noarch
python-crypto-0:2.6.1-1.fc21.x86_64
python-sprox-0:0.8.3-2.fc20.noarch
ReviewBoard-0:1.7.22-2.fc21.noarch
roundcubemail-0:0.9.5-1.fc21.noarch
rubygem-fakeweb-doc-0:1.3.0-10.fc20.noarch
rubygem-httparty-doc-0:0.10.2-2.fc20.noarch
scilab-tests-0:5.5.0-0.3.beta1.fc21.1.noarch
sfact-0:0.0-5.20130128gitbc56c68.fc21.noarch
svgalib-devel-0:1.9.25-14.fc21.i686
svgalib-devel-0:1.9.25-14.fc21.x86_64
texlive-nddiss-doc-3:svn29349.3.0-8.fc21.noarch
texlive-pst-bar-doc-3:svn18734.0.92-8.fc21.noarch
tkgate-0:2.0-16.beta10.fc20.x86_64
tkgate-es-0:2.0-16.beta10.fc20.x86_64
tkgate-ja-0:2.0-16.beta10.fc20.x86_64
WebCalendar-0:1.2.7-3.fc20.noarch
xerces-c27-doc-0:2.7.0-13.fc20.x86_64
xfig-common-0:3.2.5-41.c.fc21.x86_64
yourls-0:1.7-1.fc21.noarch
zikula-0:1.2.3-5.fc19.noarch
zsh-0:5.0.2-8.fc21.x86_64

--
Orion Poplawski
Technical Manager 303-415-9701 x222
NWRA, Boulder/CoRA Office FAX: 303-415-9702
3380 Mitchell Lane   or...@nwra.com
Boulder, CO 80301   http://www.nwra.com
--
devel mailing list
devel@lists.fedoraproject.org
https://admin.fedoraproject.org/mailman/listinfo/devel
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct

Re: Shipping .orig files

2014-04-04 Thread Oden Eriksson
fredagen den 4 april 2014 09.32.06 skrev  Orion Poplawski:
 A number of rpms currently ship .orig files, which are presumably from
 patching the files in the spec.  Would it make sense to have rpm complain
 about this?

No, if -b is used there won't be a .orig file, unless it's named so.

-- 
devel mailing list
devel@lists.fedoraproject.org
https://admin.fedoraproject.org/mailman/listinfo/devel
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct

Re: Shipping .orig files

2014-04-04 Thread Stephen Gallagher
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

On 04/04/2014 11:44 AM, Oden Eriksson wrote:
 fredagen den 4 april 2014 09.32.06 skrev  Orion Poplawski:
 A number of rpms currently ship .orig files, which are presumably
 from patching the files in the spec.  Would it make sense to have
 rpm complain about this?
 
 No, if -b is used there won't be a .orig file, unless it's named
 so.
 

I'm not sure when this changed, but I just did the following
experiment on nodejs-less (one of the identified packages from the
original message):

1) Extract tarball
2) Validate that bin/lessc is the only file in the bin/ directory
3) patch -p1  0001-Require-include-files-from-the-default-location.patch
4) Look at bin/lessc: it created a .orig file

Digging much deeper, it appears that this is because my patch applied
with an offset. The default behavior of patch on Fedora appears to be
- --backup-if-mismatch, hence the creation of the .orig file.

I'd be inclined to suggest that we may want to file a bug against
redhat-rpm-macros and get the %patch macro updated to include
- --no-backup-if-mismatch in rpmbuild.
-BEGIN PGP SIGNATURE-
Version: GnuPG v1
Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/

iEYEARECAAYFAlM+/c8ACgkQeiVVYja6o6PcYACghQzfEyCwz9+6dUQsWTwZyKGu
sCQAniG015yL2iJKUa7GL/7N5KdaPLee
=a73t
-END PGP SIGNATURE-
-- 
devel mailing list
devel@lists.fedoraproject.org
https://admin.fedoraproject.org/mailman/listinfo/devel
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct