[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: svn-inject wants to remove some files



Hi Tony,

On 9/1/06, tony mancill wrote:
As I understand it, you're supposed to be checking in the state of the
package in the build directory, and if the debian diff removed one of those
files, then that's ok.

It seemed a good solution, but for the package in exam
(wmspaceweather) I've found no deletion/replace in diff.gz of the
deleted files:

morph@morpheus:~/tmp/deb$ zcat wmspaceweather_1.04-17.diff.gz | grep ^---
--- wmspaceweather-1.04.orig/wmSpaceWeather/wmSpaceWeather.c
--- wmspaceweather-1.04.orig/wmSpaceWeather/Makefile
--- wmspaceweather-1.04.orig/wmSpaceWeather/wmSpaceWeather.1
--- wmspaceweather-1.04.orig/wmSpaceWeather/wmSpaceWeather_master.xpm
--- wmspaceweather-1.04.orig/wmSpaceWeather/GetKp
--- wmspaceweather-1.04.orig/debian/changelog
--- wmspaceweather-1.04.orig/debian/compat
--- wmspaceweather-1.04.orig/debian/control
--- wmspaceweather-1.04.orig/debian/copyright
--- wmspaceweather-1.04.orig/debian/links
--- wmspaceweather-1.04.orig/debian/menu
--- wmspaceweather-1.04.orig/debian/rules
--- wmspaceweather-1.04.orig/debian/watch

(thanks to a command posted by Christoph Haas in d-m for this list)

The orig.tar.gz file should be checked in as well as
long as you're not doing anything with merging external source versions.  So
I think the output is just a little strange, but ok.

I've studied better what svn-inject does (without mergeWithUpstream
active, i.e. my configuration):

1. inject upstream tarball
2. in the current workdir it's extracted the source package
3. debian/rules clean   <<---- this is the point
4. commits these changes

at point 3, there could be some deletion in the upstream files, due to
mainly tmp left in the tarball removed by clean.

This clarifies why svn-inject claims some files are replaced by others.

Cheers,
Sandro

PS: I've put d-mentors in CC (with maint of svn-buildpackeg for info)
so that the email will be searchable by someone else in need of light
on this argument...

--
Sandro Tosi (aka Morpheus, matrixhasu)
My (little) site: http://matrixhasu.altervista.org/



Reply to: