Jakub Wilk <jw...@debian.org> writes:
> Once we fix #560287, we could perhaps say that the directory should be:
> <packagename>-<upstream-version>-<repack-suffix>
>
> or something similar.
There isn't (to my knowledge) a defined concept of “repack suffix”.
But there is a defined concept of “Debian upstream version”; it is the
upstream version string (the version string minus the Debian release suffix)
specified in the Debian changelog entry.
I would think the correct, and expected, behaviour is to always unpack
to a directory named for the Debian upstream version.
Some examples:
* “foo (1.2.3.ds1-2)” heads the Debian changelog entry:
* The Debian release suffix “-2” says this is the second Debian
release of the same upstream source.
* The Debian upstream source is a re-packed tarball from upstream “foo
1.2.3”.
* The Debian upstream version is “1.2.3.ds1”.
* The Debian upstream source *does not match* the original, so “.orig”
is not appropriate in the directory name.
* The Debian upstream source should unpack to ‘foo-1.2.3.ds1/’.
* “foo (1.2.3~xyz-2)” heads the Debian changelog entry:
* The Debian release suffix “-2” says this is the second Debian
release of the same upstream source.
* The Debian upstream source is upstream's (say) “foo 1.2.3.xyz”,
which is *prior* to a subsequent “1.2.3” by upstream's version
naming rules.
* The Debian upstream version is “1.2.3~xyz”.
* The Debian upstream source is identical to the original, so “.orig”
is appropriate in the directory name.
* The Debian upstream source should unpack to ‘foo-1.2.3~xyz.orig/’.
* “foo (1.2.3-2)” heads the Debian changelog entry:
* The Debian release suffix “-2” says this is the second Debian
release of the same upstream source.
* The Debian upstream source is identical to upstream's “foo 1.2.3”.
* The Debian upstream version is “1.2.3”.
* The Debian upstream source is identical to the original, so “.orig”
is appropriate in the directory name.
* The Debian upstream source should unpack to ‘foo-1.2.3.orig/’.
--
\ “I don't want to live peacefully with difficult realities, and |
`\ I see no virtue in savoring excuses for avoiding a search for |
_o__) real answers.” —Paul Z. Myers, 2009-09-12 |
Ben Finney <b...@benfinney.id.au>
