Benny Pedersen
2011-11-22 08:28:18 UTC
Hello Jame!
17 Nov 2011 17:28, Jame Clay wrote to Benny Pedersen:
JC> Any idea what they mean by that? Because it doesn't make any sense
JC> to me...
its have to be sure that the tarballs does not change md5 sum, and if fetched
from non tarballs there is no md5 check, unless its using tagged fetchdownload
JC> Tags are on a particular commit (& can themselves _be_ a commit),
JC> which are global to a GIT repository and can be pushed or pulled
JC> to/from other GIT repositories like any other commit.
when its in that way i could use it as build source, if each tag does never
change, but it needs to be stable with content on one tag, else content is
changed, so if maintainers do make tags as tarballs then it works, otherwise it
does not
one example is how my binkd ebuild is currently done, it uses common filename
tarball to build from, but content changes, then my ebuild fail md5 checksum
Regards Benny
... there can only be one way of life, and it works :)
17 Nov 2011 17:28, Jame Clay wrote to Benny Pedersen:
JC> Any idea what they mean by that? Because it doesn't make any sense
JC> to me...
its have to be sure that the tarballs does not change md5 sum, and if fetched
from non tarballs there is no md5 check, unless its using tagged fetchdownload
JC> Tags are on a particular commit (& can themselves _be_ a commit),
JC> which are global to a GIT repository and can be pushed or pulled
JC> to/from other GIT repositories like any other commit.
when its in that way i could use it as build source, if each tag does never
change, but it needs to be stable with content on one tag, else content is
changed, so if maintainers do make tags as tarballs then it works, otherwise it
does not
one example is how my binkd ebuild is currently done, it uses common filename
tarball to build from, but content changes, then my ebuild fail md5 checksum
Regards Benny
... there can only be one way of life, and it works :)