Re: mk/git-package.mk and "--depth 1"
[email protected] (Eric Schnoebelen)
| Newsgroups | gmane.os.netbsd.devel.pkgsrc.wip.general |
|---|---|
| Message-ID | <[email protected]> |
Greg Troxel writes:
- Due to option "--depth 1" some packages based on git-package.mk
- fail to build like this
-
- fatal: dumb http transport does not support --depth
-
- Can I commit the following?
-
- Index: mk/git-package.mk
- ===================================================================
- RCS file: /cvsroot/pkgsrc-wip/wip/mk/git-package.mk,v
- retrieving revision 1.2
- diff -u -r1.2 git-package.mk
- --- mk/git-package.mk 3 Mar 2010 05:06:26 -0000 1.2
- +++ mk/git-package.mk 7 Mar 2010 12:49:58 -0000
- @@ -55,7 +55,7 @@
-
- _GIT_CMD= git
- _GIT_ENV= # empty
- -_GIT_FLAGS= --quiet --depth 1
- +_GIT_FLAGS= --quiet
- _GIT_PKGVERSION_CMD= ${DATE} -u +'%Y.%m.%d'
- _GIT_PKGVERSION= ${_GIT_PKGVERSION_CMD:sh}
-
- I am not a git black belt, but I believe '--depth 1' is intended to
- avoid copying lots of unnecessary stuff. So I wonder if it's a bug in
- git that some packages flame out, and if there should be some specific
- workaround for them instead of always fetching everything.
Yes, that was the intent.
I suppose the _GIT_FLAGS variable could have the '--depth 1'
conditionally removed if the service URL is 'http://.*' instead
of 'git://.*'.
But, given that clone wants to slurp the entire repository in
the absence of '--depth', I believe it is a good idea to limit
the amount of data brought onto the builders systme.
--
Eric Schnoebelen [email protected] http://www.cirr.com
"...Microsoft is to software what Miller (and the other big
American brewers) are to beer." -- Mark Bartelt in comp.sys.dec
------------------------------------------------------------------------------
Download Intel® Parallel Studio Eval
Try the new software tools for yourself. Speed compiling, find bugs
proactively, and fine-tune applications for parallel performance.
See why Intel Parallel Studio got high marks during beta.
http://p.sf.net/sfu/intel-sw-dev