Difference between revisions of "Talk:VCS package guidelines"

From ArchWiki
Jump to: navigation, search
(Incomplete Git function)
m (Updating a CVS repo: re)
Line 1: Line 1:
 
== Updating a CVS repo ==
 
== Updating a CVS repo ==
I don't use cvs. How can you describe the pkgver for cvs (for pacman 4.1)? --[[User:Dracorp|Dracorp]] ([[User talk:Dracorp|talk]]) 09:31, 6 April 2013 (UTC)
+
I don't use cvs. How can you describe the pkgver for cvs (for pacman 4.1)? <br>
 +
-- [[User:Dracorp|Dracorp]] ([[User talk:Dracorp|talk]]) 09:31, 6 April 2013 (UTC)
 +
 
 +
:CVS is not supported in pacman 4.1 like the other VCS tools. You will need to update pkgver manually until CVS support is added.
 +
:-- [[User:Jstjohn|Jstjohn]] ([[User talk:Jstjohn|talk]]) 22:44, 15 April 2013 (UTC)
  
 
== https:// vs git:// ==
 
== https:// vs git:// ==

Revision as of 22:44, 15 April 2013

Updating a CVS repo

I don't use cvs. How can you describe the pkgver for cvs (for pacman 4.1)?
-- Dracorp (talk) 09:31, 6 April 2013 (UTC)

CVS is not supported in pacman 4.1 like the other VCS tools. You will need to update pkgver manually until CVS support is added.
-- Jstjohn (talk) 22:44, 15 April 2013 (UTC)

https:// vs git://

Could we consider a guideline to use firewall-friendly protocols when possible (e.g. https://github.com/matplotlib/matplotlib.git instead of git://github.com/matplotlib/matplotlib.git)? --Mitch feaster 14:34, 15 November 2011 (EST)

Some leftover changes

Since Thestinger decided to protect the entire page (for whatever reason):

$SRCDEST is required for SVN

Thestinger's change (7:03, 10 April 2013 (UTC)):

→‎Subversion:  $SRCDEST not necessary

Wrong, see the note you left behind. svn export doesn't copy .svn, which means getting the version won't work anymore:

[$srcdir]$ svnversion
Unversioned directory
Thanks, reverted. -- thestinger (talk) 16:45, 10 April 2013 (UTC)

The Git functions

Danielwallace's change (04:32, 10 April 2013 (UTC)):

pkgver() {
   cd local_repo
−  git rev-list --count HEAD
+  echo $(git rev-list --count master).$(git rev-parse --short master)
}
- [...]
1142

That's good but you left the title and the sample output intact. Also, the first example still uses git describe.

--Det (talk) 16:41, 10 April 2013 (UTC)

Incomplete Git function

The following function will list not only the total number of commits, but also the tag of the last commit

pkgver() {
  cd local_repo
  echo $(git rev-list --count master).$(git rev-parse --short master)
}

1142

So the following should be used as the example title:

Using the total count of commits and the tag of the last commit.

And the example output should be changed to something like

1142.a17a017

--X33a (talk) 17:50, 10 April 2013 (UTC)

Or "tag of the last one" to prevent unnecessary repetition, but I fear even the most obvious changes like this will take a very long time now that we can't do them anymore. --Det (talk) 18:50, 10 April 2013 (UTC)
Thanks, Thestinger but the title still only mentions the first git command, while the other example should use git rev-parse as well. --Det (talk) 13:15, 11 April 2013 (UTC)
The other example should not use rev-parse. Using git describe when there's an annotated tag gets you the output that you want; rev-list --count and rev-parse --short is a workaround to simulate some of that output when there's no tag. Scimmia (talk) 18:00, 15 April 2013 (UTC)

Mercurial pkgver() broken

hg identify -ni

The above command does not change my pkgver variable of the placeholder. Can someone look into this? Additionally, I'm getting a sed warning about an unterminated substitution.

--Schala (talk) 04:25, 14 April 2013 (UTC)

Shorter and simpler Mercurial pkgver()

pkgver() {
  cd local_repo
  hg identify -ni | awk 'BEGIN{OFS=".";} {print $2,$1}'
  echo $(hg identify -n).$(hg identify -i)
}
2813.75881cc5391e

--N30N (talk) 12:21, 15 April 2013 (UTC)