While looking at the diffs between the 261 release tags and the 26 branch, I
noticed many items in Misc/NEWS appearing in the 2.6.1 or even 2.6 sections.
I moved all of these to 2.6.2, after checking some of them, and found all of the
checked ones be backported after the 2.6.1 release. Is there anything what could
be done to avoid these wrong merges?
I plan to check the items on the 3.0 branch soon.
I'm as ready as I'll ever be -- it's time to start setting up some
Python buildbot slaves on Snakebite. I wanted to reach out and see
if anyone had any particularly strong opinions on how the slaves
should be set up.
We've got two main options: use the existing Python buildbot master
and just create new slaves for Snakebite hosts, or, set up a new
Snakebite-specific buildbot master for now, keeping it completely
separate from the existing infrastructure.
I think for at least the first couple of months it's best to assume
the Snakebite slaves will be wildly unstable/unavailable, or at the
very least, that's how I'd like everyone to view them, just so expe-
ctations aren't set too high.
FWIW, I presented the same option to the Twisted folk, and they said
they'd prefer to use their existing buildbot master, claiming that
they're used to unreliable buildbots and false-positive e-mail noti-
fication, and noting that they have something similar to our own
'stable' buildbot view that slaves are promoted to once they've dem-
onstrated they're not falling over every other commit.
(Additionally, we agreed that we could set up a 'snakebite' buildbot
view for all the Snakebite slaves, which has numerous benefits; we
could do this for Python as well.)
....and in the same vein of being as ready as I'll ever be: it's
time to set up Snakebite accounts for everyone! As I've alluded
to in the past, Python language implementation committers get ac-
cess by default.
Can you please e-mail accounts(a)lists.snakebite.org and provide
the following details:
- Full name (first, last)
- Preferred login name 
- Contact e-mail address 
- SSH key (RSA or DSA is fine): I can't enforce this, but
please make sure you use suitably secured keys, either
via passphrases or storing the private key on encrypted
media. I'd prefer the key in the OpenSSH format as it
saves me having to convert it.
: The original plan to use <first>.<last> login names lasted
about a day; unfortunately, we're at the mercy of NIS on the
UNIX side of things, and login names <= 8 characters with no
special characters (i.e. only a-z) are the lowest common den-
ominator. So, you can have whatever login name you want, as
long as it conforms to <= 8 a-z characters. For a list of
logins already taken: wiki.snakebite.org/accounts.
: This e-mail address will be added to members(a)lists.snakebite-
.org; a private e-mail list explicitly for Snakebite account
holders (analogous to python-committers@ in a way).
I just wanted to merge a series of commits to py3k where the first
added a new file, and the others changed it. Output was:
property 'svnmerge-integrated' deleted from '.'.
property 'svnmerge-blocked' deleted from '.'.
--- Merging r71409 into '.':
svnmerge: command execution failed (exit code: 1)
svn --non-interactive propdel "svnmerge-integrated" "."
svn: Attempting to delete nonexistent property 'svnmerge-integrated'
As the error message indicates, the svnmerge properties are *deleted*
in this case. Committing will delete them in the repo as well.
Doing an "svn revert -R .", deleting the added file(s) and starting
over, going through the revisions one by one, seems to work.
-----BEGIN PGP SIGNED MESSAGE-----
Sorry for the delay, but I'm now about to build 2.6.2 rc 1. Please,
no checkins to the 2.6 branch until further notice. Ping me on irc if
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (Darwin)
-----END PGP SIGNATURE-----