Hi, Andreas Barth wrote:
> | - the release architecture must have N+1 buildds where N is the number
> | required to keep up with the volume of uploaded packages
> The reason for this proposal should be instantly clear to everyone who
> ever suffered from buildd backlogs. :)
Well, certainly the N+1 requirement makes sense. Unfortunately it won't by
itself eliminate buildd backlogs. You also need
- backup buildd maintainers / signers
- a more sensible queueing strategy than "alphabetically".
The latter is quite trivial. Buildd admins: Please apply.
(NB: Is there a mail distributor for all the buildd admins?)
--- /tmp/wanna-build 2005-03-15 14:47:50.266938341 +0100
+++ /usr/bin/wanna-build 2005-03-15 15:00:51.000000000 +0100
@@ -147,7 +147,7 @@
{ short => "O", arg => \$list_order,
code => sub {
die "Bad ordering character\n"
- if $list_order !~ /^[psncb]+$/;
+ if $list_order !~ /^[psancb]+$/;
} },
message => { short => "m", arg => \$fail_reason },
database => { short => "b", arg => \$conf::dbbase },
@@ -223,7 +223,7 @@
$op_mode = $category ? "set-failed" : "set-building"
if !$op_mode; # default operation
-$list_order = $list_state eq "failed" ? 'fcpsn' : 'cpsn'
+$list_order = $list_state eq "failed" ? 'fcpsan' : 'cpsan'
if !$list_order;
$distribution ||= "unstable";
die "Bad distribution '$distribution'\n"
@@ -1311,6 +1311,11 @@
return $x if $x != 0;
last SWITCH;
};
+ /a/ && do {
+ $x = parse_date($a->{'State-Change'}) <=>
parse_date($b->{'State-Change'});
+ return $x if $x != 0;
+ last SWITCH;
+ };
/n/ && do {
$x = $a->{'Package'} cmp $b->{'Package'};
return $x if $x != 0;
--
Matthias Urlichs | {M:U} IT Design @ m-u-it.de | [EMAIL PROTECTED]
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
