On 28/06/2011 2:47 PM, Kevin A. McGrail wrote:
On 6/28/2011 2:33 PM, [email protected] wrote:
On 06/28, Kevin A. McGrail wrote:
my vote on the matter. A brand new morning in my timezone and
1140483 appears to have resolved the issue and been published. All
$ host -ttxt 1.3.3.updates.spamassassin.org.
1.3.3.updates.spamassassin.org descriptive text "1140054"
Is that expected DNS propagation delay?
I checked with sa-update using a 3.4.0 trunk svn version. So 1140483 is
a 3.4.0 ruleset and 1140054 is the 3.3.x ruleset, perhaps because
1140054 does still have the SEM issues.
Daryl, you have any input on the different versions? I would have
thought they would be the same but perhaps the packaging creates two
different versions?
As I mentioned in the flury of emails it'll be tonight before you see
the update... stable version updates are based on the rules as they were
when the mass-check that the update is based on was run.
Therefore the cycle is...
1. rules get changed
2. version gets "tagged" for nightly mass-checks
3. mass-check results get submitted and processed
4. scores get generated based on results and an update is generated a
few hours before the next "tagging" happens for the nightly mass-check
So...
1. rules were changed yesterday
2. a version was tagged for mass-checking last night
3. results have been submitted throughout the day
4. scores will be generated and an update generated this evening (in
about 6 hours) using yesterday's rules and last night's mass-check
results for those rules
For trunk updates... it's simply a mkrules of the latest svn version at
the time they are released. There's no score generation and thus no
dependency on a mass-check happening first. The mass-check is the
reason for the delay of the stable updates.
Daryl
host -ttxt 2.3.3.updates.spamassassin.org.
2.3.3.updates.spamassassin.org descriptive text "1140054"
host -ttxt 0.4.3.updates.spamassassin.org.
0.4.3.updates.spamassassin.org descriptive text "1140483"
