KDE PIM/Meetings/KMail Stabilization Sprint 2011: Difference between revisions
Appearance
(2 intermediate revisions by the same user not shown) | |||
Line 99: | Line 99: | ||
* make optional, rather do import than full migration | * make optional, rather do import than full migration | ||
* improve mixedmaildir resource performance by applying the maildir fixes | * improve mixedmaildir resource performance by applying the maildir fixes | ||
* fix maildir to support standard nameing scheme on non-Windows | * <s>fix maildir to support standard nameing scheme on non-Windows</s> | ||
=== Nepomuk === | === Nepomuk === | ||
Line 109: | Line 109: | ||
* spam pass through filtering stops processing of further filters apparently | * spam pass through filtering stops processing of further filters apparently | ||
* tokoe is splitting it into an agent | * <s>tokoe is splitting it into an agent</s> | ||
* filter config contains ids, thus not system independent | * filter config contains ids, thus not system independent | ||
Line 120: | Line 120: | ||
* bug reports: | * bug reports: | ||
** https://bugs.kde.org/show_bug.cgi?id=281552 | ** https://bugs.kde.org/show_bug.cgi?id=281552 | ||
** https://bugs.kde.org/show_bug.cgi?id=279722 | |||
=== SQL inconsistencies for Christian Mollekopf and David Faure === | === SQL inconsistencies for Christian Mollekopf and David Faure === |
Latest revision as of 08:02, 17 September 2011
KMail Stabilization Sprint 2011
Planning in progress.
Time & Location
September 16th - 18th
KDAB Tempelhofer Ufer 11 10963 Berlin Germany Phone: +49-30-5213254-70
Getting there by public transportation:
- subway lines U1 or U6 to "Hallesches Tor"
- subway lines U1 or U7 to "Möckernbrücke"
- bus line M41 to "Hallesches Tor" (leaves from Berlin Hauptbahnhof)
Getting in:
- If the big front door is closed, there is a door bell on the right, labeled 'KDAB'.
- After passing the main gate, there are two options:
- Take the stairs on the right, go up one floor, take the door on your right and go up one more floor. You'll find a door labeled KDAB in front of your with a door bell on the right (that one might need some more pressure to trigger).
- Be lazy and go through the second big door right in front of you, turn right and take the entrance to the elevator on your right. KDAB is on the second floor.
Attendees
Attendee | Arrival | Departure | Needs Hotel | Travel Costs |
---|---|---|---|---|
Volker Krause | already there | not leaving | no | 0 |
Till Adam | n/a | |||
Kevin Krammer | Sep 15th | Sep 18th | yes/non-smoker/ok to share | |
Christian Mollekopf | Sep 16th in the morning | Sep 18th | yes/non-smoker/ok to share | ~260 euro |
Kevin Ottens | Sep 15th | Sep 18th | yes/non-smoker | ~450 euro |
Thomas McGuire | no | zilch | ||
Tobias Koenig | already there | Sep 18th | yes | 0 |
Andras Mantia | Sep 15th ? | Sep 18th ? | yes/non-smoker | ~450 euro |
Agenda
Based on http://community.kde.org/KDE_PIM/Meetings/Desktop_Summit_2011 , with already done/fixed stuff removed
(List View) Performance
- Till found the likely cause, too many (proxy) model resets on top of ETM
Migration
- make optional, rather do import than full migration
- improve mixedmaildir resource performance by applying the maildir fixes
fix maildir to support standard nameing scheme on non-Windows
Nepomuk
- merge feeders to improve scheduling
- add agateau-widgets to places needing nepomuk in kab, composer and (if needed) korg
Mail Filtering
- spam pass through filtering stops processing of further filters apparently
tokoe is splitting it into an agent- filter config contains ids, thus not system independent
Move data loss
- only unspecific and vague reports available
- originally supposed to be fixed by external payload enforcement (-> missing error handling on payload store failure)
- seems to be related to moving uncached data
- seen on local -> local move, so unrelated to net issues
- needs way to reproduce
- bug reports:
SQL inconsistencies for Christian Mollekopf and David Faure
- likely developer setup issues with mixing a lot of different setup
- tokoe was looking into making constraints more strict