+ Tibby, David, Adam, Caolan, Cedric, Kendy, Kohei, Markus, Michael M,
Tsahi, Cloph, Robinson, Andras, Ahmad
* Completed Action Items:
+ setup tb3 repo on gerrit (Bjoern)
https://gerrit.libreoffice.org/gitweb?p=tb3-django.git;a=summary + mail everyone wrt. instructions on how to un-break MSVC 2012 wrt. XP (Markus)
+ attempt export from Coverity categories of non-security critical bugs (Caolan)
[ sent as CSV to the mailing list, nice to have a better way still ]
+ take German public holiday ...
* Pending Action Items:
+ find better way to export from Coverity categories of non-security critical bugs (Caolan)
+ push new default templates in & fix if necessary (Ahmad)
[ Mirek had concerns wrt. pushing these to master, need to vote on it ]
AI: if no vote is setup in a week, push anyway (Mirek, Ahmad)
+ switch to 10.8 baseline for Mac / 64bit (Norbert)
+ setup infrastructure to build bibisect repos (Cloph / Bjoern / Norbert)
+ CC all developers' mailing list on all the easy hacks (Bjoern)
+ change wiki page to eliminate proposed easy hacks (Bjoern)
+ propose different name for proposed easy hacks (Markus)
+ produce a VM for the load / crash regressions (Cloph / Markus)
+ partition document test loads / run them on tinderboxen (Norbert)
+ produce a VM with callgrind metrics (Matus)
+ transition and automate this inside a VM for Cloph (Lionel)
+ create a script that counts bug commenters stats for minor
release update PR (Bjoern)
+ send a mail to libreoffice user list for each "API CHANGE" commit (unassigned)
- as discussed in Stephans UNO API talk
[ Markus prefer to have a new script that we plug in ] (Bjoern/Stephan)
* Release Engineering update (Cloph)
+ 4.1.2 RC3 status
+ on the mirror network
+ pending press release / announce.
AI: + poke marketing / PR for announce (Cloph)
+ 4.0.6 RC1 - deadline was Monday (? ...
+ on the mirror network; 4.0.6 built with 10.6 baseline
+ stand-in for Norbert, RC1 is not compatible with 10.4 but RC2 will be
+ 4.1.3 RC1 - deadline next Monday.
* GSOC round-up / summary (Cedric)
+ sadly had to fail 2x students
+ all others did a great job, have pushed their code to Melange
+ everything green on the administrative side.
+ questionable merit of passing border-line people at the mid-term
+ Cedric still waiting for couple of lines summary / update from each student.
* Stopping the BSA from auto-setting 'regression' keyword (Markus)
+ quality of 'regression' bugs has declined significantly, too many
unconfirmed/needinfo bugs for calc.
+ bit de-motivating, to see 30% of bugs in needinfo -> because not a regression
+ in favour of removing it (Kohei)
+ used to have a special query, to pick up bugs with 'regression' keyword
+ # of bugs with 'regression' was approaching # of open calc bugs many in error.
+ a QA person should have looked at it before it is marked.
+ nice if BSA - make these suggestions - might be a regression, but not
pollute the tags, QA person has to confirm it is (Robinson)
=> not making searches unusable.
AI: + ask Rob to turn 'regression' status off (Robinson)
* Conference round
+ great to meet everyone, and attach faces to name (Adam)
+ enjoyed hackathon & presentations
+ relaxing & casual atmosphere & diverse attendance (Kohei)
+ v. motivating to all be together (Robinson)
+ concern wrt. conf. timing - to get more students there,
need a better timeline, and travel pre-approval / budget (Michael, Caolan)
+ ideally announce @ end of 1 conf. where the next will be.
+ like to see the call for next conf. locations to start ASAP.
* QA update (Robinson)
+ https://bugs.freedesktop.org/page.cgi?id=weekly-bug-summary.html +121 -97 (+24 overall) [ conference week ]
many thanks to the top five bug squashers:
Samuel M. 9
Markus Mohrhard 7
+ Bibisected regressions waiting to be addressed (some older than 1 year)
https://bugs.freedesktop.org/buglist.cgi?bug_status=NEW&bug_status=ASSIGNED&bug_status=REOPENED&list_id=348917&query_format=advanced&status_whiteboard=bibisected%2C%20bibisect40%2C%20bibisect35&status_whiteboard_type=anywordssubstr&order=changeddate%2Ccomponent%2Cop_sys%2Cbug_status%2Cpriority%2Cassigned_to%2Cbug_id&query_based_on= + 50x bugs there, some bibisected to non-useful places (before time began eg.) (Michael)
+ if there is a dead-end need to annotate / get more feedback (Robinson)
+ Private/secure space and policy for large/confidential files needed for bug repro? (Robinson)
+ difficulty of getting large documents up-loaded & exchanged - bugzilla rejects it
+ often ask to have document paired down for fixing etc.
+ currently no structure.
+ what should our policy be ? store on servers ? or ... web-upload ?
AI: + ask Florian for a minimal / simple solution larger docs (Robinson)
+ either large, or (more) confidential
+ something like pastebin (but private) for doc submission.
* Bibisect & daily builds
+ do we have windows bibisect daily builds ?
+ who has access to the daily builds / status on the Mac tinderbox builds
+ pending Norbert's return.
* QA stats:
* Open 4.2 MAB
+ 4/10 2/8 1/6 1/6 2/6 5/6 5/6 6/6 7/7 2/2 2/2 1/1
+ https://bugs.freedesktop.org/show_bug.cgi?id=65675 * Open 4.1 MAB
+ 17/99 15/96 14/93 12/91 14/91 17/89 15/85 15/81 13/77 8/68 11/64 3/55 7/52
17% 16% 15% 13% 15% 19% 18% 19% 17% 12% 17% 5% 14%
+ https://bugs.freedesktop.org/showdependencytree.cgi?id=60270&hide_resolved=1 * Open 4.0 MAB
+ 58/198 56/195 54/193 55/194 52/183 50/180 52/180 44/171 32/157 14/138 13/137
29% 29% 28% 29% 28% 28% 29% 26% 20% 10% 9%
+ https://bugs.freedesktop.org/showdependencytree.cgi?id=54157&hide_resolved=1 * Bibisected bugs open: whiteboard 'bibsected'
+ 35/143 34/142 34/141 34/140 35/140 36/167 37/137 41/137 33/128 33/125 34/124
+ http://bit.ly/VQfF3Q * all bugs tagged with 'regression'
+ 381(+7) bugs open of 2050(+16) total
* ~Component count net *
Writer - 107 (+3)
Spreadsheet - 53 (-1)
Libreoffice - 46 (+0)
Presentation - 32 (+1)
Crashes - 28 (+1)
Database - 23 (+1)
Borders - 17 (+0)
Drawing - 17 (-1)
BASIC - 4 (+2)
Migration - 3 (+0)
+ http://bit.ly/15mM2Yn - for devs ( no NEEDINFO / UNCONFIRMED )
+ https://bugs.freedesktop.org/buglist.cgi?keywords=regression%2C%20&keywords_type=allwords&resolution=---&query_format=advanced&product=LibreOffice&list_id=36764 + Migration: https://bugs.freedesktop.org/showdependencytree.cgi?id=43489&hide_resolved=1
On Thu, Oct 03, 2013 at 04:58:34PM +0100, Michael Meeks wrote:
> * Stopping the BSA from auto-setting 'regression' keyword (Markus)
> + quality of 'regression' bugs has declined significantly, too many
> unconfirmed/needinfo bugs for calc.
> + bit de-motivating, to see 30% of bugs in needinfo -> because not a regression
> + in favour of removing it (Kohei)
> + used to have a special query, to pick up bugs with 'regression' keyword
> + # of bugs with 'regression' was approaching # of open calc bugs many in error.
> + a QA person should have looked at it before it is marked.
> + nice if BSA - make these suggestions - might be a regression, but not
> pollute the tags, QA person has to confirm it is (Robinson)
> => not making searches unusable.
> AI: + ask Rob to turn 'regression' status off (Robinson)
Given the "QA person has to confirm it is" requirement, how about not
completely turning this of, but make the BSA e.g. add "possibleregression" to
- keep queries to 'regression' keyword clean for devs
- allows QA guys to query for 'possibleregression' and confirm/reject them
(thus fasttracking real regressions to devs)
Sorry for reiterating ideas if this was already suggested in the call.
> On Thu, Oct 03, 2013 at 04:58:34PM +0100, Michael Meeks wrote:
>> AI: + ask Rob to turn 'regression' status off (Robinson)
> Given the "QA person has to confirm it is" requirement, how about not
> completely turning this of, but make the BSA e.g. add "possibleregression" to
> whiteboard status?
> That would:
> - keep queries to 'regression' keyword clean for devs
> - allows QA guys to query for 'possibleregression' and confirm/reject them
> (thus fasttracking real regressions to devs)
Looks fine to me.
Thanks for all the work and thinking on better&easier QA :)
On Fri, 2013-10-04 at 11:42 +0200, Samuel Mehrbrodt wrote:
> Am 03.10.2013 17:58, schrieb Michael Meeks:
> > * GSOC round-up / summary (Cedric)
> > + sadly had to fail 2x students
> Can you please name the projects that were not completed?
I'd suggest it is best that we keep that between us & the students
involved, they know who they are :-) Nevertheless it's great to have all
the rest pass.
@Robinson: Do you have an idea if that matches words or strings?
@All: Note that this now queries for whitespace separated full words, thus:
Whiteboard Status: EasyHack, DifficultyEasy
is NOT an EasyHack anymore, while:
Whiteboard Status: EasyHack DifficultyEasy