Working Group Minutes/EWG 2012-03-26


IRC nick Real name
apmon Kai Krueger
migurski Michal Migurski
RichardF Richard Fairhurst
TomH Tom Hughes
zere Matt Amos


  • There was some confusion over time: EWG meetings are at 1800Z - GMT or UTC - and are not adjusted for daylight savings time. The intention was to make it easier for people in various countries to not need to figure out what other countries' DST rules are.
  • zere gave a status report on the progress of the license change code
    • the rebuild code & associated tests [1] has good test coverage for nodes and ways, with the start of coverage for relations. Relations needs most work now, but more tests for all types are welcome. RichardF is helping out.
    • work on changes to the rails_port code [2] has started, but it's not far along. apmon has started to look at the branch.


19:01 < zere> minutes of the last meeting. apologies - i haven't written the summary yet
19:01 < TomH> err...... aren't you an hour late?
19:01 < TomH> or are we stickiing to gmt?
19:02 < zere> we can wait until next week for approval of the minutes, as the summary isn't available yet...
19:02 < zere> yeah, EWG meeting times are GMT... rather than faff around with daylight savings i thought we'd keep them at the same time.
19:03 < zere> more predictable that way, given that the DST changes are always confusing.
19:03 < apmon> particularly as daylight saving is different depending on country
19:05 -!- ajturner is now known as Guest7798
19:05 -!- ajturner [~ajturner@] has joined #osm-ewg
19:05 < apmon> Whats on todays agenda?
19:06 < zere> i figured that today, license change would be looming large on the agenda. but if anyone wants to add some other business, that would be cool too.
19:09 < zere> ok. license change...
19:09 < zere> there's basically two strands of activity: the rebuild code & associated tests
19:09 < zere>
19:10 < zere> this is the code which models the actions of the license change, and it's pretty important to get that stuff right.
19:10 < apmon> Can you give a short summery as to how far along it is?
19:11 -!- ajturner is now known as Guest7800
19:11 -!- ajturner [~ajturner@] has joined #osm-ewg
19:11 < zere> the edit/redact process is going to be pretty much reversible, so we don't have to be too scared of getting it 100% absolutely correct, or covering absolutely every case, but i'd like to get as close as possible.
19:11 < zere> current status is that i think there's good coverage for nodes and ways (could always be better, though)
19:11 < zere> there's the start of coverage for relations (thanks to RichardF), but that needs more work
19:12 -!- Guest7798 [~ajturner@] has quit [Ping timeout: 480 seconds]
19:12 < apmon> With coverage, you mean test coverage, or working code coverage?
19:12 < zere> test coverage, in this case.
19:13 < TomH> there is a third strand of activity, namely whatever changes you want made to the rails port code
19:13 < zere> there's a few failing tests in the "needs clarity" case, which are ones where the interpretation is not entirely clear, and LWG has been doing some work on this.
19:14 < zere> TomH: oh, rails port changes were my second strand... "the rebuild code & associated tests" was supposed to only be one strand...
19:15 < zere> so, the second strand - modifying the API & website to support redactions. work has started here:
19:16 < TomH> yes well let me know when you consider it "done" and I'll review it ;-)
19:16 < zere> but it's not far along. i've added a migration for the new "redaction" type and some tests & methods for old_nodes, but that's pretty much it at the moment. not all the tests pass - as the redact action doesn't actually do anything yet.
19:16 < zere> TomH: i can't persuade you to help out with the implementation?
19:16 -!- Guest7800 [~ajturner@] has quit [Ping timeout: 480 seconds]
19:17 < TomH> zere: Do you have a few extra hours to add to my day?
19:17 < TomH> I'm spending all ny time helping idiots get into their accounts
19:18 < apmon> zere: I have started looking at it ( ) but so far it is more or less only a trivial commit
19:18 < zere> apmon: cool, it's a start! :-)
19:19 < apmon> A lot of tests fail for me, but I think that might be because I am using a db with the latest migrations from master, and the redaction branch might be behind
19:19 < RichardF> I can try and look at getting a bunch more relations tests in there, though I'm on WW deadline this week so won't have a whole bunch of time.
19:19 < RichardF> I don't think I'll be much use on the API/website side though ;)
19:19 < zere> sure. any help would be appreciated.
19:25 < apmon> zere: How much time do you have available to invest in finishing the migration code?
19:26 < zere> not a lot. MQ are graciously allowing me some time to work on it this week, but it would be much easier with help.
19:31 < apmon> With respect to Dermot's email about formulating a plan. Your suggestion from yesterday to wait for a complete successful test run before specifying a plan sounds most sensible
19:31 -!- migurski [] has joined #osm-ewg
19:32  * migurski sneaks in, sits down in the back of the room
19:34 < zere> i'm afraid discussion has sort of slowed down. we were discussing license change-related code and such.
19:34 < zere> but please feel free to propose something else, if there's anything you'd like to discuss
20:05 < zere> oh, sorry guys - time keeps slipping away... and, to be honest, my attention is elsewhere.
20:05 < zere> apologies & hope this hasn't put you off joining us again next week - 1800h GMT :-)