Working Group Minutes/EWG 2014-04-28

Attendees

IRC nick Real name
gravitystorm Andy Allan
RichardF Richard Fairhurst
zere Matt Amos

Summary

  • OSM100
    • gravitystorm has a new project which "aims to get 100 new developers into OSM in 2014."
    • There's a repo for software to track this: [1].
    • Current output of the tool: [2].
    • RichardF wonders about mobile development and gravitystorm points to [3], which is still in a highly unfinished state.
    • zere wonders about whether most contributors are "single-shot" (i.e: scratch an itch and move on), or "low impulse" (i.e: contribute over an extended period of time, but infrequently). This would affect how we try to attract more developers.
  • RANDA
    • The swiss OSM chapter was approached by these guys: [4] about having an "OSM stream" as part of this year's week-long hackathon (Saturday, 9th to Friday, 15th of August 2014). It sounds like they are really looking for an organiser to ensure the OSM stream would be productive.
    • There was no interest.

IRC Log

17:31:17 <gravitystorm> howdy everyone.
17:31:33 <gravitystorm> I have one topic for today's meeting - osm100 - does anyone else have any others?
17:31:56 * RichardF is here but has no topics
17:33:19 <gravitystorm> OK, since there's at least 2 of us I'll get started. I've created a new project, called 'osm100'. This aims to get 100 new developers into OSM in 2014.
17:33:38 <gravitystorm> I've made a tool to measure success - https://github.com/gravitystorm/osm100
17:34:10 <gravitystorm> So far this year we have 9 new committers, so it seems like a lot to get 100 in the next 8 months.
17:34:27 <gravitystorm> but we actually had 98 new committers in 2013 :-)
17:35:15 <gravitystorm> so I'm currently looking at 1) how come we had so many in 2013 - which projects attracted them? and 2) what comes next for 2014?
17:35:21 <gravitystorm> Anyone have any thoughts?
17:36:25 <gravitystorm> https://gist.github.com/gravitystorm/11378778 is the current output of the tool, if you are interested.
17:37:09 <RichardF> interesting read. iD and osm-website clearly the leaders.
17:37:14 <RichardF> surprised JOSM is so low.
17:38:43 <gravitystorm> I suspect a lot of JOSM is missed since it's only looking at the core, not the plugins
17:38:57 <RichardF> ah, ok.
17:39:57 <gravitystorm> I was considering doing some cohort analysis too. The class of 2006 all contributed in 2007, but beyond that it's harder to tell. But then I wondered if I was just distracting myself from tackling the real problem - getting more developers.
17:41:49 <RichardF> I've been idly wondering whether we should be trying to attract more mobile developers to build surveying/editing tools for OSM. it seems to be a natural fit yet one where we're largely lacking. but I'm not entirely sure where we'd start.
17:45:01 <gravitystorm> finishing off http://code4osm.org could be a start - then we can highlight mobile projects on there
17:46:26 <gravitystorm> but, as ever, zere and I are busy with many things, so we need some help with that. If only we could ask CWG to help :-)
17:49:16 <zere> any help would be appreciated, of course.
17:50:43 <zere> (also, +topic about "randa")
17:51:36 <gravitystorm> any more comments on osm100?
17:52:19 <zere> i would be interested in the lifetime vs number distribution
17:52:59 <gravitystorm> zere: as in a histogram of lifetime number of commits per contributor?
17:53:08 <zere> i.e: what proportion of commits / new contributors are fixing a single thing, and what (presumably smaller) proportion go on to become long-term contributors
17:53:28 <gravitystorm> I suspect it'll be like this:
17:53:28 <gravitystorm> 1 [                                         ]
17:53:36 <gravitystorm> 2 []
17:53:39 <gravitystorm> 3 |
17:53:45 <gravitystorm> etc
17:54:26 <zere> sure, but i wonder if the 2 is 2, each a day apart, or 2 several months apart
17:55:01 <gravitystorm> ah, ok
17:55:27 <gravitystorm> I'll look into that
17:55:35 <zere> trying to figure out if there are "low intensity" contributors (in which case, worth having more), or "single impulse" contributors (in which case, might be better spending effort elsewjere)
17:56:31 <gravitystorm> yep
17:56:56 <zere> are we teaching men to fish - but they just don't feel the need to fish very often, or are we giving them a fish every now and again ;-)
17:57:04 <gravitystorm> #topic randa
17:57:18 <gravitystorm> zere: go for it.
17:57:28 <gravitystorm> (then I can head to the pub :-) )
17:57:56 <zere> the swiss OSM chapter was approached by these guys: http://randa-meetings.ch/ about being part of this year's week-long hackathon
17:58:17 <zere> Saturday, 9th to Friday, 15th of August 2014
17:58:55 <zere> it seems to be mostly KDE folks, but i think they have something set aside for an "OSM stream" if anyone is interested in attending
17:59:57 <zere> unfortunately, the deadline approaches quickly (4th may), and we would need to tell them pretty soon if anyone wanted to come and/or organise something.
18:00:00 <zere> so...
18:00:04 <zere> any interest?
18:01:45 <gravitystorm> Not for my personal attendance. Is it worth taking it to the mailing lists? Would we need some kind of filtering of volunteers or are they happy for anyone to turn up?
18:03:28 <zere> i think in general they're happy for anyone to register, but they wanted to know if anyone was going to be doing something related to OSM. given it's a themed hack week, i guess they don't go in for "just turn up and do whatever" (i.e: zere-style) ;-)
18:05:03 <zere> if there's no interest here (and it seems like not) then i'll see what the SOSM folks want to do - perhaps they'll advertise it on their mailing list, i don't know.
18:05:42 <zere> it sounded like they were really looking for an organiser to ensure the OSM stream was productive.
18:07:54 <gravitystorm> I think given out lack of interest/capacity, it's worth getting back to SOSM as you suggest
18:08:40 <zere> yup, ok.
18:08:59 <gravitystorm> So I'm out of topics for today. Anything other business?
18:09:52 <zere> not from me (na-na-na-naaa)
18:10:04 * zere hums
18:10:48 <gravitystorm> quite