Humanitarian OSM Team/Working groups/Technical/meeting 2013-07-29

From OpenStreetMap Wiki
Jump to navigation Jump to search

IRC log

12:53:24 PM mkl: hi
12:55:48 PM ian29: hello. just listening here.
12:58:15 PM mkl: hi ian29
12:58:26 PM mkl: who's in for the hot tech meet?
12:58:54 PM ***harry-wood is here
12:59:11 PM harry-wood: here's what we talked about last time:
12:59:54 PM ***dodobas waves
01:00:43 PM mkl: hey dodobas!
01:01:00 PM dodobas: hi mkl
01:01:47 PM harry-wood: mkl did you see I blurted out some thoughts here:
01:01:48 PM flavour [] entered the room.
01:01:48 PM mkl: anyone else here?
01:01:58 PM flavour: y
01:02:13 PM pierzen: hi
01:02:21 PM clara: sorry - I can'n make it today
01:03:03 PM mkl: clara: any quick updates?
01:03:15 PM mkl: harry-wood: yes, reviewing again :)
01:03:15 PM clara: I've got access to the drupal site since a few days, but didn't have time to do anythin
01:03:27 PM mkl: are you on to sprint?
01:03:38 PM clara: mkl: still up for doing birthday sprint
01:03:45 PM mkl: ok right on, thanks
01:04:10 PM mkl: ok give me 1 minute and we'll start
01:05:55 PM mkl: Ok
01:06:01 PM mkl: So first a couple announcements
01:06:23 PM mkl: Birthday Sprint is ON, plenty of projects
01:06:24 PM mkl:
01:06:32 PM mkl: plenty to talk about this hour
01:06:38 PM mkl: next
01:07:05 PM mkl: very happy to announce that dodobas has come on as HOT's sysadmin
01:07:19 PM mkl: effective ... immediately?
01:07:25 PM dodobas: hopefully :)
01:07:47 PM mkl: note, he's here to listen and learn, but not to fulfill a ton of requests.
01:07:53 PM mkl: let's call them ideas :)
01:08:14 PM mkl: welcome dodobas
01:08:27 PM pierzen: hi dodobas
01:08:47 PM dodobas: hi pierzen
01:08:56 PM mkl: so should we do a round of check ins
01:09:30 PM mkl: in order of appearance: dodobas, harry-wood, flavour, pierzen
01:10:16 PM ***ybon in a meeting, ping me if you need some info from me
01:10:28 PM mkl: k
01:10:41 PM mkl: dodobas: anything you want to say about where you're at?
01:10:55 PM kdanielson [] entered the room.
01:11:10 PM mkl: harry-wood?
01:11:35 PM dodobas: well, i still don't have any access to the servers
01:11:46 PM mkl: which servers?
01:11:55 PM harry-wood: Aha! Therein lies the problem :-)
01:12:08 PM dodobas: and i don't know who is responsible... or who to contact
01:12:11 PM clara left the room (quit: Ping timeout: 480 seconds).
01:12:18 PM kdanielson: (ohai guys-- just lurking, carry on)
01:12:19 PM dodobas: mkl: well, all of them :)
01:12:24 PM mkl: we started collected them here :
01:12:35 PM mkl: I don't think it's comprehensive
01:12:39 PM mkl: what are we missing people?
01:12:41 PM rbanick [] entered the room.
01:12:58 PM rbanick left the room (quit: ).
01:12:59 PM mkl: (hi kdanielson, rbanick)
01:13:03 PM mkl: (oop)
01:13:21 PM harry-wood: I think along with the idea of hiring a sysadmin we always assumed that idea of having some official HOT servers. Or at least one.
01:13:23 PM banick [] entered the room.
01:13:34 PM harry-wood: but it's something we wanted to discuss
01:13:39 PM mkl: the only services we have listed are the tasking manager
01:13:45 PM springmeyer: welcome and congrats dodobas :)
01:14:00 PM mkl: there's also at least the hot export, and hot website
01:14:12 PM dodobas: oh springmeyer, you see, you can't just simply get rid of me :)
01:14:33 PM mkl: is exports hosted by geofabrik?
01:14:35 PM springmeyer: dodobas: this is good
01:14:37 PM mkl: what about the tasking manager?
01:14:52 PM mkl: the website is hosted on pantheon right now
01:14:58 PM dodobas: so, do you know where are services hosted ?
01:15:05 PM harry-wood: pantheon?
01:15:06 PM springmeyer: mkl: yes on exports is hosted by geofabrik as far as I know
01:15:13 PM hjart [~hjart@] entered the room.
01:15:29 PM mkl:
01:15:40 PM mkl: mike mccaffrey is paying for the account
01:15:50 PM harry-wood: Yeah I have access to the hot-export server. It's a virtual machine geofabrik are providing
01:15:50 PM mkl: should definitely be switched soon
01:16:20 PM mkl: harry-wood: can you capture these in the wiki quickly ... service, url, host, contact?
01:16:33 PM harry-wood: I can for hot-export yeah
01:16:41 PM mkl: are there any others we need to think of? right now we have export, website, tasking server
01:17:26 PM pierzen: LearnOsm?
01:17:40 PM mkl: LearnOSM ... that's hosted all on github I think
01:17:53 PM pierzen:
01:18:04 PM dodobas: what about MapMill ?
01:18:10 PM mkl: yes, even
01:18:35 PM mkl: that's gitpages hosting
01:18:40 PM mkl: mapmill ... we'd need to ask schuyler
01:18:46 PM harry-wood: dodobas yes mapmill is a good one to investigate actually. I guess schuyler knows about this more
01:18:49 PM pierregiraud [] entered the room.
01:18:55 PM pierregiraud left the room.
01:19:03 PM pierregiraud [] entered the room.
01:19:21 PM harry-wood: this is broken
01:19:30 PM mkl: so let's at least capture this list, and the information we do have, in the wiki ...
01:19:39 PM mkl: yea, that sucks, cause i often mention it
01:20:07 PM mkl: dodobas: besides getting access to all these services, anything more to bring up, or should we move on to harry-wood
01:21:28 PM dodobas: mkl: after getting the access i can start collecting information on them
01:21:33 PM mkl: ok perfect
01:21:40 PM mkl: harry-wood, do you want to check in, or pass?
01:21:43 PM dodobas: and add it to the wiki
01:21:47 PM harry-wood: I will request access to the hot-export VM for you dodbas
01:22:16 PM mkl: I made some quick notes, but it needs to be filled in more:
01:22:30 PM dodobas: harry-wood: do you need my DNA sample ? :)
01:22:56 PM pierzen: plus fingerprint and pict :)
01:22:59 PM harry-wood: hehe. No but… a public key
01:23:30 PM harry-wood: Frederik (geofabrik) set me up with ssh access given my public key
01:23:49 PM dodobas: k, it's like a DNA... not sure if I got your mail...
01:24:02 PM dodobas: i can probably scan the mailing list...
01:24:18 PM harry-wood:
01:24:23 PM mkl: ok, flavour. want to check-in?
01:24:25 PM dodobas: k...
01:24:32 PM harry-wood: I have no further updates
01:24:58 PM mkl: k
01:24:58 PM dodobas: mkl: one more thing... there was a mail on the list (25.7) -> HOT - EXPORT service for Nepa
01:25:00 PM flavour: Not much to report here...still working on Sahana which includes interfaces with OSM & is used by Hums
01:25:06 PM flavour: No especial news today
01:25:29 PM mkl: thanks flavour
01:25:43 PM mkl: dodobas: hot export for Nepal ... meaning adding coverage for Nepal
01:25:50 PM harry-wood: dodobas yeah good point. I'm actually not sure what the procedure should be for processing that kind of request
01:26:00 PM mkl: harry-wood: ask Frederik for guidance
01:26:09 PM dodobas: harry-wood: exactly...
01:26:09 PM mkl: he can at least explain the process to us
01:26:19 PM harry-wood: Yes. I'll put the question to him
01:26:30 PM harry-wood: at the same time as getting access for dodobas
01:26:36 PM mkl: right on
01:26:44 PM mkl: pierzen, you're up
01:26:48 PM pierzen: LearOsm Friendly Editing / Translations tool
01:26:48 PM pierzen: Functions that should be simple to do without using a F1 car to go to the nearby store :
01:26:48 PM pierzen: - Add a new file in a directory and edit
01:26:48 PM pierzen: - wysiwyg results
01:26:48 PM pierzen: - upload images simply
01:26:49 PM pierzen: - see final result
01:27:24 PM pierzen: Tools adapted to contributors that do not use github regularly
01:27:40 PM mkl: pierzen: i think this is about getting two things
01:27:51 PM mkl: one is workflow documented for translators
01:28:07 PM mkl: another is someone more technical to lead the process, handle the git side of things
01:28:29 PM pierzen: does it take care of creating new file and add to the repository, does it import images?
01:29:05 PM mkl: i think so, checking
01:29:10 PM pierzen: or is this only a wysiwyg editor of text?
01:29:22 PM mkl: new file, yes
01:29:25 PM mkl: image upload, yes
01:29:53 PM mkl: wysiwyg: no, it's markdown
01:29:59 PM mkl: but you can preview at any moment
01:30:00 PM pierzen: We still need simple and clear infos, which branch to work on, how to pull, still waiting for this.
01:30:17 PM mkl: yes that's true
01:30:20 PM skorasaurus [] entered the room.
01:30:28 PM mkl: we need someone technical to take the handles of git for learnosm
01:30:32 PM pierzen: If this is simple, 1-2 pages of tutorial with examples.
01:30:34 PM mkl: document the workflow
01:30:49 PM mkl: so you are saying, we need a LearnLearnOSM
01:31:03 PM pierzen: edit workflow
01:31:09 PM skorasaurus: mkl, just quickly jumping in since i'm on lunch break.
01:31:24 PM skorasaurus: yeah, some of our guys are having trouble with git, which is understandable.
01:31:37 PM mkl: skorasaurus: they really shouldn't have to worry about it
01:32:03 PM mkl: there should be a single tech lead for each translation
01:32:04 PM pierzen: when you translate and wait 4 days to see your results with images then, ...
01:32:23 PM mkl: (could be the same person for multiple translations)
01:32:36 PM mkl: they then add translators git accounts to that fork
01:32:49 PM mkl: translators can then edit in
01:32:52 PM skorasaurus: mkl: that's a good idea, just a matter of someone finding the time. :)
01:33:01 PM mkl: results are visible within 10 minutes on
01:33:32 PM mkl: when ready, lead for the translation would make a pull request to the main learnosm
01:33:35 PM pierzen: my pull request were refused a few days ago. I still wait to see the result.
01:33:38 PM mkl: then it's accepted goes live
01:33:41 PM mkl: refused?
01:33:55 PM pierzen: yeap, using the wrong branch, I was told.
01:34:05 PM pierzen: I wait for better instructions,
01:34:07 PM mkl: right
01:34:10 PM harry-wood: I wonder if it would it help if placeholder files were put in the place. Make's the workflow easier?
01:34:18 PM mkl: so who has the time?
01:34:26 PM mkl: harry-wood, what do you mean by that?
01:35:07 PM harry-wood: Well if we want a complete french translation, we could start by just copying all the english files
01:35:13 PM harry-wood: does that make it easier? not sure
01:35:45 PM pierzen: it is done. Then when I update and try a Pull request, the fun is re-starting, again and again andd ...
01:35:45 PM mkl: isn't that what happens when a branch is made?
01:35:57 PM mkl: pierzen shouldn't be making pull requests :)
01:36:03 PM mkl: we need someone to take the lead on this
01:36:04 PM pierzen: As I said, clear instructions.
01:36:08 PM mkl: anyone?
01:36:20 PM mkl: ok, i'm noting it on the wiki. big need here.
01:36:20 PM dodobas: if I understand the process correctly, idea is to let someone create a fork of them gh_pages branch, and then create a pull request, which would be 'bindly' merged into main repo
01:36:38 PM skorasaurus: I'd like to but I seriously don't have time at the moment.
01:36:53 PM mkl: dodobas: since it's in a translation directory, should be pretty self contained
01:36:58 PM pierzen: skorasaurus is eating too much!
01:37:12 PM skorasaurus: :p
01:37:19 PM harry-wood: I'm not much of git expert. Maybe I can pierzen though, if he's even less of a git expert :-)
01:37:20 PM dodobas: by blindly, i mean.. just make sure that the directory structure is sane and that it doesn't overwrite any existing binary files
01:38:14 PM pierzen: one time we talk about wrong directory, you change, next time you are told you dont synchronize, etc,
01:38:29 PM pierzen: this is why I say we need a complete recipe of the workflow.
01:38:54 PM mkl: ok
01:39:09 PM mkl:
01:39:12 PM dodobas: is there a written summary of the current workflow ?
01:39:19 PM mkl: dodobas: no
01:40:07 PM harry-wood: I gotta go I'm afraid
01:40:16 PM mkl: ok harry-wood
01:40:21 PM mkl: pierzen: anything more to check in on
01:40:29 PM mkl: this need is duly noted, and i agree, high priority
01:40:38 PM pierzen: If we want to bring more translators, not all experimented with tools such as Github, we need to find simple solutions.
01:40:43 PM pierzen: that's all.
01:41:12 PM banick: just to break my lurking, I agree...
01:41:34 PM banick: I've been thinking of bringing on volunteers to do translation, but the workflow is muddlesome and I don't have time to teach
01:41:43 PM harry-wood left the room (quit: Quit: Leaving.).
01:41:49 PM mkl: it's what we got
01:41:54 PM mkl: it can be made simple
01:41:55 PM kdanielson: +1 "muddlesome"
01:42:02 PM mkl: but someone needs to put in the work to document
01:42:04 PM mkl: ok
01:42:24 PM mkl: checkins: ybon, springmeyer, kdanielson, rbanick, skorasaurus
01:42:26 PM mkl: ybon?
01:42:58 PM mkl: springmeyer?
01:43:27 PM banick: If I could get someone to sit down and explain with me and my intern in DC then I could get him to document
01:43:30 PM springmeyer: nothing to add except I'm around if anyone needs something
01:43:45 PM mkl: springmeyer: see above :)
01:43:54 PM mkl: banick: we could schedule something
01:43:56 PM springmeyer: k, skimmed quick, must of missed it
01:44:14 PM banick: mkl: awesome blossom, let me know. I'm out in 2 weeks, so before then is extra bueno.
01:44:24 PM mkl: birthday sprint?
01:44:32 PM dodobas: springmeyer: learnosm new translators hand holding guide
01:44:34 PM kdanielson: (Aug 10-11)
01:45:35 PM mkl: banick: let's sync up on that. email me and your intern and we'll find a time. maybe loop alex in for good measure
01:46:01 PM ybon: humm, nothing particular to add that can be expressed in the time I have ;)
01:46:09 PM ***springmeyer knows little of learnosm actually, so not helpful there
01:46:13 PM ybon: About learnosm, I've expressed my opinion on the ML
01:46:51 PM mkl: ybon, that's more about the advanced chapters, right?
01:47:12 PM mkl: i don't think we're even ready to get there yet. still gotta nail down workflow for the beginning chapters :)
01:47:32 PM mkl: kdanielson: do you want to check in? anything to contribute on the birthday sprint?
01:47:44 PM kdanielson: Just a heads up that it's happening:
01:48:05 PM kdanielson: it seems to be turning into a HOT-heavy event (lolz), which is fantastic
01:48:29 PM mkl: btw, Mike McCaffrey offered to contribute a mockup of the cross-site tool bar
01:48:38 PM kdanielson: but I know this morning a few folks were volunteered as sprint leads and I wanted to make sure that was cool with them... not sure if any of them are on the chat atm
01:48:43 PM kdanielson: great
01:49:11 PM pierregiraud: oh, someone want to integrate iD in the tasking manager?
01:49:11 PM mkl: yea, i don't think any of those folks are here
01:49:17 PM ***pierregiraud didn't know
01:49:27 PM mkl: pierregiraud: it's part way done
01:49:34 PM pierregiraud: oh really?
01:49:44 PM pierregiraud: interesting
01:49:53 PM mkl: the iD button is there
01:50:01 PM pierregiraud: oh
01:50:06 PM mkl: it doesn't open with the tile config however
01:50:10 PM mkl: but that should be easy to fix
01:50:11 PM banick: yep
01:50:27 PM pierregiraud: I was thinking about something more intrusive
01:50:47 PM mkl:
01:50:58 PM mkl: oh you mean, right there in the browser?
01:51:07 PM mkl: that's possible. fieldpapers now has that
01:51:33 PM mkl: we at least should get the button fully working, should be an easy win for someone
01:51:52 PM pierregiraud: understood
01:52:26 PM mkl: pierregiraud: anything more to check in?
01:52:54 PM pierregiraud: unfortunately not
01:52:56 PM mkl: i think only skorasaurus is left after that
01:53:19 PM pierregiraud: I haven't worked on the tasking manager for months
01:53:43 PM pierregiraud: though there are many issue to address
01:54:19 PM banick: I have a small update
01:54:24 PM mkl: any other top issues? could be offered up for the sprint
01:54:26 PM mkl: yes, banick
01:54:40 PM banick: We've been worknig with ybon to refine a JD for a volunteer to help develop uMap
01:54:59 PM pierregiraud: JD?
01:55:00 PM banick: looking to see if the Red Cross name is enough to pull in someone who wouldn't otherwise know of / take an interest in the project
01:55:02 PM banick: it might not work
01:55:11 PM banick: JD = job description, just a loose term
01:55:31 PM mkl: job or volunteer?
01:55:35 PM banick: volunteer
01:55:41 PM banick: VD has bad connotations :)
01:55:48 PM dodobas: what is uMap ?
01:56:16 PM banick: We're basically trying to figure out if the Red Cross name works here like it does for other volunteer positions. Like I said, it might not work. ybon's been providing very helpful feedback.
01:57:40 PM mkl: that would be cool
01:57:54 PM mkl: seems to work well. only thing for me, i find the UI a bit confusing
01:58:19 PM banick: our feelings also. We want to figure out if we can find a way to print...possibly externally hosting a print solution ourselves on an RC server
01:58:40 PM banick: anyways the JD is up here:
01:58:47 PM mkl: anyone notice this ?
01:58:51 PM banick: feel free to do the normal github pull / push shuffle
01:58:52 PM mkl: i see a merge possibility
01:59:19 PM skorasaurus: checking in regarding ?
01:59:21 PM banick: we were thinking about that would make sense
01:59:47 PM pierzen: one more point : Nicolas is reporting a lot of problems with Field teams in Africa, with bad internet connection.
01:59:49 PM mkl: skorasaurus: we've just been structuring the meetings with a go round of everyone present
02:00:02 PM mkl: if you have nothing to add, jsut pass
02:00:07 PM mkl: pierzen: yes?
02:00:20 PM pierregiraud: mkl: yes, another MapBox (tmcw) neat tool
02:00:39 PM pierzen: we have difficulty communicating plus they have difficulty with the osm workflow in such a context.
02:01:11 PM pierzen: skorasaurus: and ybon: worse then haiti, this is possible !
02:01:57 PM mkl: what can be done? besides getting them a set of satellite phones (not happening ;)
02:02:19 PM skorasaurus: oh sry, was away for a few minutes, screen didn't scroll down.
02:02:20 PM pierzen: As discussed before, there could be various ways we could look at.
02:02:21 PM banick: do they have mobile hotspots? those can pull in a bare bones connection sometimes. we had success with them in uganda
02:02:22 PM ***skorasaurus catching up.
02:02:33 PM skorasaurus: banick: yes, they do.
02:02:47 PM skorasaurus: but even then, it's only advertised as 3g. :/
02:03:02 PM mkl: we need BRCK
02:03:11 PM mkl: but that's 6 months away
02:03:30 PM mkl: 3g is definitely good enough to map
02:03:35 PM pierzen: and for the next 6 months we have 4 projects in Africa.
02:03:52 PM pierzen: to support.
02:04:58 PM banick: balls
02:05:06 PM mkl: how can we do mapping without internet?
02:05:16 PM banick: I think its "3G" in quotes to be honest
02:05:25 PM banick: single G is more realistic
02:05:26 PM pierzen: that is the 2cents question :)
02:05:52 PM skorasaurus: yeah, that's what I was meaning ;)
02:05:53 PM banick: I think this is hitting larger workflow questions
02:06:13 PM banick: What's our offline workflow? it needs to be separate
02:06:29 PM kdanielson left the room (quit: ).
02:06:38 PM banick: render everything locally for a while and then push up all at once when we get connectivity?
02:06:55 PM pierzen: yes, If people have local imagery, then can work offline and upload later.
02:07:06 PM mkl: we can't cache bing
02:07:23 PM mkl: but people can map with gps, right?
02:07:30 PM mkl: that's how we used to do it in the old days
02:07:51 PM pierzen: not as productive as imagery let's say.
02:07:56 PM dodobas: mkl: even if you use an internal proxy... like some large companies do ?
02:07:56 PM skorasaurus: *cough* legally, we can't cache, as far as i know *cough*
02:07:57 PM pierzen: It takes more time.
02:08:07 PM dodobas: to temoprarily cache imagery ?
02:08:28 PM mkl: yea, can't even do that. it violates agreement with bing. we're lucky to have it, even this way
02:08:34 PM dodobas: ?temporarily
02:08:45 PM banick: I agree with mikel, we should be 100% respectful of our bing agreement
02:08:53 PM mkl: would you like to speak to msft's lawyers about it? ;)
02:09:29 PM pierzen: Other imageries to look at in such context?
02:09:36 PM banick: what about rendering our GPS traces etc locally?
02:10:00 PM banick: using the HDM or some such, then hosting locally
02:10:14 PM pierzen: If you are teaching people how to use imagery and you have no Imagery, you start to have a problem.
02:10:15 PM banick: it's not imagery but it's at least a little clearer what's being collected and entered
02:10:20 PM mkl: a lot of geometries can be completed remotely. the key thing you can't get, actually being there, is what's under the roof
02:10:43 PM mkl: why are we teaching imagery when there's no access to it?
02:10:47 PM mkl: aren't there gps units down there?
02:11:00 PM mkl: (btw, 1 hour is up, so we can call the official meeting closed)