PSC Meeting 9 May 2014¶

Proposed meeting time:¶

Friday, Mar 09, 1500 UTC

previous meeting

next meeting

Members Present:¶

• Richard Duivenvoorde
• Jürgen Fischer
• Anita Graser
• Tim Sutton
• Paolo Cavallini
• Marco Hugentobler

Agenda:¶

• Alex wants to discuss about current server setup, see http://wiki.osgeo.org/wiki/Infrastructure_Transition_Plan_2014
• Richard needs a secondant for the qgis2 setup
• Take a decision on tracker (Git*)
• What's the status of icon/symbology/style sharing web infrastructure? (Relating to Paolo's request [Qgis-developer] Adding SVG symbols)

Log¶

--- Log opened Fri May 09 16:21:38 2014
16:21 -!- jef [[email protected]] has joined #qgis_meeting_140509
16:21 -!- Irssi: #qgis_meeting_140509: Total of 5 nicks [1 ops, 0 halfops, 0 voices, 4 normal]
16:22 -!- Irssi: Join to #qgis_meeting_140509 was synced in 33 secs
16:22 < anitagraser> everyone in early today? :D
16:23 -!- timlinux [~smuxi@qgis/developer/timlinux] has joined #qgis_meeting_140509
16:24 < timlinux> hi
16:24 < anitagraser> hi timlinux
16:24 < timlinux> am I early or late?
16:24 < anitagraser> i think we're all early
16:25 < timlinux> ok I take it we start officially in >30min then?
16:25 < timlinux> just gonna make a call quick
16:26 < pcav> ok 
16:26 < anitagraser> http://hub.qgis.org/wiki/quantum-gis/PSC_Meeting_2_May_2014
16:28 < anitagraser> I have time until 17:30 CEST before I have to leave. 
16:37 < timlinux> Im ready whenever others are
16:37 < anitagraser> same here
16:37 < pcav> same
16:38 < timlinux> mhugent , jef : you around?
16:38 < jef> ack
16:38 <@mhugent> yep
16:39 < timlinux> who will chair?
16:39 < timlinux> mhugent: ?
16:40 <@mhugent> ok, let's start with first topic: Alex wants to discuss about current server setup
16:40 < pcav> should we invite alex?
16:42 <@mhugent> duiv: are you familiar with that topic?
16:43 < timlinux> I think more generally we should discuss if it makes sense to continue using telascience servers
16:43 < timlinux> or just go with hetzner one (or more)
16:44  * anitagraser doesn't know where telascience is located and what's hosted there
16:44 < pcav> could someone briefly outline the current situation?
16:44 < timlinux> telascience is free (good) but the time we spend administering it to manage free space etc is bad
16:45 < timlinux> anitagraser: qgis.org is hosted there plus downloads and the little ping file that qgis uses to see if your version is up to date
16:45 < timlinux> and a few other bits
16:45 < timlinux> I think the debs were there not sure if that is still the case
16:46 < timlinux> I think wildintellects argument is that they have very good infrastructure there
16:46 < timlinux> whereas rduiv and myself are both happy with hetzner infrastructure
16:46 < timlinux> poke duiv 
16:47 < timlinux> does someone have him on whatzapp or something?
16:47 < anitagraser> so the question is which one to choose or if we just leave it as is?
16:47 < pcav> is juergen around?
16:47 < timlinux> yes he is
16:47 < timlinux> jef: 
16:48 < pcav> jef: any opinion?
16:49 < jef> all downloads are there
16:50 < timlinux> my suggestion would be that we keep the machine but use it as a mirror
16:50 < timlinux> and move the main qgis site to hetzner
16:50 < timlinux> and have like us.qgis.org
16:51 < timlinux> and then on the downloads page add a 'choose a mirror nearby you' option
16:51 < timlinux> we could similarly set up mirrors in other parts of the world
16:51 < pcav> so no cdn?
16:51 < timlinux> or just use cloudflare or similar yeah
16:52 <@mhugent> I have to catch the train. Trying to connect via phone...
16:52 < anitagraser> sidenote: if we do that, can we hide the mirror thing from the users? at least by default? 
16:52 -!- mhugent [[email protected]] has quit [Remote host closed the connection]
16:52 < timlinux> anitagraser: actually pcav's solution is better - less work for us
16:52 < jef> cdn?
16:52 < timlinux> I use cloudflare and its dead easy
16:52 < timlinux> content distribution network
16:53 -!- mhugent [[email protected]] has joined #qgis_meeting_140509
16:53 < anitagraser> ok, never herd of
16:53 < timlinux> like akamai
16:53 < timlinux> trying to think which other ones are famous
16:53 < timlinux> anitagraser: normally as user you are unaware of them
16:53 < pcav> well, debian uses it, etc
16:53 < timlinux> they just act like a caching proxy between you and the actual host
16:54 < anitagraser> i'll look it up later ;)
16:54 < timlinux> cloudflare is free for the basic offering which is all I need
16:54 < timlinux> the one downside with both mirroring nad cdn is losing decent download stas
16:55 < anitagraser> stats weren't that great so far either. so wouldn't say that's a blocker
16:55 < timlinux> ya
16:56 < jef> probably already lost that.  we must count download redirects
16:56 < duiv> hi, sorry, was afk
16:56 < timlinux> duiv: np
16:56 < timlinux> duiv: can you see up in the logs?
16:57 < timlinux> maybe you can read them quick and then give your ideas?
16:57  * duiv reading up
16:59 < duiv> ah. I agree with doing building and all on Hetzner, then rsync to mirrors
17:00 < duiv> one thing for not doing that yet is there are a lot of redirects and some complex apache conf files which we have to copy/rewrite
17:01 < duiv> I do not want to do/fail that alone, to many descisions for me to take in too short time... blocker :-)
17:01 < duiv> cdn vs cloudflare, no opinion
17:02 < timlinux> sorry cloudflare is a cdn - sorry if that wasnt clear
17:02 < timlinux> its just one I already used and am happy with
17:03 < duiv> not sure if we need it though, if we managed to do it with one qgis server, we should manage with two :-)
17:03 < timlinux> what are the redirect rules for?
17:03 < timlinux> connecting some old legacy urls to new ones?
17:04 < duiv> yes
17:04 < duiv> wiki,
17:04 < duiv> on godaddy we did a *.qgis.org to current server, so we can split that I think...
17:06 < duiv> muninstats, aliases to api docs, download, durep (??), debian*, etc
17:06 < timlinux> ok
17:07 < duiv> we should just go through them
17:07 < duiv> and decide what to do with it
17:07 < duiv> is it difficult to 'mirror' ourselves, with just two servers?
17:08 < duiv> or is that difficult, and easier to do with cdn's?
17:08 < timlinux> its trivial to do with cloudflare
17:08 < timlinux> you just let it manage our dns
17:08 < timlinux> (update godaddy to tell it that cloudflare is our nameserver)
17:09 < duiv> and they to that for 'free'?
17:10 < duiv> or is it that they do not offer diskspace/bandwidth... just redirecting/proxy/dns magic?
17:11 < timlinux> duiv: http://i.imgur.com/JeRYaZY.png
17:11 < timlinux> you get a panel like that 
17:11 -!- mhugent [[email protected]] has quit [Ping timeout: 240 seconds]
17:12 < timlinux> and for each a record you can say if it should be cached (arrow goes through the cloud) or uncached (arrow goes around)
17:12 < timlinux> its all free
17:12 < timlinux> you pay for premium services
17:12 < timlinux> like extra security stuff
17:13 < timlinux> they can do e.g. examining the contents and rewriting it to remove stuff etc
17:13 < duiv> ok, how does it handle new builds then? You give all a ttl of 1 day?
17:14 < duiv> and... then there is no point anymore to mirror ourselves?
17:15 < timlinux> you also get a stats page like this
17:15 < timlinux> http://imgur.com/e7lDyoJ
17:15 < timlinux> I didnt play with configuration too much - I think it is mostly magic :-P
17:16 < timlinux> and they have a threat control system
17:16 < timlinux> http://imgur.com/lxDkHjK
17:16 < timlinux> where you can block hosts etc
17:17 < duiv> so this would help us to tame the crawlers and bots better?
17:17 < timlinux> so like if a spam bot is trying to ddos us (happened many times in the past to qgis.org)
17:17 < timlinux> then they block it 
17:17 < timlinux> or you can set overrides in the above panel
17:17 < duiv> well, I'm fine with it. It's just.... takes some time to DO it ...
17:18 < timlinux> I can do the godaddy + cloudflare setup
17:18 < timlinux> and put the cloudflare account details into the keepass
17:18 < timlinux> oh another nice thing is that dns changes are basically instant
17:20 < timlinux> for physical migration of stuff off telascience
17:20 < timlinux> jef you build on hetzner and copy over to telascience right now? (for debs) ?
17:21 < timlinux> and for nightlies you build at home/work and copy to telascience?
17:22 < timlinux> you can see what you get for the free versus paid cloudflare sites here:
17:22 < jef> yes, nightly debs on hetzner, windows nightly in the office and then to osuosl.
17:22 < timlinux> https://www.cloudflare.com/plans
17:22 < jef> telescience isn't used anymore, I think.
17:22 < timlinux> so that should be fairly trivial to push them to another / and additional host
17:23 < timlinux> sorry I still think of it as telascience s/telascience/osuosl
17:23 < timlinux> duiv: what else will present difficulties?
17:23 < timlinux> redmine is probably the worst
17:23 < jef> ah, and the windows nightly got to osgeo4w aka download.osgeo.org anyway.
17:23 < timlinux> suggestion:
17:24 < timlinux> jef ah yeah
17:24 -!- mhugent [[email protected]] has joined #qgis_meeting_140509
17:24 < timlinux> suggestion: leave redmine to run on its own on osuosl server
17:24 < timlinux> and move everything else for now
17:24 < duiv> +1
17:24 < timlinux> then come back in a second round to look at redmine
17:25 < duiv> my difficulty is just that I do want to do this together with somebody
17:25 < timlinux> note that if you want to run ssl on cloudflare you have to pay USD 20 a month
17:25 < duiv> happy to plan one day or part of it somewhere in next week
17:25 < timlinux> so I will do qgis.org migration with you
17:25 < timlinux> DNS + cloudflare
17:26 < timlinux> do we have qgis.org in a docker container somewhere?
17:26 < duiv> yes, we have a sphinx container, which can build docs and website
17:26 < pcav> duiv: could alex help?
17:26 < duiv> on hetzner/qgis2
17:26 < timlinux> we can first do like test.qgis.org get everything looking nice on qgis2 and then just switch qgis.org over to that same host
17:27 < timlinux> ok we can go into the details offline
17:27 < duiv> timlinux: fine. Next meeting point then?
17:28 < timlinux> Richard needs a secondant for the qgis2 setup <-- not sure what that means
17:28 < duiv> that is what I just asked
17:28 < duiv> so if either timlinux or alex is available to do some pair-setting-up it is ok for me
17:30 < duiv> timlinux: https://github.com/qgis/QGIS-Sysadmin/tree/master/docker/sphinx creates a building image for you
17:30 < timlinux> ok
17:30 < pcav> I'd agree, better if we have a backup for the mian admin
17:30 < pcav> main
17:30 < anitagraser> sorry guys, i'll have to leave. i can try to reconnect once i'm in the train. if this is still going on
17:31 < duiv> anitagraser: one point
17:31 < timlinux> ok anitagraser cya!
17:31 < duiv> never mind... jef already put this osgeo logo/link on the frontpage
17:32 < anitagraser> duiv: ok
17:32 < jef> duiv: is the site rebuilt automatically?
17:32 < anitagraser> duiv: is the script fetching meetings and blog posts broken?
17:33 < duiv> nothing is automatically currently
17:33 < duiv> anitagraser: why?
17:33 < anitagraser> can't see any
17:33 < duiv> I ran it this afternoon
17:33 < duiv> I have to figure out to do this in a way that it can be automated AND be started/stopped by other people
17:34 < jef> like when donors get added every two seconds. ;)
17:34 < timlinux> duiv: we could set up jenkins
17:34 < timlinux> and build via that
17:35 < duiv> anitagraser: I see... it now. Will run it again
17:35 < anitagraser> duiv: ok. cu later
17:35 -!- anitagraser [[email protected]] has left #qgis_meeting_140509 []
17:37 < timlinux> can we move on?
17:37 < timlinux> Take a decision on tracker (Git*) <-- paolo yours?
17:37 < duiv> nope, mine
17:38 < duiv> alex puts a lot of energy in investigating current situation of wiki/redmine
17:38 < pcav> duiv: go :)
17:38 < duiv> and finding alternatives...
17:39 < duiv> but he will not do anything untill we decide what to do
17:39 < timlinux> well my preference is for us to use github issues
17:39 < duiv> we talk a lot about moving to github, or upgrading current setup
17:39 < timlinux> but alex doesnt like that idea
17:39 < pcav> duiv: is someone manaing RM actually?
17:40 < duiv> alex tries to tweak it a little
17:40 < timlinux> I think there is an email thread somewhere with reasons why he doesnt like gh issues
17:40 < timlinux> but I like getting rid of stuff we have to manage
17:40 < pcav> I remember he considered it too simple
17:40 < duiv> me too
17:40 < timlinux> and I like using gh issues 
17:41 < pcav> timlinux: do we have an easy and safe way out of GH
17:41 < pcav> in case we want and need it?
17:41 < timlinux> it is simple (just don't know if I would put 'too' word in front)
17:41 < jef> did anyone try redmine on a 'real' box?   i/o seems to be the big problem on the vms
17:41 < timlinux> pcav: they have an api you can import / export your issues with it
17:41 < duiv> main point for opponents was the centralisation of application/plugin issues
17:41 < timlinux> I used to use it on my server jef 
17:41 < timlinux> it is always really slow
17:42 < timlinux> and it is very teadious to make and comment on issues
17:42 < jef> did you run it with a real database?
17:42 < timlinux> hmm I think I just used sqlite
17:42 < timlinux> what are we using?
17:42 < jef> mysql
17:43 < timlinux> I assume by 'real database' you meant 'one that is not sqlite' :-P
17:43 < timlinux> only real downside if gh issues is making zip attachments with e.g. shapefiles is not possible
17:44 < jef> 7z is also fine
17:44 < pcav> no attachments?
17:44 < timlinux> only images
17:44 < pcav> urgh
17:44 < timlinux> images you can drag and drop into the issue
17:44 < pcav> not so cool
17:44 < timlinux> jef: does gh take 7z?
17:44 < timlinux> I never tried it
17:44 < pcav> so we should drop all existing attachments
17:45 < mhugent> attachments are very important for small testcases
17:45 < timlinux> yup
17:45 < jef> timlinux: probably not.  no idea.
17:46 < timlinux> (yup was for mhugent)
17:46 < duiv> mhugent: don't you think people can make them available elsewhere?
17:46 < pcav> without attachments I think it' a no-go for us
17:46 < pcav> duiv: it si surely possible
17:47 < pcav> but much less safe
17:47 < timlinux> why less safe?
17:47 < jef> no dead links.
17:48 < timlinux> ok not safe as in security
17:48 < pcav> less stable I mean
17:48 < pcav> links may come and go
17:48 < pcav> original files may e deleted
17:48 < pcav> etc
17:49 < timlinux> yeah
17:49 < timlinux> so lets do nothing for now and run redmine on dedicated server
17:50 < timlinux> and move off all other services
17:50 < timlinux> maybe one day gh will allow attachments and we can raise the issue again
17:50 < timlinux> or we can build a little attachment service where users can upload the test datasets
17:50 < pcav> did we open a feature req on GH?
17:51 < duiv> yes, only images and zipped shapefiles allowed ;-)
17:51 < timlinux> pcav: i suspect they know people want it but dont want to deal with storage of gumpf
17:52 < timlinux> there is also a fork of redmine
17:52 < timlinux> I forget its name
17:52 < timlinux> I dont know if it improves it at all though
17:52 < duiv> I think alex was talking about that too
17:52 < duiv> problem is not possiblities, but hands who want to do redmine/ruby
17:53 < pcav> timlinux: I imagine
17:53 < pcav> but a reasonable limit is ok for us
17:53 < timlinux> not me I am a ruby idiot
17:53 < timlinux> pirmin was giving some support
17:54 < timlinux> maybe we can ask for more help if you have specific ideas of what we need
17:55 < duiv> do general tasks, like upgrading, tweaking, backups etc etc
17:56 < duiv> but this will be easier if we have all other things off that server...
17:56 < timlinux> ok so shall we just ask him more formally if we will be maintainer for redmine? Otherwise ask on dev list for a volunteer
17:56 < pcav> +1
17:56 < duiv> maybe we can then ask for a fresh VM and create a fresh install there then
17:56 < duiv> +1
17:56 < mhugent> +1
17:57 < duiv> who is 'him'? pirmin or alex?
17:57 < timlinux> which ever one is available and knows redmine + ruby is find for me
17:57 < timlinux> (I meant pirmin above)
17:58 < duiv> ok. After we moved other stuff to hetnzer, I'll ask both
18:00 < duiv> next point?
18:00  * duiv have to leave in about 10 mins
18:00 < pcav> What's the status of icon/symbology/style sharing web infrastructure? (Relating to Paolo's request [Qgis-developer] Adding SVG symbols)
18:00 < pcav> I think it's a bit of a shame
18:00 < duiv> I think we first have to draw a firm plan for that
18:00 < pcav> having left this in the outer space
18:01 < pcav> comments?
18:01 < timlinux> pcav: Ale's opinion was that it is not production ready
18:01 < timlinux> and needs work
18:01 < pcav> I know
18:02 < timlinux> so I have a guy starting for me on 1 June who is a django boffin
18:02 < pcav> on the other hand, Nathan always said it is ok
18:02 < timlinux> I can get him to put in a few days to review and clean it
18:02 < pcav> thanks, good news
18:02 < timlinux> it will be in second half of june only
18:03 < pcav> ok, no objections I suppose
18:03 < pcav> should we move on?
18:03 < duiv> yep
18:03 < timlinux> yup
18:04 < pcav> so
18:04 < pcav> PR management
18:04 < duiv> next time we should put the 'owners' of the agenda points next to them
18:04 < pcav> a long standing issue
18:04 < duiv> there was a discussion on the list isn't it?
18:04 < pcav> duiv: yes
18:05 < pcav> several PSC memebrs casted a +1
18:05 < pcav> mhugent has serious issues about this
18:05 < pcav> which I understand
18:05 < timlinux> pcav: I think he has good reservations
18:06 < pcav> timlinux: yes
18:06 < timlinux> one suggestion is to pay the pr assignee
18:06 < timlinux> where possible
18:06 < pcav> I do not know whether I'm pessimistic
18:06 < timlinux> so mhugent assignes to e.g. nathand
18:06 < timlinux> nathan
18:06 < timlinux> and if it helps we offer to pay a day of his time (either directly to him or to his employer)
18:07 < pcav> I'm more worried by unmanaged PRs
18:07 < timlinux> unmanaged?
18:08 < jef> https://gist.github.com/jef-n/e8f79df615379c748495
18:08 < pcav> I mean, if we say "Thanks, we found issues, and we do not have the time to review it more thoroughly" 
18:08 < pcav> this is accpetable
18:08 < pcav> just leaving them there
18:09 < pcav> may put good people off
18:09 < timlinux> most of those unassigned are quite obvious who can address them
18:09 < timlinux> e.g. labelling to dakarto
18:09 < pcav> timlinux: exactly
18:09 < pcav> thet's why I think a manager is important
18:10 < pcav> to make sure nothing slips through the cracks
18:10 < duiv> should we make it an agenda point on our meetings?
18:10 < duiv> so we now assign the one not yet assigned?
18:11 < pcav> I'm seeing this wrt plugins
18:11 < pcav> with a reasonable amount of work
18:11 < timlinux> duiv: that sounds fine
18:11 < pcav> I was able to reply to all pending plugin authors
18:11 < pcav> making them a little bit happier
18:11 < timlinux> you mean like a standing agenda point to quickly go over PR's?
18:12 < duiv> timlinux: yes
18:12 < jef> hm, in the psc?
18:13 < timlinux> mhugent: has already the PR list I think
18:14 < pcav> I do not think we can say much in the psc
18:14 < timlinux> so lets see if he is ok with your idea
18:14 < timlinux> pcav: we could just do assigning unassigned if neede
18:14 < timlinux> maybe have a system like:
18:14 < timlinux> mhugent: does them
18:15 < timlinux> anything not assigned by day of psc meeting
18:15 < timlinux> we can quickly go over and assign
18:15 < pcav> I still think we need more management work
18:15 < pcav> replying to authors
18:16 < pcav> checking that the assigee agrees in reviewing it
18:16 < pcav> etc
18:17 < duiv> timlinux: will ping you soon for the hetzner move
18:17  * duiv going afk now
18:17 < timlinux> duiv: sure
18:17 < duiv> hyo
18:17 < timlinux> ok
18:20 < timlinux> pcav: so you are looking for someone to visit each PR and just let the author know they have not been forgotten?
18:20 < timlinux> and ask them to merge their branch again if needed?
18:21 < pcav> yes
18:21 < pcav> plus, does an initial review, merge it if already sure it's safe
18:21 < pcav> etc.
18:23 < timlinux> I agree it would be good to do
18:23 < timlinux> I'm not sure who has time for it though
18:23 < timlinux> my suggestion is lets first focus on the server migration
18:23 < timlinux> then on getting the symbol sharing stuff
18:24 < timlinux> and then we can tackle the PRs
18:24 < timlinux> unless someone has time already to work on PRs
18:24 < timlinux> I want to use my spare time to finish the work I am doing on the plugin builder
18:27 < timlinux> pcav: looks like the meeting is coming to an end 
18:27 < timlinux> maybe we should wrap up for now
18:27 < pcav> timlinux: my suggestion was to pay one core dev to manage PRs
18:30 < timlinux> 1 sec
18:30 < timlinux> I just want to read mhugent's comments again
18:31 -!- anitagraserAndro [[email protected]] has joined #qgis_meeting_140509
18:32  * anitagraserAndro waves hello
18:33 < timlinux> hi anitagraserAndro 
18:33 < jef> wb
18:34 < timlinux> pcav: do you have specific cases of where someone left the project because their PR was not handled?
18:34 < timlinux> we did not even use PR's for that long
18:34 < pcav> I remeber of a Brazilian guy
18:34 < pcav> hae to look bak in the archives
18:35 < timlinux> I also maintain that if you make a PR it is not enough to simply put it online - you have to also market it
18:35 < timlinux> some devs are very good at that
18:35 < timlinux> e.g. the simplification code
18:36 < timlinux> not that I am saying it is ok to ignore PRs
18:48 < pcav> ok, so
18:48 < pcav> how do we proceed?
18:51 < anitagraserAndro> How many cases of ignored prs do we have now
18:51 -!- pcav [[email protected]] has quit [Quit: pcav]
18:51 < jef> https://gist.github.com/jef-n/e8f79df615379c748495
18:52 < anitagraserAndro> Are they being ignored because of quality or because of lack of resources
18:53 < jef> not sure. probably both.
18:54 -!- pcav [[email protected]] has joined #qgis_meeting_140509
18:56 -!- AndChat|47025 [[email protected]] has joined #qgis_meeting_140509
18:56 < pcav> sorry
18:56 < pcav> meeting ended?
18:56 < AndChat|47025> Do we distinguish between prs for adding features and those fixing bugs?
18:56 < pcav> not that i know of
18:57 < pcav> BTW: I think we still have some patches on RM hanging around
18:57 < AndChat|47025> I would be fine with throwing money at prs which fix bugs
18:57 < AndChat|47025> If money is what it takes
18:59 -!- anitagraserAndro [[email protected]] has quit [Ping timeout: 240 seconds]
19:00 < AndChat|47025> Have you asked the devs who have assignments about what's their biggest problem? Could they be doing more with more money or are they just out of time
19:05 < pcav> this is probably variable among dves
19:06 < jef> *headdesk*
19:06 < pcav> but I'm sure time is an issue for most or all
19:06 < jef> oh, ww
19:08 < AndChat|47025> Anyone got an idea what we can try or do we leave it 
19:08 < timlinux> Imm still floating around
19:10 < timlinux> I suggest we leave it as is for now - just ask mhugent to assign unassigned ones for now
19:11 < AndChat|47025> Fine with me
19:14 < pcav> ok
19:14 < pcav> have to leave
19:14 < pcav> thanks
19:16 < AndChat|47025> Bye pcav
19:17 < pcav> bye
19:17 < AndChat|47025> Anything else on the agenda?
19:25 < timlinux> AndChat|47025: no that was it
19:26 < timlinux> jef: why are you *headdesk*ing?
20:08 < timlinux> jef: this is what git pulls shows https://gist.github.com/timlinux/036114d92ef2ba9e2a78
20:09 < timlinux> yours is probably more useful since it shows unassigned
21:26 -!- mhugent [[email protected]] has quit [Ping timeout: 245 seconds]
22:39 -!- anitagraserAndro [[email protected]] has joined #qgis_meeting_140509
22:41 -!- AndChat|47025 [[email protected]] has quit [Ping timeout: 240 seconds]
--- Log closed Sat May 10 00:00:09 2014