Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Repo info
Activity
    Kate Chapman
    @wonderchook
    we have a bit of time proposals are due on April 17th
    Ian Ross
    @ian-ross
    @mojodna Message from rbanick that appeared in our Gitter room: hi field papers lovers. we're getting an "Atlas Creation Failed" error when we try to create an atlas. Could someone look into it? It's a little urgent as we have some field surveys starting tomorrow.
    Robert Banick
    @rbanick
    hi @ian-ross sorry -- internet problems at my previous location meant I thought I was posting to here
    and yes, our team is reporting some atlas creation problems
    Alan McConchie
    @almccon
    @rbanick @ian-ross @mojodna I'm about to get on a plane, but I'll try to restart the server now
    Okay, I think I just triggered a restart. Let us know if it starts working again in a few minutes. If not, @mojodna can probably debug in a couple hours.
    Robert Banick
    @rbanick
    I'm still getting an "Atlas Creation Failed" with this: http://fieldpapers.org/atlases/2zigye1o
    wah wah :(
    a couple hours is OK though, don't panic. we have backup plans.
    Alan McConchie
    @almccon
    I'm not sure if an in-process atlas would survive reboots. The depth of my knowledge ends about there. Does it work if you try to recreate an atlas with similar settings?
    Robert Banick
    @rbanick
    that's actually what I did
    sorry that wasn't clear above
    Alan McConchie
    @almccon
    oh, I see
    hmm. And I don't think there's any way to figure out the parameters that caused the failure, without diving into the database (and maybe not even then). What, roughly, were the settings you were using?
    Like, the extent, the tile layer, etc.
    Robert Banick
    @rbanick
    I had a 4x3 tile grid in a very small (500m x 500m or less) area of Sri Lanka
    standard OSM basemap
    this is more or less standard for the areas our team creates. sometimes they go a little larger
    Alan McConchie
    @almccon
    So, like something like this? http://fieldpapers.org/compose#15/7.4695/80.6253 (picking a random spot it Sri Lanka)
    So you're probably around zoom 15?
    Robert Banick
    @rbanick
    actually probably z17 for us
    sorry if I seem vague, I am always relaying requests from our team here and they are very resistant to providing details about errors
    "it is broken robert"
    Alan McConchie
    @almccon
    Like so http://fieldpapers.org/compose#17/7.47171/80.62365
    Okay, it's failing for me, too.
    Well, to be fair, the error message doesn't give your team very much info to work with! It seems like we should print out the parameters on that error page, at least.
    Robert Banick
    @rbanick
    that would be lovely! for us and you both
    Alan McConchie
    @almccon
    Actually, my z17 tests are still running... the z15 test failed pretty quickly.
    I'll see if we can get more info queued up for @mojodna, since I'm pretty sure I won't be able to fix it myself.
    Alan McConchie
    @almccon
    @rbanick I think it's the default OSM style that's failing, or is too slow. I just got it to work with satellite, with toner, and with humanitarian
    http://fieldpapers.org/atlases/2huqak3f
    http://fieldpapers.org/atlases/3uvauzxr
    http://fieldpapers.org/atlases/32diqsp2
    You might want to tell your team to use the Humanitarian style for now
    I'm not sure what's wrong with OSM right now. Now I'm going to be AFK for a bit, hopefully this helps!
    Seth Fitzsimmons
    @mojodna

    nick and i were wrestling with this in DC a couple weeks ago and basically decided that anything that uses a web-based tile service is inevitably going to have problems

    (OSM + the humanitarian style are the 2 biggest offenders)

    my inclination is to fetch extracts from Overpass that match the size of the atlas and use that to locally render the OSM style

    that way we have a bit more control over things (though not Overpass)

    Robert Banick
    @rbanick
    @mojodna that seems a pretty fair appraisal. I like the local overpass rendering fix although it does sound more labor intensive
    @almccon thanks for troubleshooting! I will send my team the tip re: humanitarian style
    Seth Fitzsimmons
    @mojodna
    This message was deleted
    Alan McConchie
    @almccon
    Hey @/all, do the scanned maps look really bad in iD on Retina screens for anybody else? I'm not sure if this is a known issue, or something new, but I don't remember noticing it before.
    For example, here's a scan that I'm testing with (it's pretty good quality): http://fieldpapers.org/snapshots/1pkbvagg#19/43.59052/-116.19452
    But when I try to use it in iD on a retina macbook, the resolution is pretty much unusable.
    Alan McConchie
    @almccon
    @/all looks like everyone's atlases are stuck right now. This seems to be happening no matter which basemap we choose.
    So this seems to be a different problem than the one that was happening previously, where the default OSM style was getting throttled.
    Alan McConchie
    @almccon
    Rebooted. But I guess it could still be the same problem: the atlases using the OSM default style eventually fill up FP's queue, preventing any new atlases from generating, no matter what basemap they use.
    Kate Chapman
    @wonderchook
    @almccon would it make sense to have a different default style that might not have that issue?
    Alan McConchie
    @almccon
    @wonderchook hmm, maybe we just switch to the HOT style as the default, which should work fine in most cases where people don't have a preference?
    Kate Chapman
    @wonderchook
    I think that would work, might be worth mentioning to HOT or OSM-fr I think hosts it.
    Kate Chapman
    @wonderchook
    hi all! Cadasta is going to participate in Outreachy and hopefully Google Summer of Code this next round. I was wondering if anyone would be interested in mentoring any field papers type internships. One possible GSoC one would be to make it so larger format scanning is possible. I guess it would require creating an android app or finding one we could use to scan a large plotter size piece of paper and then making that work in the OCR
    Alan McConchie
    @almccon
    @/all There's someone trying to work on Field Papers for the Mozilla Global Sprint happening today and tomorrow. Anyone on hand to answer a few questions? Unfortunately we don't have any capacity (or much knowledge, frankly) at Stamen to help out currently.