[HOT] TM3 and validation

john whelan jwhelan0112 at gmail.com
Tue Oct 24 17:01:13 UTC 2017


>I had one problem when validating multiple tiles of a user in JOSM today.
I clicked the button to download the data to JOSM but I didn't get any OSM
Data. I just got the imagery and the TM-Tile-Borders, but the
"OSM-Data"-Set was empty. I guess this is not what it should be? In single
tile validation I get the OSM-Data.

I use hovering over the start button to identify the tiles then validating
the tiles individually.

As Michael has commented trying to slippy map the area and download more
than one tile is awkward and its easier to validate one tile at a time.
Unfortunately when you download tiles they go to a new layer so you can't
do checks across all the tiles at once.

Cheerio John

On 24 October 2017 at 12:34, Florian Kratochwil <florian at kratochwil.at>
wrote:

> After checking out the new TM as validator I can say: Respect! I like it!
> It is way better than before and I am looking forward to validating.
>
> I had one problem when validating multiple tiles of a user in JOSM today.
> I clicked the button to download the data to JOSM but I didn't get any OSM
> Data. I just got the imagery and the TM-Tile-Borders, but the
> "OSM-Data"-Set was empty. I guess this is not what it should be? In single
> tile validation I get the OSM-Data.
>
> And I am curious what is behind the random task selection for validation.
> It is really random or is there some algorithm behind which gives me users
> which
>  - haven't been validated recently or
>  - have been invalidated recently or
>  - are beginners or
>  - have finished the tile just minutes ago or
>  - ...?
>
> I really was waiting for this button, because now validators can focus on
> validating without browsing for tiles. I think I will use this button a lot.
>
> Cheers,
> Florian
>
>
> Am 2017-10-21 um 22:27 schrieb john whelan:
>
> I'm after the most recently mapped tile because that one is the way they
> are currently mapping and that's what I'd like to correct.
>
> If I get a name that gives me one or more tiles.  Which is the most
> recent?   If they are mapping in a group they may well have corrected
> themselves or even by themselves they may have experimented then read the
> documentation and corrected themselves.
>
> I've had mappers with three of four years experience who have benefited
> from feedback.  One was using highway=track rather than
> highway=unclassified because he once put in name=unclassified in JOSM got
> an error there after used highway=track.  English even the US variety
> wasn't their first language.
>
> You should know by now I'm awkward and do things differently.
>
> Cheerio John
>
> On 21 October 2017 at 16:08, Blake Girardot <bgirardot at gmail.com> wrote:
>
>> I think you just want to sort by "last feedback" then.
>>
>> That should tell you how long it has been since the mapper last got
>> feedback and if it is the same as their registration time, you know
>> they have probably never gotten feedback.
>>
>> Can you try that and see if it is a fair replacement for your most
>> recently mapped task. I think it will get you the same thing
>> essentially.
>>
>> regards
>> blake
>>
>> On Sat, Oct 21, 2017 at 3:48 PM, john whelan <jwhelan0112 at gmail.com>
>> wrote:
>> > Remember I'm basically lazy.  I do two things when I validate.  The
>> first is
>> > clean up the second is provide feedback.
>> >
>> > The sooner someone receives feedback the better the response.  You get
>> more
>> > tiles out of them and they are of a higher quality.  This means less
>> clean
>> > up for me to do.
>> >
>> > Someone confirmed the feedback loop and confirmed that within 24 hours
>> is
>> > best then you get a trickle down effect so after a week you get much
>> lower
>> > responses.
>> >
>> > The other part is if I'm validating tiles that are a week old am I
>> going to
>> > annoy the mapper because they may have already corrected their bad
>> habits as
>> > a result of earlier feedback and yet these are still being mentioned.
>> So
>> > the mapper doesn't get a warm fuzzy and to be honest it takes longer
>> for me
>> > to provide proper feedback than to have mapped the tile myself.
>> Remember
>> > I'm lazy.
>> >
>> > So when I land on a project the ideal is validate the most recently
>> mapped
>> > tiles first then quietly clean up the rest.
>> >
>> > Does that make it clearer?  It's about problem avoidance or prevention
>> > rather than verifying the mapping as such.
>> >
>> > The how long have they been mapping helps, some mappers I'll take a
>> quick
>> > glance at, newer mappers it's zoom in and look for a few more things.
>> >
>> > Thanks John
>> >
>> > On 21 Oct 2017 3:32 pm, "Blake Girardot HOT/OSM" <
>> blake.girardot at hotosm.org>
>> > wrote:
>> >>
>> >> Hi John,
>> >>
>> >> If you could explain just a bit more about why the most recently
>> >> mapped task is your main priority (sometimes)?
>> >>
>> >> I do not understand.
>> >>
>> >> I understand it has something to do with someone who mapped 3 months
>> >> ago, we now also send emails so time passed since they last mapped is
>> >> much much less of an issue (Martin is going to maybe look at redoing
>> >> his analysis).
>> >>
>> >> We actually started off with what you are asking for, but feedback
>> >> from users was they were much more interested in getting to the newest
>> >> mappers the fastest, which is what we ended up going with.
>> >>
>> >> Maybe it is easy to add back in, but that table is already pretty
>> crowded.
>> >>
>> >> Regards,
>> >> Blake
>> >>
>> >> On Sat, Oct 21, 2017 at 2:37 PM, john whelan <jwhelan0112 at gmail.com>
>> >> wrote:
>> >> > I've found what you are doing on the video, your accent is fine you
>> know
>> >> > of
>> >> > course that people studying English like what it is spoke years ago
>> >> > travel
>> >> > to parts of the US to record the Americans as their language has been
>> >> > isolated for many years so it much closer to English what was spoke
>> when
>> >> > their families emigrated.
>> >> >
>> >> > Yes I can pick out a user who registered ten minutes ago and the
>> tiles
>> >> > they
>> >> > mapped but that isn't what I'm after.  I'm after the most recently
>> >> > mapped
>> >> > tile and that's the one that is the most valuable to give feedback
>> on.
>> >> >
>> >> > By the way I'm using cut and paste on the name displayed below to
>> give
>> >> > feedback with the @ sign or <Crtl>i to grab the username in JOSM, its
>> >> > faster
>> >> > than typing it in as you were doing.  Some mappers are naughty and do
>> >> > not
>> >> > completely map a tile, but their mapping still requires feedback
>> even if
>> >> > someone else finishes it off.
>> >> >
>> >> > Thanks
>> >> >
>> >> > Cheerio John
>> >> >
>> >> > On 21 October 2017 at 14:20, Blake Girardot <bgirardot at gmail.com>
>> wrote:
>> >> >>
>> >> >> Hi John,
>> >> >>
>> >> >> Thank you very much for your feedback on the TM3, this and the
>> >> >> previous. And thank you for putting them in GitHub, that is really
>> the
>> >> >> best place for them.
>> >> >>
>> >> >> The TM3 has better than the most recently mapped task.
>> >> >>
>> >> >> The most recently mapped task tells you nothing about the skill
>> level
>> >> >> or the newness of the mapper. It could have been completed by me, 7+
>> >> >> years in OSM.
>> >> >>
>> >> >> What I really want is to find when I am validating is the newest
>> >> >> mapper who someone has not already given feedback.
>> >> >>
>> >> >> No point in me giving a 2 hour old user feedback if you got one of
>> >> >> their tasks 5 mins ago.
>> >> >>
>> >> >> That is what you can see in the Tasking Manager 3.
>> >> >>
>> >> >> I know my accent bugs you, so I made a silent video of how I use the
>> >> >> TM3 to get the newest mappers feedback as fast as possible:
>> >> >>
>> >> >> https://youtu.be/BX8rzi2NqQY
>> >> >>
>> >> >> Let me know what you think of the workflow in the video if you would
>> >> >> please.
>> >> >>
>> >> >> Respectfully,
>> >> >> blake
>> >> >>
>> >> >>
>> >> >>
>> >> >> On Sat, Oct 21, 2017 at 11:06 AM, john whelan <
>> jwhelan0112 at gmail.com>
>> >> >> wrote:
>> >> >> > Typically when I validate I validate the most recently mapped
>> tiles
>> >> >> > first
>> >> >> > working on the principle that the faster feedback the more likely
>> it
>> >> >> > is
>> >> >> > to
>> >> >> > be effective.
>> >> >> >
>> >> >> > If someone has already changed their habits telling them that two
>> >> >> > villages
>> >> >> > two kilometers apart are probably not connected by a primary
>> highway
>> >> >> > but
>> >> >> > by
>> >> >> > a minor road three months after they have mapped is a waste of
>> time.
>> >> >> >
>> >> >> > I can't see how to do this in TM3.  I need a link to the tile.
>> >> >> >
>> >> >> > _______________________________________________
>> >> >> > HOT mailing list
>> >> >> > HOT at openstreetmap.org
>> >> >> > https://lists.openstreetmap.org/listinfo/hot
>> >> >> >
>> >> >>
>> >> >>
>> >> >>
>> >> >> --
>> >> >> ----------------------------------------------------
>> >> >> Blake Girardot
>> >> >> OSM Wiki - https://wiki.openstreetmap.org/wiki/User:Bgirardot
>> >> >> HOTOSM Member - https://hotosm.org/users/blake_girardot
>> >> >> skype: jblakegirardot
>> >> >> Live OSM Mapper-Support channel - https://hotosm-slack.herokuapp
>> .com/
>> >> >
>> >> >
>> >> >
>> >> > _______________________________________________
>> >> > HOT mailing list
>> >> > HOT at openstreetmap.org
>> >> > https://lists.openstreetmap.org/listinfo/hot
>> >> >
>> >>
>> >>
>> >>
>> >> --
>> >> ----------------------------------------------------
>> >> Blake Girardot
>> >> Humanitarian OpenStreetMap Team
>> >
>> >
>> > _______________________________________________
>> > HOT mailing list
>> > HOT at openstreetmap.org
>> > https://lists.openstreetmap.org/listinfo/hot
>> >
>>
>>
>>
>> --
>> ----------------------------------------------------
>> Blake Girardot
>> OSM Wiki - https://wiki.openstreetmap.org/wiki/User:Bgirardot
>> HOTOSM Member - https://hotosm.org/users/blake_girardot
>> skype: jblakegirardot
>> Live OSM Mapper-Support channel - https://hotosm-slack.herokuapp.com/
>>
>
>
>
> _______________________________________________
> HOT mailing listHOT at openstreetmap.orghttps://lists.openstreetmap.org/listinfo/hot
>
>
>
> _______________________________________________
> HOT mailing list
> HOT at openstreetmap.org
> https://lists.openstreetmap.org/listinfo/hot
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/hot/attachments/20171024/2d2b3a7f/attachment.html>


More information about the HOT mailing list