[OSM-talk-be] TEC: what I did and why I stopped (long ago)

Tue Apr 7 23:14:20 UTC 2015

Checked the stops of line 65. Fixed some of the remarks. Couldn't fix
others. André's findings conflict with the data coming from TEC. He has the
local knowledge, so the most likely explanation is that he's right.

I never claimed the stops were at the correct positions. They're all in
now, so feel free to move them to the proper place if you pass them by.
That's what I do for the ones that are nearer to me, when I happen to go
there.

For the non existing stops, which are still part of a bus route in TEC's
data, it's annoying. No idea what to do about those.

I did a best effort attempt to reconcile what TEC provide us with, with
what I could figure out from the aerial imagery. If, at some point, people
will provide us with Mapillary pictures from the surroundings of the stops,
it will become possible to verify their positions in a better way. But last
summer there was no such thing and since we're not allowed to look at
Google street view, I refuse to do so.

I also saw nobody stepping in to use the data in the file that I made
available on Dropbox, so I took it upon myself to add them. If I had know
that this was going to cause bad blood from Brugge all the way to Liège, I
might have chosen another hobby to waste my time on. Everybody happy.

The way I figure it, we're constantly trying to improve the data,
incrementally.

Oh yes, source tags belong on the changesets, so I remove them.

tags like todo_by_Papou don't belong in the data either, so they annoy me
and cause me to write inflammatory comments on changesets. Odd that it
takes months before a reaction surfaces. And of course not as a personal
message, but as a public shaming. Way to go.

There is one comment I do like. When I draw shelters, I've never added
shelter=yes to the highway=bus_stop node. Often thought about doing that,
but there are also contributors who'd consider that redundant tagging. I
don't mind a bit of redundancy, myself. So if others think it's a good
idea, I might start doing that.
Oh, I also never removed bin,bench or shelter when I drew them as separate
objects. That's indeed not totally logical. Oh well, nobody's perfect.
Least of all me.

Jo

2015-04-07 21:46 GMT+02:00 André Pirard <A.Pirard.Papou at gmail.com>:

>  Hello Julien and TEC collaborators,
>
> I finally decided to explain here why I stopped working for the TEC
> project.
> Mediocre result
>  I had almost totally mapped TEC line 65 (before the stops were announced
> to be complete). I had made all the hard work finding where the TEC bus
> stops are (1). I still needed to review the whole thing with aerial maps,
> but while waiting for an illusive confirmation from TEC of about 10 bug
> reports I sent them,  I was giving myself a break and I was playing with
> the GPX trace that I'm probably the only one to have made.  I was having
> fun writing a program that analyzes GPX, automatically detects where the
> car stopped, computes the average of the oscillating stop position and
> makes a POI of it, all that displaying nicely in JOSM.  The pre-alpha was
> producing amazingly good results for a first try. (That program might
> analyze GPX data produced by a 60€ GPS smartphone installed on the buses).
>
> (1)  TEC's data can put them 100 m and more away from the true location,
> even on another road. I had put most of them less that 5 m where they are,
> and as carefully tagged as possible with shelter etc...
>
> The fun was totally spoiled when Polyglot sent me an e-mail saying that
> (without first contacting me in any way) he had made modifications to many
> of my pending bus stops and that the schoolmaster or is it OSM chief was
> not pleased.
> He had posted the following public insults in what the OSM.org map readers
> see when they look at my bus stops data (left pane).
> We should try to explain to Polyglot that OSM is a geographic database and
> not a database of insults.
>
> *reviewing all bus stops added for route 65. Performing conflation where
> required and improving positioning where needed. Why were 3 versions
> necessary to get this mediocre result? *
>
> *routes for TEC 65, mapped properly, please, if somebody offers you to
> show how it's done, grab the opportunity to learn. Alternatively look at
> all the other routes this person already did over the past years *
>
> And maybe more.
>
> What a sin to make 3 OSM updates for about 100 bus stops, isn't it! Read
> below how many he made !!!
> Regarding what the ignorant people we are must "learn" from the teacher, I
> had asked before on the mailing list several questions and they were never
> answered.
> The tagging conventions are here
> <http://wiki.openstreetmap.org/wiki/WikiProject_Belgium/Conventions/Bus_and_tram_lines#Tagging>,
> the answers should be there, and I did exactly what they say.  Obviously,
> if we are commanded to look at other routes, their links should at least be
> in those conventions.  Etc.
>
> Being disgusted (and because it's alleged be a way to "learn"), I looked
> at just a dozen of those so-called corrections, no more, I stopped out of
> disgust.
> Let us see what "mediocre result" means.
> Mostly, they move by 2 m the bus stops that I had painfully found 50 or
> 100 m away and correctly moved.
> Or they do such minor details.
> Or they just introduce plain mistakes.
>
> *Please* note that I do not disparage other people's work, quite the
> opposite.
> But when my work is disparaged, I compare.
>
> *Please* let me make my final line 65 modifications of the following
> before doing any more anyone.
>
> In this place <http://www.openstreetmap.org/node/2955197984>, rue Hamal,
> Polyglot made a slight position change of "Liège Opéra".
> I was waiting for and I now have confirmation that this bus stop does not
> exist.
> I think that it was a temporary stop during the road works in 2009 and
> that it was never removed from the map.
> I am going to remove it, at least for line 65.
> Is moving non-existing bus stops really a correction?
> Which is a mediocre result?
>
> This nearby <http://www.openstreetmap.org/node/2955197983> place is the
> line terminus of 65 and, as I know it, quite logically the place where the
> bus waits for passengers and starts.
> I had put the line start where the buses start but Polyglot put it at the non-existing
> stop <http://www.openstreetmap.org/node/2955197984> here above.
> Which is a mediocre result?
>
> In this place <http://www.openstreetmap.org/node/2955198000>, I put the
> bus stop quite normally on the right side of the road.
> Polyglot moved it to the left side, on the sidewalk behind the parking!
> Now, TEC must buy English like buses with doors on both left and right
> sides to service that stop !!!
> Checking it today, I see that Polyglot has put it back where I had put it
> (In 3 additional updates !!!)
> Which is a mediocre result?
>
> I was waiting confirmation that this bus stop
> <http://www.openstreetmap.org/node/2955198001> does not exist.
> Although that's indicated in a fixme, Polyglot moved once again a
> non-existing bus stop. (I passed by that place and I didn't see any bus
> stop with a careful look).
> Are the bus stops moved just in order to write insults?
> I will remove it.
> Which is a mediocre result?
>
> When I checked this stop <http://www.openstreetmap.org/node/2955197938>
> with my GPS, I noticed a bus stop opposite.
> I first thought that it belonged to line 65.
> But a close look at the photo I took showed that it belongs to line 727.
> So, it should be on the map, now that all bus stops are finished, but it
> is not.
> I will do nothing because it's not my line 65 matter.
> Which is a mediocre result?
>
> And so, I wondered if that stop opposite
> <http://www.openstreetmap.org/node/2955197938> did not belong to line 227
> too.
> And it does, I did see a 227 bus stopping there, but that's not indicated
> in the tags.
> Which is a mediocre result?
>
> Near this bus stop <http://www.openstreetmap.org/node/1645537259>,
> Polyglot drew a shelter. Very nice.
> But he did not shelter=yes in the stop.
> I would probably have done that in my last review because I'm not doing
> mediocre tagging.
>
> Polyglot had written:
>
> If you insist on putting them [source=TEC 2014-04] on the objects, take
> the prepared osm file. Select all objects, and add the source tag you like.
>
> I did put source=TEC 2014-04 in this bus stop
> <http://www.openstreetmap.org/node/857875464/history>, as an experiment,
> to see what would happen.
> What happened is that Polyglot removed it.
> Regarding 3 updates, I had corrected the TEC position by more than 100m
> and put the correct tags in 1 update; Polyglot used another one for that
> freakish update. A third one will be needed to add the shelter he forgot.
> What's the name of that game?  "Mediocre Results"?
>
> Regarding the 3 updates insult, this bus stop
> <http://www.openstreetmap.org/node/857875464/history> had correctly been
> tagged and moved by 20m+ in *1 update* by Papou.  Polyglot used *2 more
> updates*, and now 3, to move it by 2m and to remove source= again. Really
> important and worth an insult !!!
>
> I'm just back from totally different work where I happened to see 2 bus
> stops 10m and 30m off their position.
> And, how can one possibly insult other mapper when that line is going
> contraflow (contra one-way) there !!!
>
> I made 4 spelling corrections to the place names (see below). Polyglot
> managed to reintroduce one of them in the line names.
>
> Come with us and play the Mediocre Results game!
>  History Let us briefly recall what happened before that.
> I shared Julien's enthusiasm by making a beta TEC_2014_04.osm import file
> containing bus stops and lines.
> I tried to start a so-called necessary discussion about tags, lines,
> capitalization etc on talk-be and there was absolutely no reply.
> Julien published
> https://wiki.openstreetmap.org/wiki/WikiProject_Belgium/Mapping_resources/TEC
> containing his suggestions and a reference to my file.
> I discovered that Polyglot had silently replaced my name and file by his
> and totally ignored what Julien had written in this document. He claimed
> that he didn't find my file but he never asked (I had renamed it to avoid
> misuse). I had made instructions to use my file; he erased them and refused
> to make instructions for his.
> I said that Polyglot's file should contain the TEC lines with their
> numbers. My file did contain the lines.  They are necessary (for example to
> avoid contraflow like I have found above). Polyglot refused because "De
> Line's files do not contain lines" (and that maybe causes wrong routes).
> That may be too why line 227 stops are missing.
> Let us notice that OSM presently contains bus stops but no lines (except
> "mediocre" ones).  I wanted to add more nearby ones, correctly, but I will
> not.
> I offered my help to convert the place names to lowercase and fix the
> accentuation and Polyglot refused it. My idea was to match the names with
> those of nearby places of the OSM data.  Polyglot used Wikipedia
> municipality names which, of course, miss the villages.  As a result, for
> just one line I had to fix 4 names which the OSM map was displaying
> differently for the village and for its stops inside.  Polyglot
> reintroduced one mistake and there are more of them.
> I said that the TEC data should obviously contain source=TEC 2014_04 so
> that it can be queried (with overpass) which release they contain and
> Polyglot obstinately refused such a mere request. TEC data is now at TEC
> 2014_07, even 2014_10 and probably later they say, which is unused, and
> just wonder how you would differentiate a mix of 04 and 07 if it were.
> Question: if someone were not fearing insults and introduced 2015 data,
> how would he do that and how would the others know (with overpass)?
> That essential question wasn't even considered.
> In consequence, I thought that I would be more happy if I stopped trying
> to organize TEC and limited myself to mapping a few lines.  But I wasn't
> happy even so.  And I have the feeling that the same thing happened to
> Julien.
>
> I may write a second e-mail about the lines.
>
> Best regards.
>
>   André.
>
>
>
> _______________________________________________
> Talk-be mailing list
> Talk-be at openstreetmap.org
> https://lists.openstreetmap.org/listinfo/talk-be
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/talk-be/attachments/20150408/ea2fd1b9/attachment.htm>