Hungerburg's Diary Comments

Diary Comments added by Hungerburg

Post When Comment
Open Government Data mit jq und JOSM 1 day ago

Noch eine Zahl zu den Forstwegen, für die bicycle:conditional=Yes @ (Apr - Oct) gilt - durchs Hintertürl gesehen sind es einmal sicher so um die Drei, Vier- von den 18 Tausend km, wenn man alle in Openstreetmap erfassten MTB-Routen darauf hin anschaut, wo sie über tracks führen - siehe gibt aktuell 3030 aus - und das mapping ist nicht ganz vollständig

When is a path not a path? 7 days ago

It is not that much different here in Tyrol (Austria). Administrative data by the BEV (the local OS) contains both outdated and plain rubbish pathways, all the wile being quite precise on what is there most often. At times, some of the paths they map have special appeal in rare occurrences, because local law does not allow to create new paths, but allows grooming paths existing on paper.

Open Government Data mit jq und JOSM 10 days ago

Manchmal dauert es, bis dass der Groschen fällt: Nach dem Beitrag hier aber darf nicht verwundern, warum in Tirol – im Unterschied zu Ober- oder Niederösterreich laut Forum topic Anekdoten – so viel Fahrverbote ausgeschildert sind - Weil 80% der tracks Forstwege sind, und von diesen das Forstgesetz es so verlangt! Auf dass über allen Wipfeln sei Ruh? Was so beobachtet werden kann verlängert sich die Praxis nahtlos auf Feldwege …

How to show paths on a map? about 1 month ago

I am still fond of this comment on the tagging ML, - The interesting content is below the fold ;)

Consuming conditional access tag values 3 months ago

Hallo Arminus, das ist wohl wahr! Die Zeichenkette heißt ja auch “day”, weil sie einen Tag will :) In der regex muss dann der Tag optional werden, und in Folge ein Tag eingesetzt werden, wo der fehlt. Ich werde das gleich oben ändern.

PS: Du mappst ja selber Steige und bereitest Daten für Karten auf? Fällt dir dazu etwas ein - - ich will das bald zur Abstimmung bringen und bin über jeden Rat froh.

Consuming conditional access tag values 12 months ago

Late Update: The data -, The consumer (xctrails) already used Simon’s Lexxer/Parser on the server, before I wrote my exercise, but the work still was not in vain ;)

2021 editor usage stats - some interesting parts about 1 year ago

StreetComplete is a very valuable tool to enter immediate observations on the ground. Still, please do no oversimplify use cases:

I guess that sidewalk=no, cycleway:both=no, noname=yes, lit=no also belong here.

Tagging a residential road sidewalk=no gives valuable information, while on a motorway it is redundant. cycleway:both=no a bit less so, because a cycleway is not so common as a sidewalk. Such issues are best handled on the SC tracker. Some heuristics may need some polish.

SC needs to grow user base a lot, if it was to replace work done on the desktop. Please think twice before gamification - something that applies to desktop editors just the same - ticking off validator suggestions blindly should not give users Bonus points.

Consuming conditional access tag values about 1 year ago

Cool, A real parser in the .jj file. I guess from coercing openstreetmap restrictions into BNF or some such formalism, a number of proposals can spring from. Sadly my mind has deteriorated too much to be on any help, since writing a Bison parser for something much more simple years ago; and even then… Still, there are lots of smart people! Writing down the grammar might be a “good first issue” :)

Misuse of sac_scale in the Alps about 1 year ago

Hello SK: Early autumn, I was in the Karwendel, paying visit to friends that herd some cows there, with a hotel-hut close to the Alm; I had dinner in the touristic hut. I like to ask people, to let me have a glimpse on their Navi. I found does a good job at rendering “trail_visibility” - A value of “no” makes a path nearly invisible on the map. Of course, thats the one I chose earlier that day. In fact, there was no path, a pleasant hike over grassy slopes up to a summit, that once had a bicycle as a cross mark. I was the only one there, caravans went to the prominent summits that day. After dinner, the host advised people, to go to the second prominent summit then, because of snow fields on the most prominent one. The photos he showed, did not show anything difficult. I tried to pith the other summit as truly easy, he just said, 10 per year go there, nothing further. The next morning I went examining another “mysterious” OSM “path”; Again, a very interesting hike, mostly along a pressurized water pipeline; I had to delete a part of this path, because it created a routable connection, even though no signs on the ground showed actual use as such shortcut there. On the retour, I went the touristic path, which features some assisted passages; The next day, a person fell to death on this one. I remember, reminding myself to be careful in the very location, but I never had thought, that it was about life or death. In English, its called accident, in German its called Unfall, both means, German maybe a bit more concise, what one is not expecting to see happening. Otherwise, it probably would not happen.

Misuse of sac_scale in the Alps about 1 year ago

Hello SK, I am not the least that familiar with Swiss customs as you are, being from a neighboring country, where trails are graded like skiing pistes: blue, red and black. I like it, that openstreetmap has all the paths, that disappear from commercial maps, with each new edition the more. Mostly, they are not fashionable either, so one can hike them in splendid isolation, so to say. I sometimes delete ones, that are obviously wishful thinking or plain misinterpretations of the aerial. I also like to remove sac_scale tags from ‘informal’ paths, not the least, because those tags make them more prominent on OSM derived hiking maps ;) Actually, I am less concerned about dangerous trails being mapped, than about users, who delete paths, because they are dangerous in their mind or give ridiculous sac_scale values, just to make graphhopper not route there. When a hiking accident makes it into the news, my first is, to look what OSM has there. Over the years, I learned of a prominent case, where the “Bergretttung” had to engage for at least two times, where nothing was in OSM, while google maps was the “Navi”. I learned of fatalities, among many, on paths, that are in the official state map too and once on a path, that is a bit en vogue, got covered by many blog articles and is in all the basic hiking guides. The last one surely classifies as T6, in the meaning, WS climbing, no possibilities for securing. The entry is two to three hours uphill walk from the bus stop or car parking. So I do not consider this a danger to the public ;)

Misuse of sac_scale in the Alps over 1 year ago

Let me first point at a misunderstanding in the first paragraph: That there are different scales, a hiking scale and a climbing scale, does not mean, that one starts, where the other ends. It is mentioned in the SAC document, that T6-hiking may comprise WS-climbing, but without the possibilty of securing by rope. The grading document further says, that it is meant to describe the difficulty in optimal conditions, i.e. fine wheather, on dry ground, suitable snow-cover, etc.

This is not to say, that there are trails in the OSM data, that are beyond SAC difficult alpine hiking, and consequently, must not be mapped as a path. So ZS climbing is out of scope, as well as UIAAIII. I find it interesting, that by omitting the “hw=path” tag, the trails can be kept, but hidden from display in OSM-Carto. Is there a tag, that may show them on special “climbing” maps?

SwissTopo has an overlay at that shows all the signed hiking trails. You will get to see the very nice hill-shading in close zoom :) You will notice, that there are not that many blue ones. I looked up “Niesengrat” from the search box, because this is mentioned as a sample for T6 difficulty. There is no trail shown in the Swisstopo overlay to the Fromberghore. That may explain, why there are so few blue trails to be seen. I guess, that is because the demanding and difficult alpine trails are not signed.

Luftbilder im "source" Tag? over 1 year ago

Hmm, das schweift vom Thema das Blogposts ab. Nach dem was mir oben so durch den Kopf ging finde ich nun den source Tag nicht den passenden Ort, die Chronik zu duplizieren.

Im Blogpost geht es vielmehr darum, dass manchmal in Foren behauptet wird, das Luftbild müsse als source am Objekt angeführt werden (das würde dann auch einzelne Bäume, Sitzbänke usw. betreffen), um dessen CC-BY Lizenz zu erfüllen, und warum dem meiner Ansicht nach nicht so ist.

Dass das source Tag als veraltet angesehen wird, scheint genau in dieselbe Richtung zu deuten :)

Luftbilder im "source" Tag? over 1 year ago

Die verbreitetsten Editoren machen das eh schon von selbst so, dass die eingeblendeten Luftbilder im Changeset angeführt werden. JOSM verwendet “source”, iD und weitere “imagery_used”. Die Frage bringt mich aber auf eine Idee.

Ein Mapper in der Gegend hier zeichnet sehr oft Straßen schön, so dass Kurven auch in zoom 19 noch Kurven sind. Er setzt an den Weg dann ein source mit Luftbildquelle + und + Jahr des Luftbilds. Der Vorteil davon ist, dass man nicht die Chronik aufschlagen muss um herauszufinden, wo ein Versatz herkommt, oder der Verlauf nicht mit dem übereinstimmt, was man im eigenen Luftbild sieht.

Von daher könnte die Praxis Sinn machen. Es wird aber bald kompliziert: Wenn ich an einem Detail was ändere, was dann: Die Quelle ersetzen oder eine zweite anhängen? Was ich so beobachte, passiert das nicht. Eher bleibt die “source” länger kleben als etwas von ihr noch da ist.

Auf der Suche nach dem verlorenen Wald almost 2 years ago

This updated and a followup too, presenting, below the fold, results of laymen research into local forestry regulations.

Auf der Suche nach dem verlorenen Wald almost 2 years ago

Dear Penegal, You are welcome! Feel free to task me. I wrote this in German, because of the matter seemingly being most controversive in the German community. There are some things to amend, I hope the piece will not get longish and boring. Kind Regards

Sidewalks by length about 2 years ago

The story continues here Sidewalks by popularity - I rather not modify this post, as it has discussion and is linked from the Wiki. You can do your own plots from the OHSOME dashboard, but it does not do ratios.

Sidewalks by length about 2 years ago

PS: I know that you are aware of confirmation bias, because it was you, who introduced the “regional variations” section in the sidewalks article in the Wiki.

PPS: Someone needs to tighten the separate way section in the Wiki, it even contradicts itself. It does not do a good job at advertising the usefulness of this approach.

Sidewalks by length about 2 years ago

Hello Mateusz,

Its been fun to explore what a database like OSM holds :) I was quite surprised to see the results to come out so contrarian. It is too much work to change the legend, think of it as a shorthand, restating the names of the OSM keys, that are used to map the feature.

Regarding path, that is only documented for some weeks, but as you said, surely existed for much longer. I recreated the stats locally and there is no difference at all in Britain and negligible difference in Poland.

Please note, this is the actual length of the features mapped, according to OHSOME down to less than 1% deviation, not some arbitrary number, such as count of OSM entities.

Note, that this also does not count “sidewalk=no”, which some consider humbug, but others consider worthwhile, e.g. developers of pedestrian routers. Some negative sidewalk tags can replace hundreds of positive ones, as, at least in my opinion, most of the sidewalks on the ground in OSM are mapped as “highway=residential” (implicitly that is; much like the developers of the the abstreet game e.g. who would be much better served with highway tagging than with separate ways.)

Gehsteig oder Begleitweg? about 2 years ago

Diesmal ohne Schummeln: Das Foto ist in Österreich. Hier wird zw. Gehsteig und Gehweg unterschieden. Ersterer ist kenntlich durch bauliche Gegebenheiten, zweiterer durch Beschilderung. Beide sind per se ausschließlich für den Fußverkehr da. In OSM ist beides unter /footway/ zusammengefasst, kommt aufs selbe heraus, also kein Malheur.

Auf dem Bild erkenne ich im Vordergrund keinen Gehweg. Wie auch, ist ja kein Schild da! Das könnte genausogut ein Bringungsweg sein. Der könnte auch in den querenden Bringungsweg münden. Dass das hintere ein Gehsteig ist, der - merke - kein Schild braucht. Wenn man bei dir Straßenbau studiert hat, dann kennt man’s :)

Gehsteig oder Begleitweg? about 2 years ago

Hallo Robhubi, das Bild kenn ich! Es zeigt einen Gehsteig, von dir selbst so bestimmt. Aus reiner Faulheit hab ich selber schon an Straßen derartiges als “sidewalk=*” drangehängt, wozu auch parallele Linien zeichnen zu etwas, das schon da ist? Allerdings kenn ichs von fleißigeren Mappern gern als eigenständigen “path” oder “footway” kartiert, mit oder ohne “sidewalk” tag. Ab und zu auch mit “bicycle=yes”, ohne dass das irgendwo ausgeschildert wäre. Wo kein Kläger, da kein Richter?