Vincent de Phily has commented on the following diary entries

Post When Comment
Rising participation in the discussion of tag definitions about 6 hours ago

YMMV, I really don't like forum interfaces. And yet-another place to find info I want ? Nah. I'm subscribed to tagging, but that doesn't mean that I read all threads.

The forum vs mailinglist (vs $FOOMEDIUM) debate will never die out, because each has its own advantages. As Richard pointed out, you can use a forum-like interface for mailinglists. The reverse is not always the case for forums (I've never looked at the osm forum so I don't know in that case).

Website for asking to improve just one area in OpenStreetMap (or paying for improvement) 20 days ago

Mapbox and Telenav are two companies (out of the top of my head, maybe lists others) that employ full-time mappers. You could try contacting them for a once-off single-area job.

Regarding the "maps for self-driving cars" idea, AFAIK those cars currently need very detailed maps. Armchair mapping certainly isn't sufficient. OSM still has good but not stellar data for for the satnavs usecase (missing width, lanes, restrcitions), so I think than osm-based self-driving cars are a bit far off.

How to improve OSM: kill the bureaucracy 21 days ago

Contrary to imports, there's no such thing as a tagging approval process. There are conventions to help reaching consensus, which is quite different. OSM has always been pretty much the "let's allow things to happen, then fix them when they break" phylosophy.

The "descriptive vs normative" can be a bit hard to navigate, but really the wiki should do both, and it usually does. Many pages describe both "what's in the db" and "what is the prefered method".

The wiki certainly does need TLC. The "process" to tidy the wiki is to just do it. That doesn't necessarily mean conclude the discussion, but the pros and cons of each view can be more clearly documented.

The level of bureaucraty is subjective. You're turning people away for reasons I wouldn't have batted an eylid about.

OSM has loads of communication channels. Many people actually prefer mailing lists, but there's IRC, a forum, blogs, social media, private messages and changeset comments... You name it. Besisdes, you want to fix wiki-related problems by adding another wiki ??

Vision? about 2 months ago

That's very close to what the french community did with their BANO dataset : essentially a big CSV of addresses with many sources, tools to compare back and forth between BANO and osm, and the ultimate long-term goal of getting most of it merged into osm.

How can I get accurate locations in a forest ? 2 months ago

Just do your surveying in two steps : the first in summer while carrying a flame-thrower, the second a few days later while carrying your GPS. Problem solved :p

what are these oddly shaped circular features? 2 months ago

I've usually mapped those as orchards. They're a bit too low for palm trees. My guess is pineapple, but don't quote me on that.

As for there being no paths to reach the area, remember that main access is by foot and is often not visible from imagery.

No Waiting 2 months ago

Those notes seem to be of little value indeed. You should try asking on the mailing list, to identify the guilty program and fix the root cause.

First thoughts 3 months ago

Welcome :)

It can be disheartening to redo a lot of work that you already did for Google, but on the other hand, your edits are much more likely to eventually reach stanavs (even if that likelyhood is still small) and varied usecases.

The inside-out vs outside-in question is a classic one with not definite answer, but since you already have GIS experience and are interested in the whole county, I suggest starting with the big picture before moving to smaller roads. 3 months ago

So where is the "checked" information stored, if not in the object's tags ? A third-party db ? If so, why do you need OAuth ?

Other than that: nice tool, thanks :)

No joy this weekend 4 months ago

Ireland is fairly well-mapped, with what I guess is a comparatively high number of veteran mappers per km2. Yet I'm often (depending on my mood) spending most of my osm time coaching new users. Just one anecdote in the "this problem is not specific to Brazil" camp.

It's tedious work, and I can see how disheartening it can be if you're the only one doing this in your area, but it's more useful than just fixing mistakes. It's also for advocacy and for recruiting. Some of those map-breakers will eventually become fully competent and share your load; coaching them increases the likelyhood of that.

We can no longer go on like this 4 months ago

Validation is hard. JOSM has the best in-editor validator, but it is far from being flawless. Tools like osmose have very good algorythms too, but they are after the fact and not worldwide.

It'd be great to have something like a changeset review board that newbies (or anybody having doubts) could upload to instead of uploading to the APIs directly. That board would run a standard set of checks, allow other people to look, have some shortcuts to IRC, and have a "send to DB" button. A man can dream, can't he ?

New Bing imagery 4 months ago

You'd think that they would have spoted the hole during QA and kept the old imagery for it :/

They probably still have the old imagery somewhere, it's worth a try alerting them about the regression (sorry, no idea who to contact for this).

Trace GPS loin de la réalité ! 4 months ago

Les GPS ont du mal dans les zones encaissées (vallées, hauts bâtiments...) ansi qu'en forêt ou par très mauvais temps. C'est normal qu'un tracé dans les Gorges du Tarn soit de mauvaise qualité.

Les combinés GPS+Glonass ainsi que les appareils spécifiques randonée font parfois mieux. Pense à attendre le "fix" GPS en terrain dégagé avant de t'engager en terrain difficile. Utilise plusieur appareils en même temps et fait plusieur passages à au moins 1h d'intervale, pour faire une moyenne entre tous les tracés.

Native English speakers: Please take action when a Tag name is wrong! 5 months ago

Sometimes, after much arguing, the AE term ends up as the consensus winner. See the shop=fishmonger -> shop=fish -> shop=seafood debate.

4 826 424 "addr:country=DE" 6 months ago


I'm not sure I follow the point in your 1st paragraph. My point is that when you want to route to a particular housenumber, the routing algorythm will find the closest matching way and direct you there. Most of the time, doing that matching using just the street name, even if there are multiple corresponding ways, is correct. But sometime (typically in housing estates) the name-matched osm way that is closest to the house is the wrong one. If matching using a relation instead, there's only one osm way to choose from and no mistake made.

Regarding the ease of contribution (for newbies and veterans), there is no obvious winner. Creating the relation is harder for newbies, same amount of work for veterans. Maintaining it is easyer for both.

It's also worth it to add a house to the associatedstreet relation even before you know its housenumber. Firstly because it makes surveying housenumbers easyer, secondly because it is usefull information even in the absence of housenumbers.

4 826 424 "addr:country=DE" 6 months ago

Let's run some actual data compression numbers :

  • download extract from geofabrik and decompress
  • use grep to filter out the addr:country tag into a new file
  • recompress both files

$ \ls -Sl brandenburg-latest.* -rw-r--r-- 1 work work 2019804582 Jun 24 00:34 brandenburg-latest.osm -rw-r--r-- 1 work work 2007110472 Jun 24 00:38 brandenburg-latest.nocountry.osm -rw-r--r-- 1 work work 159076860 Jun 24 00:34 brandenburg-latest.osm.bz2 -rw-r--r-- 1 work work 158989054 Jun 24 00:38 brandenburg-latest.nocountry.osm.bz2 -rw-r--r-- 1 work work 119848272 Jun 24 00:34 brandenburg-latest.osm.xz -rw-r--r-- 1 work work 119720968 Jun 24 00:38 brandenburg-latest.nocountry.osm.xz

The gain is 0.6% for plaintext, 0.05% for bzip2, 0.1% for xz... All in all not very impressive, but not as insignificant as the 40K that flohoff expected. Compressing a string disseminated in a file is not as easy a compressing a file made entirely of that string.

Also, remember that downloading and decompressing data might not be your bottleneck. Parsing the decompressed data has its cost.

Since the numbers are small (I could increase them by using an adress-focused extract, but why bother ?), I agree that size is not very important in this case. But it's still good to have proper numbers :)

4 826 424 "addr:country=DE" 6 months ago

I'm in the "no redundancy" camp, mainly because it is less time-consuming to create and maintain (from fixing a typo on a street name to the anexation of Crimea). The fact that the data is smaller and tidyer is a bonus. For the specific case of the associatedstreet relation, it also fixes the problem of finding the right street to park in front of a house when the nearest-matching-way algorythm fails.

Normalizing the data does make querying more complicated, but the fact is that relations offer advantages and/or make things possible that wouldn't be otherwise. They are used more and more. So if your tool doesn't handle relations... It doesn't handle OSM data very well at all. Go fix it.

If you "don't want a geodatabase" because querying street and country at runtime is too costly, add those tags to each object at import time.

There's a balance to be found between coders/computer ressources and data contributors. On that point, while it is probable that geocoding will eventually be computationally "solved and finished", adding and curating data will never be. So I lean on the data contributor's side in this equation.

Of course there are compromises to be made. "Don't tag to the renderer/geocoder" only goes so far, pragmatism is also needed. But in the addr:* case, I feel that this pragmatism is a bad idea.

Why doesnt my edictions are good to gpsyes and others?? 6 months ago

A quick search in help will give you a few similar answers on that topic.

JOSM tip for squaring buildings 7 months ago

Alt+extrude will create a second store.

The building tool is great for simple things, but I find that it asumes too much (such as the value of the building tag or the angle of the next building) while not handling L/T-shaped buildings and leaving unshared nodes everywhere. Extrude is less advanced but more versatile.

JOSM tip for squaring buildings 7 months ago

Thanks, didn't know about this one. For the given example I'd have used extrude instead anyway, but I can see other cases where it'll be usefull.