When inspecting data/lpp/stops.csv, I noticed that LPP’s capitalization of bus stop names seems to be inconsistent. LPP seems to spell some, but not all, stops with ALLCAPS.
For consistency and readability, I suggest to divert from LPP’s capitalization scheme in OSM:
Only capitalize the first letter of a bus stop name (e.g. Barje), even if LPP capitalizes the whole word (BARJE).
Exception: If the bus stop name is an acronym (e.g. AMZS), then capitalize the whole word.
In composite words, apply this rule to the components, e.g. spell BTC-Atlantis in OSM.
I fully agree on this. Most search engines will disregard the casing, looks nicer on the maps, but if any rendering requires ALL CAPS for aesthetic reasons it can still do so.
Also, there is ref if it needs a perfect match.
The purpose of that github repository was to compare the OSM data with official LPP one. Currently it is only possible to see the current data and changes in the bus lines / stops via git diffs. The data is being updated daily.
A similar list as under “LPP - Ljubljanski Potniški Promet”, including QA, can be created using PTNA.
I can configure this for you and it will run each night and create an overview on what’s expected and what’s actually mapped in OSM.
Example: Zagreb
Besides this, do you have GTFS data for your area? PTNA can analyze and prepare it as well.
I was not familiar with the PTNA, but it looks really promising, thanks for setting up PTNA for Ljubljana!
We should probably consolidate network and operator names, as they vary a bit (abbreviations etc). Analysis diff view can help us track changes
Indeed, the LPP operator in Ljubljana offers the official GTFS feed at https://data.lpp.si/api/gtfs/feed.zip and could probably be nice to include it for reference.
I’m not yet sure about the wiki sub-pages purposes:
CSV list could be useful, but likely redundant to GTFS feed? We could also generate it from lpp/lines.csv, which is scaraped from official lpp.si website
/Analiza wiki sub-page with just the link can be included in the main page and
we can also discuss PTNA in the Talk of above page there if needed?
@ToniE, please replace the “MPP” in the name with “LPP” as this is the company name abbreviation (Ljubljanski potniški promet). MPP stands for Medkrajevni potniški promet, which are wider regional bus routes, operated by the same company, but with different tariffs and less frequent.
You’re welcome. the two links here do not work any longer: replaced ‘MPP’ in the URL by ‘LPP’
Yes, they are included as a side effect of the Overpass-API query: this searches for parent (parent, parent, …) relations of route and route_master relations. As there are some Flixbus routes passing your area and some of them use the ferry at Constance @ Lake Constance, those relations are included in the dataset, and as the ‘network’ value is empty for the ferries, they are listed.
Yes, I had a brief look into the data, seems OK. I’ll check that - latest on Saturday.
Yeah, a static URL is preferred. Also a Webserver supporting HTTP “HEAD” with “last-modified” header will ease the evaluation of the release date (curl -sI)
Yes, agreed. The OSM-Wiki-page allows some layout control thought.
PTNA’s GTFS analysis supports creating this list from GTFS sources - without layout commands
Both pages are for historic reasons but not needed. Discussions (of general interest) can also take place on my private OSM-Wiki page for PTNA. LPP related discussions can take place here or per PM (PM not useful to share information globally or country-wide)
@ToniE the license is not explicitly stated, but it is coming from a company owned by the city municipality of Ljubljana, which must respond to freedom of information requests, and such data is usually licensed with some sort of CC-BY license
Regarding licensing, I think I might be able to get explicit permission from the agencies for inclusion in OSM. Is there any kind of template for that?
For a template: all waiting, negative and positive attempts to get permissions will finally be documented in the OSM-Wiki permissions page, maybe you’ll find something appropriate - I know only some in German language.
looking good. And “re:” the “casing differences”: doesn’t make a difference for you at the moment.
This is relevant only if you have two (or more) bus lines with the same number (‘ref’) and same ‘operator’ in different areas. PTNA then uses ‘from’ and ‘to’ of the relations to compare them with data from CSV (from and to). So, you seem to be safe at the moment.
Example: Busses ‘A’, … appear quite often in DE-SN-VMS (‘A’ 11 times) in different cities, partly with same ‘operator’
)