Grants:IEG/Wikimedia Maps Rendering Improvements
I am withdrawing my IEG proposal, as I have taken on other work which conflicts with this time-wise. I'm leaving up the original text, as the Discovery team may want to refer to it |
Project idea
[edit]What is the problem you're trying to solve?
[edit]The current road labeling for the Wikimedia Map style has defects that result in too many duplicate road labels and unclear road labeling. The lack of public transport information is a critical issue for acceptance on Wikivoyage.
The current toolchain for borders used does not capture the details of border disputes which are tagged in OpenStreetMap.
What is your solution?
[edit]- Reduce duplicate labels by name "glomming"
- Shorten labels by abbreviating road names
- Develop a new border toolchain which captures disputed borders
Project goals
[edit]Improve the quality of maps available for Wikimedia projects
Project plan
[edit]Activities
[edit]- Road glomming
- Derive test dataset: Developing the SQL will be easier against a test dataset, where different scenarios can be crafted
- Fix bugs in existing roads query: There are some bugs (e.g. misordering of labels) which should be fixed before developing a more complicated query
- Design query to glom roads: This will take a few days of thinking, drawing on the whiteboard, and testing with psql + qgis.
- Integrate sub-query: The sub-query will need integrating into the roads query. CTEs might be useful here, but care needs to be taken not to be an optimization fence
- Test query with tile boundaries: Label clipping at tile boundaries is a well-known issue. There are almost certain to be problems.
- Fix label clipping: This is expected to take a few days, based on experience with past label problems. Multiple iterations are likely
- Submit pull request for style
- Reply to feedback on changes
- Abbreviation
- Research existing abbreviation lists: Others have tackled this problem before
- Derive list based on existing lists
- Test list based on a multilingual list of names: Using English abbreviations in a German road name would be bad
- Publicize list for wider feedback: Both the Wikimedia and OSM communities may see problems with the abbreviations
- Borders
- Evaluate osm-borders tool: There's an existing tool for extracting border information an OpenStreetMap planet file, but it hasn't been updated in years and relies on depreciated software
- List OSM tags used to indicate a disputed or maritime border
- Identify if sufficient information is in the rendering database: It's not immediately obvious that there is enough information in the rendering DB to construct border linestrings and get disputed information from the ways
- Design test dataset: It's easier to test on something that isn't planet-sized
- Investigate SQL query to create admin lines table
- Investigate ways to update the admin lines table without regenerating all of it
- Evaluate osm-borders tool: There's an existing tool for extracting border information an OpenStreetMap planet file, but it hasn't been updated in years and relies on depreciated software
Budget
[edit]The sums are before taxes. If the scope changes, costs will need to be revised
Developer: 15 000 USD
weeks | @ | sum | |
Road glomming | 2 | 2500 USD | 5000 USD |
Abbreviation | 1 | 2500 USD | 2500 USD |
Borders | 3 | 2500 USD | 7500 USD |
Total | 15 000 USD |
Community engagement
[edit]The community will be engaged through the maps Talk page, and development will be in the open.
Sustainability
[edit]All work is expected to be in existing codebases maintained by Wikimedia. I will be making an effort to keep the code reusable by other projects.
Measures of success
[edit]Need target-setting tips? Note: in addition to your project-specific measures of success, you will also be asked to report on some Global Metrics at the end of your final report. Please keep this in mind as you plan, and we'll support you as you begin your project.
Get involved
[edit]Participants
[edit]Paul Norman
[edit]I am an OpenStreetMap community member and developer. I have done PostGIS, PostgreSQL, OpenStreetMap and Open Data consulting, with expertise in
- Extracting data from OpenStreetMap
- Raster-based tile rendering
- Vector-based tile rendering
- Map style design and implementation
- PostGIS optimization
- Transforming data to different schemas
All of these skills are relevant to the proposed project.
I am currently a maintainer of
- osm2pgsql: This is the data converter Wikimedia uses to load OSM data into a database, as well as nearly everyone else
- OpenStreetMap Carto: This is the Standard style on OpenStreetMap.org
- ogr2osm: This software is used for converting data to OSM format
I have also worked on the MapQuest style, MapQuest Open style, and a new style designed to replace it.
This summer I was a Google Summer of Code mentor for a student who successfully redesigned the OpenStreetMap Carto road colours. I am a past Google Summer of Code student, having worked on the OSM API.
Community Notification
[edit]Please paste links below to where relevant communities have been notified of your proposal, and to any other relevant community discussions. Need notification tips?
- https://www.mediawiki.org/wiki/Talk:Maps
- https://en.wikivoyage.org/wiki/Wikivoyage_talk:Mapmaking_Expedition
- https://en.wikipedia.org/wiki/Wikipedia_talk:WikiProject_Maps#Proposed_IEG_project
- Issues will be filed on Phabricator or other relevant issue trackers for ongoing communications
Endorsements
[edit]Do you think this project should be selected for an Individual Engagement Grant? Please add your name and rationale for endorsing this project in the list below. (Other constructive feedback is welcome on the talk page of this proposal).
- Community member: add your name and rationale here.
- Support 100% - Any help with the border dispute indicators and better labeling will be great. Also, Paul has already consulted us on a number of issues with the Maps, and having someone with Paul's expertise would be a major plus to the Maps initiative Yurik (WMF)
- Support it looks like a good idea. If you need some help with abbreviation in Italian, contact me.--Alexmar983 (talk) 19:41, 2 October 2015 (UTC)
- Support I'd really love to see this happen Tfinc (talk) 23:24, 9 October 2015 (UTC)
- Support Sturm (talk) 03:07, 16 October 2015 (UTC) Sounds a very desirable effort. If you need some help with abbreviation in Portuguese, contact me.
- Support Many of the Wikimedia projects (most notably Wikipedia and Wikivoyage) will strongly benefit from having better maps. Useful developments in this direction should be supported. --Alexander (talk) 14:59, 29 October 2015 (UTC)
- Support Csyogi (talk) 18:14, 29 October 2015 (UTC)
- Support Wikivoyage would greatly appreciate better maps more suited to our guides. JamesA (talk) 00:09, 30 October 2015 (UTC)
- Support PNorman is a competent coder with deep understanding of the topic, and an improved vector maps style and process is very important for the future of wikimedia maps. Would allow for so much more flexibility. 79.27.230.90 08:54, 30 October 2015 (UTC)
- Support Paul Norman is deeply experienced with the technical challenges of rendering OSM data, and he is also extremely familiar with the nuances of the OSM data model and its community's tagging practices. The Wikimedia maps will benefit significantly from his skills and knowledge. --Almccon (talk) 18:36, 1 November 2015 (UTC)
- Support This seems like a small amount of money for much needed expert discussion and insight and the requesting person already has experience working in this space. The kinds of activities proposed here are the sort that volunteers cannot do but also would bring outcomes that would increase the efficacy of what volunteers can do. Just having someone do documentation in this space about problems and challenges would be very useful. Blue Rasberry (talk) 14:22, 6 November 2015 (UTC)
- Support I totally support Paul, he is a very competent developer and very knowledgeable of both the OSM ecosystem and technology stack. --Simone Cortesi