Working Group Minutes/EWG 2021-10-18

From OpenStreetMap Foundation
Jump to navigation Jump to search

These notes are a draft.

Formal proceedings

Minutes written by Roland.

New members

Should we clarify the criteria for acceptance to EWG, since there are more applications recently? No, because we have no criteria. Everybody against whom we do not raise objections can join right now.

Actual business

Reply from the DWG

We agree that the list is too much to get through in an hour for a meeting. Instead we organize it as a homework for the next meeting:

Every member on its own shall rate as many of the feature as possible whether they are considered to be how hanging fruit, rather complex tasks, or it is unknown how difficult it is. Comparing different ratings or finding task where there is only one or no rating at all then shall help to find good candidates to start with.

In the beginning we will also need to set up and then get used to the process, thus we shall start with small but at least somewhat useful projects.

Our ideas

Again homework: Every member on its own shall make cost estimates no matter how sketchy for each of the ideas from the pad. Also we shall ask the maintainers whether they would merge a result if the task is done. Also of interest is the expected benefit to the OSM ecosystem.

Again in total we aspire to get able to decide which projects are doable in limited time as a first round of productive activity. We do not need precise figures right now, but it does make sense to prepare tenders for projects that are small enough that they can be funded.

We do not talk about all of the ideas from the pad, and about most of them rather briefly.

We do not expect that projects fall in the scope of the Data Working Group, but rather of in the scope of us because nobody else is preprared to fund software deveopment.

Pedestrian and bike routing on areas

Upgrade an existing OSM-based routing engine (preferably GraphHopper or OSRM) to support routing across plazas mapped as highway=* + area=yes.

There are well known algorithms for this and probably implementations for other routing engines. The rationale is to de-incentivize the mapping of virtual paths across Pedestrian areas, thus there should be something to point at when discussing with fellow mappers.

Preparatory steps for data model improvements

This is a large topic that we won't be able to wrap up with a single grant project. So the idea is to focus on merely getting the ball rolling again.

No actual API changes yet, focus on potential data model changes first.

Data model changes in scope for this project: - Area datatype - Identity-free way nodes - possibly some minor additional improvements (e.g. legal characters in keys, clarifying edge cases)

It is suggested to start with a conversion software for the then still experimental file format. We probably can only realistically evaluate whether the data model change is an actual improvement when we see it in action on the whole Planet file. Thus there should be a conversion tool that can handle both excerpts and the full Planet file.

Only when the data model change indeed turns out to be beneficial then there should be a proxy such that on the inside the new data model can operate and on the outside older software can interact based on the old data model.

We plan to do at the moment only rather step one.

Unicode compliance audit of OpenStreetMap

As a prerequisite to better multilingual capabilities in OSM, we should do a review of character encoding within the OSM dataset. Current known problems include Burmese, where there's a significant amount of data in non-Unicode that results in garbled text on OSM Carto. Brandon (EWG) has reached out to Burmese users for comments.

Brandon will writes a blog post about it, the working group will promote it. There is the local chapter's congress on 6th Nov, and the local chapters may know other places where non-Unicode data plays a role. Could ask the local chapter community - there's a Telegram group for the public under the "Get in Touch".

Can the main database catch this when entering data?

Idea pad

Bookmark: Pad for idea brainstorming https://pads.ccc.de/P5UpTz8M4K

Budget

We do want to ask the board for this budget.

We have different expectations whether is realistic to have project ideas from outside this working group ready to implement already in 2022. Therefore it does not make sense to breakdown the budget in that dimension.

Any other business

Further homework: Please review the Microgrant applications based on the report about the first round of Microgrants.

Next meeting: 2021-11-01