Why are we creating a design system for OSM?

General talk
21

Indeed, this should be another forum topic to not derail the discussion here.

Since there is no single central “OSM”, just a loose-ish collection of mostly individual projects, yes it is mostly up to the people who put up the effort operate those individual projects to decide what they are willing to integrate or change about them. If you want to see something changed, you need to do the work and that might have to go beyond just creating a fire and forget pull request if it’s a major overhaul. You might have to commit yourself long-term to the maintenance and operation of the project.

Not being good enough code or not complying with the code style seem like perfectly good reasons to not accept a pull request. Or are you implying that the developers/operators are just looking for made up reasons to deny changes?

Aside from ‘simply’ writing the code, part of the responsibility of anyone requesting a change or submitting patches to a project is to check if the proposed changes are

  • a cultural and social fit for the project
  • fall within the technological scope of the project and need to be implemented there and not somewhere else
  • are functionally sound and well implemented
  • are possible to sustain with the resources available to the project, both technological and social
  • …

I’m pretty sure plenty of pull requests do not end up being integrated fall short on some of those measures.

You also seem pretty dismissive about the amount of work that goes into the day to day maintenance and operation of these projects and the amount of work that is required for reviewing pull requests. Calls to help out with those aspects of OSM projects mostly end up falling on deaf ears.

1 Like
22

I’m not, and I’m glad there are people doing the free and the hard work. However, at the same time, I believe I can criticize things that clearly are not working well.

I think it was needless to say that I was just considering “good PRs” (hence my “flawless PR” term in my last message). No one wants bad code, no one wants useless things, this is clear I believe.

But, just a random case, because I don’t want to create a huge off topic and not waste my time going to Github looking for stale PRs (because I know they are there): let’s remember the busway thing about Carto. There is simply no excuse to refuse that. Like, 0. And still, 2-3 open PRs, lot’s of supportive comments by fellow mappers, and still, it will never get merged.

This is not an isolated case about a single project or a single dev/person. This is entrenched in OSM.

Have you considered that perhaps people gave up contributing because of the current OSM culture?

Food for thought: a single dude was so upset with OSM that he decided, by himself and alone, to create a whole freaking new OSM. How upset must be someone to act this crazy?

However, I can understand him. Not only me, but it seems that several people are working for/supporting him. These same people are, however, not contributing to current existing OSM code…

2 Likes
23

The German goal is NOT to improve osm.org Map but make alternative, a better one (at osm.de)
The 3 projects could do that togeter, merge their ideas and work power. But there will be different code languages, tools, frameworks, human characters, just a dream.

We could make an OSM Wiki page with a table to compare the 3 (or more?) atempts by lines of details. So we could see commonness, gaps, incompatibilities.

2 Likes
24

In the OSM-NG thread there is endless discussion of technical merits of specific design choices and how best to proceed to get a positive outcome. Just as in Gustavos github issue.

At worst there is some subtle hinting that the projects were/are far too ambitious in scope and baby steps would be a better way to achieve some of the goals.

And yes Gustavo is proposing a awful lot of work for himself for max one and a half websites and that for an organisation that doesn’t even have a plain old style guide. Not that there’s even a slight chance of anybody following it.

4 Likes
25

politics.

The first stage is basically “creating the thing that is better”, the second is creating a track record of keeping that thing maintained, and the third is the politics associated with persuading people that you’ve done those first two things well.

Normally the third of those is the biggest challenge, but with website code the second is a big bar to clear too, it’s not like an alternate map style where anyone can develop and maintain something for a period of time, and if it breaks, people would just use something else.

6 Likes
26

Very little of it is ops. It’s the website maintainers who are responsible for the website.

To summarize the long thread, the maintainers gave him a good amount of information to move forward, the issue went mostly idle for a couple of years, then Gustavo decided the discussion would be better on a different issue tracker.

5 Likes
27

I’m still trying to understand the whole project of yours. When you posted it yesterday, I was about tired but I wanted to take a look at it, so I gave the diary entry to ChatGPT to give me a summary of what the design system is about and how it would benefit OSM currently.

Tldr, combining with the Wikimedia’s Codex you mentioned at the bottom, what I understand so far is that you want a standardization of the front-end elements and UI of OSM so that it can be used afterwards by other projects of OSMF or OSM-subprojects which may appear. In other words, I see it as the bases of what would possibly the OSM’s future, similar to what Wikimedia is today.
Did I understand it correctly? That the goal of the design system is more focused on the OSM’s expansion in regards of OSMF?

1 Like
29

Olá, @Gustavo22Soares
Qual o nome do grupo da UnB e quais são esses professores e estudantes envolvidos, por gentileza?

1 Like
30

That’s the answer what I looking for. The most important thing is not the solution, is the “convincing people who don’t want to be convinced” (or politics, as you nicely said). This should be stated in the beginning of the Contributing.md file, it would avoid a lot of waste of possible contributor’s time.

To someone who might not be aware of this behavior, this could be called a case of weaponizing someone by analysis paralysis or if want to go a little further, perhaps concern trolling. Open any PR in the OSM-website repository and you will see this behavior.

In other words, you start saying "hmmm, I support your idea, but I am just concerned about these 500 things first. " And then the person comes with the solution of the first thing from the 500 things list and then you go again “hmmmm, that’s nice you did that, but this is too ambitious, why don’t you actually do just this”. And this goes ad infinitum, until the point the person gives up and you feel good that you didn’t have to say “I won’t support your idea, no matter what you do”.

Bringing to this case, what is the minimum Gustavo could do? He suggested doing an overall redesign of the site. Of course it was too much. Then it was suggested that he tackles just one item of the list, which he just proposed (Creating a new design system). Now, we just discovered that this is still ambitious (like if this was a bad thing).

What is, then, the very minimum he could do, to be remotely accepted by the maintainers? Redesign a single button, one at a time? I’m not a designer, but I believe this is not how a redesign of anything works.

Like, again, no one here in this thread made a SINGLE criticism about the project. They complained about:

  • who created the survey (posting a link with the answer in that same link, in the 3rd line);
  • the wording of the diary (“I’m just concerned you phrased like this”);
  • the software used to create the designs and that the website is not 100% done yet (even though it’s just a proposal).
  • in the Github thread, they complained that just a Figma/Penpot page is not enough, it has to be a website instead, so he did it (and now they complained about it);
  • then someone complained that OSM doesn’t need a design system (although he proved, not just stated, that yes, we do need it).

Can you imagine Steve Coast pitching, 20 years ago, the “creation of a free map” to the current OSM dev community?

I could suggest Gustavo rewriting the diary substituting "will’ for “could”, but I understand that, no matter what he (and anyone else proposing something new) does, someone will always be “concerned” about something, except for change.

I’m glad that, despite all of this, people are still being ambitious (this redesign, the OSM-NG, the better-osm script, the Strava browser plugin, the CoMaps fork, RapiD, and all the other things thriving with community collaboration).

12 Likes
31

Politics, at this level, is simply “navigating around how humans behave”. I don’t think that that makes sense to include in any “Contributing” file. People have written books about this sort of stuff, made documentaries, etc. …

I think that’s unfair. The first advice started with “not try to do too many things at once” - that’s the reverse of what you suggest.

I have to ask at this point “what project”? I don’t think that there is anything to look at yet, is there? The alleged website 404s. Similarly, the initial post here is essentially content-free - it uses nice words like “empathy” but does not say what is planned, how it will be implemented or what tangible benefits there will be.

7 Likes
32

This is the first comment by a maintainer, with lots of questions. Then, the comment you quoted, also goes in a similar way, with “advice” and then lots of questions. If these messages (and the entire osm-webiste repository behavior) is not concern trolling, then I don’t know what it is.

If you open the links in the first message, you can see the project:

And all the other things you are asking (like how it will be implemented etc) are here: Microgrant Proposal: Atlas Design System & Modernization of OSM.org · Issue #26 · osmfoundation/ewg_bidding · GitHub

So, perhaps now people can criticize the project?

2 Likes
33

Yes, what we want is for Atlas to be used not only by OSM.org but also by other projects in the ecosystem. As a mapper and contributor, I believe OSM should follow the path of Wikimedia. I’m not part of the Wikimedia community, but I find the work they do very interesting.

1 Like
34

Somos um projeto de extensão SIGAA - Sistema Integrado de Gestão de Atividades Acadêmicas

35

I found it by clicking on the link that I linked to above - at https://github.com/openstreetmap/openstreetmap-website/issues/3785#issuecomment-1630882157 the link to https://playzinho.github.io/gaia/index.html is the one that 404s.

The new link https://gustavo22soaresh.github.io/gaia/login.html is to 5 screens (they’re non-functional, which I guess is by design) but you can go between them with “prev” and “next”. To me the look slightly different to, but broadly the same as, the current login screens. How are they “better” (or even “different”)?

How would I use this? I currently look after this site, and it looks like it was designed by someone with the artistic capabilities of a hedgehog, which it was. What would I need to do?

4 Likes
36

This project was more of a proof of concept showing that it could indeed be implemented. It was created before working on Atlas, with improvements in how the information is presented and also better adaptation to larger screens.

I’m glad you’re interested in implementing it on your site. We are currently developing the components, and as soon as we have a stable version, you’ll be able to use them. You can learn more about how to use Atlas here: Accessing Altas | Atlas — we are still working on making the components available.

2 Likes
37

(you may be already aware, but) on https://design-atlas.vercel.app/using-atlas/accessing-atlas.html the link to https://design-atlas.vercel.app/icons/overview.html 404s, as does on https://design-atlas.vercel.app/using-atlas/adrs/overview.html the link to https://github.com/joelparkerhenderson/architecture-decision-record/blob/main/templates/decision-record-template-by-michael-nygard/index.md.

2 Likes
38

Obrigada pelo esclarecimento. Você poderia consertar o nome OpenStreetMap, que está escrito de modo errado na descrição do grupo? É uma marca registrada.

39

I’ve fixed it, I apologize for the mistake. The documentation was copied from Codex, so some links may be broken or outdated, I’m checking that now.

1 Like
40

I’m curious about the previous and ongoing communication with the CWG about this project. What topics were discussed and did that lead to any other further communication with any other parts of the OSM social sphere (website maintainers, ops, the wider osm community, …)?

1 Like
42

errr sorry but he doesnt explain the project at all,at least not through any link i can see in the post.
There are lots of generalisations and friendly language but it would be helpful to understand what it is!!

1 Like