Using the forum as the main discussion platform for tagging

I’d say the mailing list does require an external tool: an email address and client. Most likely everyone already has this, but it is still an external tool. With the forum, a new user can follow a link, sign in, and immediately reply to a thread right on the same platform where they read it. With the mailing list a new user follows a link to the archive page and they can read the thread there, but not reply. In order to reply they must move to a different platform (email client) instead of the original place where they read the conversation (list archive).

This discussion is taking a hilarious turn. If you do not have an email address you cannot sign up to OSM and if you cannot sign up to OSM then you cannot participate in this forum. If you use a throwaway email address you will not receive messages when people comment on your mapping and you will rightly be kicked out for not being a good OSM citizen. Having people who are not good OSM citizens have a say in what tags OSM is using is ridiculous. No email - no service.


As someone who has been subscribed to the mailing lists for a very long time, I’m sure it sounds ridiculous to you, but please consider what it’s like for someone who is not yet subscribed, or does not wish to subscribe, to the mailing list. Of course they have an email address, but subscribing to the mailing list is a separate process than signing up for OSM. The point is that if I send a link to this mailing list thread to another mapper who is not subscribed, they can’t reply at the link I sent them. With a link to the forum, they can. It’s very simple.


I feel that focusing this discussion on mailings lists vs discourse is probably not going far, individuals will keep having their own personal preferences.

Instead, as I suggested, I would frame a small test (one or two tags only) to validate if these forums can productively host a tagging discussion and see and present the outcomes to everyone, based on data, not only opinions or preferences.



Agreed, though the focus here is not in and of itself forcing users to switch to discourse, but finding advantages and disadvantages; so that it will become easier for people to decide whether to use the forum or not and it’ll be clear to everyone what are the problems with the forum in order to find some workarounds.

We are not sure about how discourse behaves in particular situations; could you please address these points?

  • When a new topic is created from an existing one, are all the participants noticed about it? (here)
  • Is it possible to easily understand which topics are related to each other? (here)
  • Are users participating in the discussion notified about edits to previous posts? (here)

Knowing how the forum works will make it easier to find a good way to utilize it.

Original topic owner is notified.

Replies created as new topics will appear as link to the original topic message. I’d suggest appending a unique tag to all topics to keep them searchable.

People is notified if the original message has been edited or a message you replied or liked was edited.

I think trying to anticipate every use case (or if they are blockers for most people or not) is very time consuming, that’s why a simple test will help people understand if there are any other additional gaps through the process.

As the admin of the tagging@ mailing list I’ve long anticipated that tagging discussions will move here in time.

The list format is not particularly conducive to participation by subject experts, because the prime way of interacting with it is by subscribing to the list; unless you use an email client/provider with significant threading and filtering capabilities, that means your inbox ends up bombarded with tagging discussions, most of which you will have no interest in. The result is that the tagging@ list is primarily composed of “people who like talking about tagging” which is not a great outcome.

However, is still effectively in setup mode and there is a lot of ‘meta’ talk, so the time is not yet.

I wouldn’t anticipate that the wiki “voting” procedure needs to move here. If people want to vote on what particular words appear on a wiki page (which is all the voting procedure is), it makes sense to keep that on the wiki.


2 posts were split to a new topic: Edit messages time limitation

True, but smaller discussion can also happen in one post. See this and original post. The original topic has over 50 comments already. Several small things were covered in a single topic which went fine

Have you noticed that now the original topic offers a summary of messages? It will pick the 10% most relevant ones (likes and replies).

Multiple replies to one message can be expanded from the original message, allowing one level of threading some people was asking about.

1 Like

Sweet. That is indeed an useful feature.

this is relatively new though, before you could use nabble to read and post, but they are not there any more (AFAIK)

Why not move the whole proposal process here?

I mean, create proposal as wiki post, have discussions and voting on Discourse and then copy the final version of wiki post to wiki, with a link to proposal (and discussion and voting) in Discourse?

Whole proposal process should be dealt in one place. It does not matter which one as long it is one place.

There are some complications to this.
As explained here (Edit messages time limitation) the creator of the topic cannot edit the post for as long as he wants.
Uses with trust level 0 and 1 can edit posts only during the 24 hours following the post creation; while users with trust level 2 can edit post for one month.
Attaining level 2 should not be too complicated, but it still requires the user to log in the forum in 15 different days; thus requiring at least half a month before they can manage such a discussion.

30 days to initiate and conclude a discussion looks like a good amount of time, however the approval of any feature requires at least one month to be concluded: the voting on the proposal has to be started at least 2 weeks after the RfC and the voting itself has to last two weeks.
Thus, one month to maintain the discussion could not be enough.
Under these circumstances this solution might not be appropriate to deal with this kind of discussions.
Some other possible solutions to keep the discussions ordered have been proposed:

  • Turn a topic into wiki mode where anyone with TL1 can edit. (here)
  • Append tags to a topic and all the related ones to make the easy to search and to subscribe to the topic (here)
  • Increase the edit time limit or remove the limitation completely (here)

The main problem with this would be splitting the place where the proposal will be published and the place where it is voted. The full text of the proposal might be proposed on the forum to obviate to this, but that might generate some problems since the data would then have to be transferred to the wiki.

The wiki is one necessary step in the proposal process, since approved proposals need to be published there. If the whole process has to be dealt in one place, then that place should be the wiki because it’s the only platform that is strictly necessary for it to take place.


The wiki is the proposal process. The proposal process determines how tags are documented on the wiki. Nothing else. You are free to use whatever tags you like in OSM no matter what the wiki says.


That’s exactly what I meant by

Yes, we should keep proposal process in one place. So let Discourse be this place and only after accepting the proposal, the article would go to wiki. With a link to proposal on Discourse so that everyone can easily find the discussion and voting.

Please, discuss about voting and proposal in the dedicated topic: Moving to the new forum for proposals and voting

Here we should focus on identifying problems and advantages about using this forum as the main place where discussion about those proposals should take place.

@nukeador is there any way to move these posts to the correct topic?

Can you help flagging the messages that should be moved? Thanks!