[OSM-talk] junction_ref=<whatever>

David Earl david at frankieandshadow.com
Fri Mar 16 12:27:57 GMT 2007

> I appreciate the work you're putting into the proposals process.
> However, I won't vote on this, or practically any proposal, without
> seeing what is actually being proposed (as opposed to the general
> concept, which is often quite clear). I see a long conversation on
> that page, but nothing else. Nothing technical, nothing specific.

That's because the page is a link from the proposed_features page (linked
from the top), which is where the tag and context is stated explicitly.

> On the page for each proposal, please make sure that the proposal is
> clearly stated. Is this applied to a node, a way, an area etc?

It quite clearly says in the proposed_features page in the appropiate column
that it is applied to a node.

> Is it junction=, junction_ref= or what?

It is junction_ref, again as stated in the proposed_features page in the
appropriate column.

> It's even worse if I read the
> caveats that people have put on their existing votes...

Well, I can't stop people doing that! But see below.

> (and on this one, i'd rather see numbers and names in a separate
> space, rather than backslash, semi colon or colon deliniated as the
> various comments are intermingalling. But I won't vote either way when
> I can't see what I'm voting for.)

The proposal is for junction_ref=string applied to a node, with no specific
syntax in the string (that was someone else's comment: I made no other
proposal). There is some discussion about alternatives. Again, thosedon't
constitute a proposal.

But I do agree that this voting system doesn't work properly:

(a) too few people participate to make it really meaningful (part of that is
because you have to notice a vote is on, unless mailing lists are directing
you to it, as I have tended to do). That also means things can get through
on one vote because not enough people see it in time to vote against.
(b) it relies on the implementors of the renderers to actually act on
approved proposals, or they are meaningless. Essentially they have a veto.
(Yes, you can change style sheets, but only for your own renderings, not the
central slippy map, or tiles that a real consumer is going to use; and then
not all proposals follow the pattern of just some new line/icon/text).
(c) there is no mechanism for making alternative suggestions and refinements
and incorporating them into the proposal.

In a voting system governed by standing orders, there are rules to avoid
this kind of confusion. What happens is the proposer makes a proposal.
Alternatives are addressed by amendments which are voted on in turn and
individually (ideally with all the amendments available before the process
starts). Amendments which are accepted becoem part of the substantive
proposal, and those not accepted fall by the wayside. After each approved
amendment therefore there is a revised proposal, which can be further
amended, and after all the amenments have been considered there isa final
proposal which can be accepted or not in its final form.

This is cumbersome, and lots of people get put off from particiaption in
democratic processes because of it, but it is a tried and tested way of
getting to the end while always beng clear what the proposal is. Given these
are only Wiki pages, though, it is hard to see people sticking to the rules!

In the meantime I have clarified that I intend the vote to be on the
original proposal and reproduced the information in the proposed features

I believe from user Higgy's comments his *approve* should be *disapprove*
and C2r is probably still *approve*. Would either of you like to amend your
entry to make your vote clear.


More information about the talk mailing list