[Tagging] Data redundancy with "ref" tag on ways vs relations

Richard Fairhurst richard at systemed.net
Wed Aug 1 17:15:29 BST 2012


Petr Morávek [Xificurk] wrote:
> I apologize if my words sounded harsh or offending. I admit that I'm 
> not regular user of Potlach, so my knowledge of it is kind of limited.

I can tell... you can't even spell it. ;) (Sorry, cheap shot. But it's
PotlaTch.)

> 1) Pointless members of relations, e.g. a German train station added 
> to the Czech administrative boundary relation. My guess is that the 
> user mistyped the relation id when he/she wanted to add that station 
> to some route relation. I have no idea, why he/she did not use the 
> method, you've described, but it happens.

There's not a whole bunch that P2 or any editor can do to stop people
mistyping numbers (i.e. when loading a relation by id that's not in memory).

I think this goes back to the point I made earlier in the thread: that if
relations are to be adopted as the replacement for the ref tag, then we need
to give the API "a fast, efficient bbox-limited search function", such that
the editor can automatically load the Czech admin boundary relation via the
API rather than having to copy out an id from the wiki.

> 2) Generally broken relations, e.g. the user deleted a way that was 
> part of large forest polygon, probably because the way itself had no 
> tags. I presume this is caused by not very good visualization of 
> relations in Potlach, thus most of the time users are simply unaware 
> that they are breaking something. A simple task like seeing that 
> the way/node is part of some relation requires a user to switch to 
> Advanced mode and look at the list.

Ok. So I've just five minutes ago committed this:
https://github.com/systemed/potlatch2/commit/54e3ae56e128523104aa786363b5337f9f5a68e9

which builds on the earlier
https://github.com/systemed/potlatch2/commit/6df69becdd6990294c964b2e469c76631a3e312a

to give an indication of unrecognised relation memberships in Simple view.
That should fix that.

So can I ask you for a quid pro quo?

tagging@ is not the Potlatch 2 bugs channel, especially when phrased in
flamebait terms such as "poor editing capabilities" and "overly
complicated". There is a potlatch-dev@ mailing list and a P2 trac component.
Use them. Venting elsewhere might be therapeutic and make you feel happier,
but it generally doesn't get things fixed (it's just by chance I'm reading
this thread).

By analogy, it used to be really very common for relations to be deleted by
accident with JOSM. I posted to josm-dev@, the developers responded
helpfully, we had a conversation, and things improved (see
http://gis.19327.n5.nabble.com/Another-66-relations-bite-the-dust-td5333242.html
for the right thread). If you have a suggestion for P2 or any other
component of OSM, that's great: but write it constructively, and in the
right place.

cheers
Richard





--
View this message in context: http://gis.19327.n5.nabble.com/Data-redundancy-with-ref-tag-on-ways-vs-relations-tp5719083p5719286.html
Sent from the Tagging mailing list archive at Nabble.com.



More information about the Tagging mailing list