[OSM-talk] Advanced Relationships

Peter Miller peter.miller at itoworld.com
Wed Jun 20 08:14:42 BST 2007


 

>It is already a widely (not universally) accepted consensus that we don't
need segments in their present form. Abolishing segments would make >many
things easier. It is just that nobody got round to implement it yet (and it
has to be implemented in several places at the same time).

 

I think you need to be careful about saying that a change is 'widely'
supported. My experience says that there is a considerable risk of a
minority agreeing on a particular change and working up their ideas and
enthusiasm only the find serious resistance when actual implementation is
proposed. This can lead to frustration and worse. I think it is essential to
agree the objectives first and get buy-in for these and for the process
before going much further. Could you clarify:

 

What problems are being solved by the 'abolishment of segments'?

What other changes would accompany the 'abolishment of segments'?

How will you get agreement for the changes?

What will be the implications of the change for the tools and how will the
editing process change?

 

Personally it would seem to be appropriate to deprecate adding tags to
segments and indeed change JOSM so that it isn't possible to attach tags to
segments as this was one of the major confusions I fell into when I started
using it.

 

 

 

 

Regards,

 

 

 

Peter

 

 

 

 

------------------------------

 

Message: 3

Date: Tue, 19 Jun 2007 10:09:16 -0400

From: Matthias Julius <lists at julius-net.net>

Subject: Re: [OSM-talk] Advanced relationships

To: talk at openstreetmap.org

Message-ID: <87r6o8jasz.fsf at julius-net.net>

Content-Type: text/plain; charset=us-ascii

 

Ian Sergeant <isergean at hih.com.au> writes:

 

> Since all the current proposals being made both here and on the wiki 

> involve linking segments, if you think that is not a good idea to link 

> segments, then now is the time to explain why and possibly expand on a 

> better alternative.

 

As I understand it advanced relationships are intended to be a generic way
to link any types of object together.  They don't depend on the existence of
segments.

 

Matthias

 

 

 

 

------------------------------

 

Message: 4

Date: Tue, 19 Jun 2007 16:21:30 +0200

From: Frederik Ramm <frederik at remote.org>

Subject: Re: [OSM-talk] Advanced relationships

To: Matthias Julius <lists at julius-net.net>

Cc: talk at openstreetmap.org

Message-ID: <7CE0DC09-61A9-4D6C-BD98-0362923EED2C at remote.org>

Content-Type: text/plain; charset=ISO-8859-1; delsp=yes; format=flowed

 

Hi,

 

> Ian Sergeant <isergean at hih.com.au> writes:

> 

>> Since all the current proposals being made both here and on the wiki 

>> involve linking segments, if you think that is not a good idea to 

>> link segments, then now is the time to explain why and possibly 

>> expand on a better alternative.

> 

> As I understand it advanced relationships are intended to be a generic 

> way to link any types of object together.  They don't depend on the 

> existence of segments.

 

I concur.

 

It is already a widely (not universally) accepted consensus that we don't
need segments in their present form. Abolishing segments would make many
things easier. It is just that nobody got round to implement it yet (and it
has to be implemented in several places at the same time).

 

I would be prepared to use (or re-introduce) segments as part of a
well-strucured data model like the GDF wher they form the base layer on
which everything else is built. But currently segments are really just an
extra layer of indirection, tempting people to do all sorts of things they
really shouldn't.

 

The "advanced relationship" proposals do not require segments.

 

Bye

Frederik

 

--

Frederik Ramm  ##  eMail frederik at remote.org  ##  N49?00.09' E008?23.33'

 

 

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/talk/attachments/20070620/a8414077/attachment.html>


More information about the talk mailing list