[OSM-dev] EDIT-Button on osm-website

[Martin Koppenhoefer]

2009/3/29 Dave Stubbs <osm.list at randomjunk.co.uk>:
> 2009/3/29 Martin Koppenhoefer <dieterdreist at gmail.com>:
>> I wonder if it was possible to link the "EDIT" button on the main
>> website to a page (or AJAX) with checkbox-choice for the editor:
>> Potlatch, JOSM (and any other Editor offering a
>> "remote-call-function"). This would underline the possibility of
>> alternative editors to newbies as well as offering comodity to
>> JOSM-users. It would also point out, that Potlatch is not "more
>> recommended" than JOSM. And it would maybe prevent some problems
>> connected to Potlatch-Edits (inconsistent and erratous data like
>> multiple nodes and deleted ways that still appear in Potlatch (in
>> general errors as result of not using API), insufficient zoom).
>>
>
> can you point me to the trac bug reports for these potlatch
> problems you mention?

Actually I don't know about potlatch-trac-records, but I find double
(and more) nodes on a regular basis while editing (noticed by
JOSM-Validator) and history-files always indicated potlatch-edits.

> I'm not currently aware of any problems introduced by the AMF API
> specifically, just a couple of errors due to the general lack of
> transactions.

With the term "inconsistent data" I was exactly refering to these
"couple of errors due to the general lack of transactions".

Insufficient zoom-capabilities might be the reason for some edits,
where I noticed that details I inserted were "generalized" (erased) by
other users (with potlatch). This might as well be a result of
different ideology (like "we don't get more than 5 m precision with
our GPS-Devices, why should there be details?"). Another issue in this
context is, that most potlatch users seem to don't even zoom in to the
highest level, because then the WMS-Layer disappears ("Grey
Yahoo-tiles: no data for this zoomlevel), so they might stick to even
greater distances, where the aerials are still visible.

Martin