[Talk-us] Brainstorming an Import Tool

Joe Woycke woycke at gmail.com
Tue Aug 17 13:31:20 BST 2010


i think that more than a tool there should be a focus on a way to evaluate
data that is transparent.  whether it is the state of arkansas or the new
tiger data is how the data is processed and which set of data is decided on
as the most accurate.  the current model relies on the community to review
data and the wiki wars to settle the question of which data is "best".  but
that will become more problematic to deal with as the amount of data grows.
 not to mention, how do you deal with information that is no longer there
(an abandoned street or changed alignment) but still has relevance.

transparency ought to be a concern.  if all of the data is available, with
one map presented to the casual user, the more committed members of the
community could actually dig deeper and examine the differences between say
the tiger data, the reports from the community, the streets reported by the
local government, and the state or any other group or entity that is
reporting. the problem then becomes a question of processing information and
doing so in a clear transparent way that mediates concerns between usability
and integrity of the data.

joe woycke

On Tue, Aug 17, 2010 at 8:18 AM, Ian Dees <ian.dees at gmail.com> wrote:

> On Tue, Aug 17, 2010 at 3:55 AM, Frederik Ramm <frederik at remote.org>wrote:
>
>> Ian,
>>
>>
>> Ian Dees wrote:
>>
>>> I got the impression after SotM US that there was a huge interest in
>>> doing imports correctly. For me, correctly means the following:
>>>
>>
>> 0. Discuss with community (don't import if no community exists)
>>
>>  1. Get permission
>>>
>>
>> 1a. Discuss with community (don't import if no community exists)
>>
>>
>>  2. Convert to OSM format
>>>
>>
>> 2a. Discuss with community (don't import if no community exists)
>>
>>
>>  3. Compare to existing data
>>>
>>
>> 3a. Work with community (make tools that let LOCAL community do this
>> comparison)
>>
>>
>>  4. Upload to the data
>>>
>>
>> no:
>>
>> 4. provide tools/mechanism to let community upload the data in their area
>> of interest. If no community exists, do no upload data.
>>
>> I think that far too much weight is put on the technical side here. In a
>> place with a thriving community, it may not take more than a WMS server to
>> get the "import" going. In a place without a thriving community, NOBODY
>> NEEDS A (third party) IMPORT - the import will only make it harder for a
>> community to form.
>>
>>
> I agree, but in places that do have a community, I think a tool would be
> helpful. That's why I asked about brainstorming for an import tool, not
> brainstorming for community. While that's extremely (and much more)
> important, we did a lot of that during SotM US and the need for a tool like
> this was identified -- especially when the 2010 Tiger data comes out.
>
> _______________________________________________
> Talk-us mailing list
> Talk-us at openstreetmap.org
> http://lists.openstreetmap.org/listinfo/talk-us
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/talk-us/attachments/20100817/c6e8f77c/attachment.html>


More information about the Talk-us mailing list