[Imports-us] Tracking imports

Jason Remillard remillard.jason at gmail.com
Thu Jul 11 21:30:11 UTC 2013


Hi Serge,

I don't understand where you are trying to go here?

Why do we care about older imports that are completed and are
basically "good enough". It seems like only the ongoing imports (which
should already have wiki pages) and the older imports that we are
thinking about reverting are interesting? There are a many good enough
imports, that are what they are.

If you have a list of imports that you would like to kill via the DWG
because of a license problems, or just because the data is crappy,
perhaps just a single new wiki page keeping track of them would do it.
Or if you really want to go crazy, make new import category on the
wiki page [Sucky Import], and make import pages for the old imports.

Of coarse for fun we can start sticking the new category tag on each
other import wiki pages too :-)  Kind of like a badge of shame for an
import. It would be tons of fun...

Jason

How many are you talking about, can't be more than 10/20 in the US
that might need to be



On Thu, Jul 11, 2013 at 8:58 AM, Serge Wroclawski <emacsen at gmail.com> wrote:
> Folks,
>
> I've been thinking about how we can track the various imports going
> on. I've put a lot of thought into this, so it's a relatively long
> email.
>
> We really have three kinds of imports we need to track:
>
> 1. Imports working through the committee
> 2. Old imports that have been done in the past and are understood
> 3. Rogue imports, some of which may need to be addressed by the DWG
>
> Let's break each of these types down into what they need:
>
> 1. Imports working through the committee
>
> For imports working with the committee, we need a way to track them,
> whose the point of contact for them (or points of contact), the data
> source, the license, the programs used to do the data analysis, what
> account will be used for upload, etc.
>
> We need a way of tracking where in the process we are, and ideally a
> way to use this information in the future when going back and
> analysing a previous import.
>
> 2. Old, good imports
>
> We have a bunch of old imports which are not properly documented on
> the wiki. Unfortunately we also don't have a mechanism to edit those
> changesets and add any appropriate tags (such as links to the wiki).
>
> We should be tracking these, as least in making sure they're all
> properly documented
>
> 3. Rogue imports
>
> It seems like every time I scratch the OSM database a little more, I
> see more rogue imports.
>
> Some of these imports fit into the second category, and need documentation.
>
> Some of the imports are made by folks who might be good importers if
> given the opportunity.
>
> And some of the imports are unsuitable for one reason or another and
> should be either deleted by us, or handled by the DWG.
>
>
> I've been thinking about a practical solution that will fit all three
> situations.
>
> My proposed solution (imperfect as it is) is to set up a ticket
> tracking program, such as Roundup or RT, and then use as a mechanism
> to track things down, and keep the wiki updated.
>
> Ideally, I think there should be one tool that does both the
> documentation part of an import, holds the original data, has a
> searchable index, etc. but I started such a tool a few years ago, and
> it was quite an undertaking, and I think we need something right now.
>
> I'm really interested to hear others' thoughts on this.
>
> - Serge
>
> _______________________________________________
> Imports-us mailing list
> Imports-us at openstreetmap.org
> http://lists.openstreetmap.org/listinfo/imports-us



More information about the Imports-us mailing list