[OSM-talk] [Imports] Import guidelines proposal update

[Lester Caine]

Pierre Béland wrote:
> Check the list of arguments presented here for the mandatory separate account:
>
> 1. "it's easier to separate from normal contributions"
> 2. "it's more effecient for sourcing"
> 3. "it's easier to identify the source if we change the license. We
> faced that issue in the past for ODbl transition"
>
> Lester
>
> Is a separate account is the better and only way to have some metadata
> documenting imports? I don't think so.There are various ways to document imports.
To be honest I think that the 'separate account' was originally recommended for 
a single import of a complete set of data. So we all knew that this data came 
from 'xxx', but I'm not even sure now that when you select an object it still 
tells you that information?

> There were discussions on the Import listin 2009. Andy Allan opinion was that
> metadata like attribution should be on the Changeset and not on the geo feature.
> Other like Pieren suggested that it is sometime necessary to give attribution on
> the geo feature. Andy Allen also stated that using a dedicated account was
> something he less bothered about.
>
> When uploading to the OSM database, I think that the Changeset comment field can
> be used to both give attribution and indicate that it is bulk edit. This will be
> simple and as efficient. It will be easier to manage for both the contributor,
> the local chapter and the DWG.

Comment fields are not documented as well as they should be and the 'problem' 
that instigated this thread is to my view of what's on line a very good example 
of why there WAS a problem. Correctly flagging information is essential and we 
do perhaps need a little more 'automatic' actions. I can see that the French 
data is perhaps not suited to a 'single import' which is then the problem, since 
multiple imports already processed in some way are just as much a problem? Lets 
try and make the 'initial' import as clean as possible even if that has to be to 
a staging area from which packets can be taken and manually processed. 
Identification can then be married back to the raw data in a location where 
anybody can see it?

If that Knight Foundation grant is suitable I'd like to propose that it is 
directed towards the very tools I am talking about to take all the currently 
available data sources and importing them in as raw a format as possible into an 
overlay system from where they can be merged with the main database. Rather than 
the quite heroic efforts that are currently being used to import them?

-- 
Lester Caine - G8HFL
-----------------------------
Contact - http://lsces.co.uk/wiki/?page=contact
L.S.Caine Electronic Services - http://lsces.co.uk
EnquirySolve - http://enquirysolve.com/
Model Engineers Digital Workshop - http://medw.co.uk
Rainbow Digital Media - http://rainbowdigitalmedia.co.uk