[OSM-talk] SteveC should decide
Andy Robinson (blackadder-lists)
ajrlists at googlemail.com
Thu Oct 1 15:28:01 BST 2009
Matt Amos wrote:
>Sent: 01 October 2009 2:25 PM
>To: Russ Nelson
>Cc: talk at openstreetmap.org
>Subject: Re: [OSM-talk] SteveC should decide
>On Thu, Oct 1, 2009 at 1:56 PM, Russ Nelson <nelson at crynwr.com> wrote:
>> Andy Allan writes:
>> > can win, and I decided about 6 months ago to ignore (or at least not
>> > rise to) the eternally pointless tagging debate.
>> The reason it's eternal is because there's no one to choose.
>surely it has more to do with different people having different opinions?
>> refuses to do it because too many people give him hell when he does,
>> and nobody defends him for doing it. Fine. I'm officially defending
>> his right to say "This is the SteveC(tm) compliant way of tagging."
>> without catching grief from people with more time to argue than edit.
>i absolutely agree. i'd also defend frederik's right to say "this is
>the Frederik Ramm approved tagging scheme" without catching grief, or
>andy to say it's the One True Gravitystorm way, etc... etc...
>> I'm fine with anarchy; I'm an anarchist. But chaos is not appealing
>> to me, otherwise I'd move to Somalia.
>you're fine with anarchy, but you'd like an organisational method of
>resolving arguments? anarchy + rules?
>free-form tagging was genius, but it was based on the idea that these
>"folksonomies" (urgh, i hate that word) naturally converge by people
>using them, tools consuming them and the feedback loops that creates.
>we need to be strengthening those processes - not defining rules,
>elevating adjudicators or otherwise compromising the original awesome
>genius of free-form tagging.
The reality is that there is natural convergence for most tags. Now thats
not to say there won't be more than one tag for the same thing, there will.
It just means that something is tagged predominately one way rather than any
other way. Does that make the other tags wrong? Of course not. It would only
be wrong if it made no logical sense and nobody could understand it.
So I agree with Matt. To me its not the tags that need controls its the
process by which we select them. Mostly I guess tags (as the originator of
Map Features I can remember most of the basics and have a good idea, my
personal idea, of how to build up tags beyond that. Now, if I guess and had
some tool tell me that my guess was close to some other tags (A top 10 list
if you like) I might choose differently. That would be better than
guesswork, but still allow the freeform tagging that permits me to get stuff
mapped quickly. I dont have time to read the manual!
More information about the talk