[OSM-talk] Flickr Now Supports OSM Tags

[Dave F.]

Frederik Ramm wrote:
> Hi,
>
> Eugene Alvin Villar wrote:
>   
>> What's so hard about standardizing on the boolean values given 
>> appropriate changes to editor presets, good wiki documentation, and a 
>> deprecation period for other boolean values?
>>     
>
> It's a kind of slippery slope situation. There is fear that once it has 
> been proven that standardisation works for true/false values, there will 
> be demands to standardise everything else as well.
>   
Not *everything* just the things that people feel need changing

Fear is not a good reason for the status quo.
> This would be positive for the users of our data in the short term 
> because it means they would not have to interpret the data; however it 
> would remove dynamism from the project and require mappers who want to 
> invent something new to apply to the standardisation committee first, 
> and we feel that this would be a severe detriment to further 
> participation on the mapper side. OSM flourishes partly because mappers 
> feel that they can help shape the project, and contribute what they 
> think is important, rather than just being mechanical turks (without the 
> payment).
>   
People would still think they would be contributing even with certain 
restraining guidelines.

> In the long term, standardisation would kill the project, and thus not 
> be desirable even for our users. - Coming from the outside and not 
> having the knowledge about OSM that we have
I find it very disappointing you feel there is a them & us situation.
> , users can be forgiven to demand 
Whoa, there. Who's demanding? Please, don't make things up.
> things that would ultimately destroy OSM,
>   
>  but it is our duty to 
> educate them and to explain to them that they can either take OSM as it 
> is, with some interpretation required, or they can demand that OSM 
> change but that would, in the long run, probably mean no OSM at all.
>   
This is starting to sound like quasi religious mumbo-jumbo:
'if you do this, the sky will fall on your head'

> I run a small company that, among other things, sells standardised 
> derivates of OSM data. I spend a lot of time trying to stay ahead of the 
> game, analyse what tags people use and for what, and try and convert 
> these into consistent and reliable values.
If there were certain restrictions you'd have to spend less time.
>  If OSM changes from 
> "landuse=forest" to "russ_nelson_sees=trees" because that's what mappers 
> what to use, then I can adapt and my customers don't have to, and 
> neither does the OSM community have to twist and turn just because some 
> users want consistent tagging.
>
> In my eyes, this is the way to deal with standardisation - do not force 
> it upon the mappers, but instead create a "filter layer". In my case 
> this is a commercial operation, but I have been suggesting for ages that 
> instead of writing bots to streamline OSM data, why don't people write 
> generic filters/standardising engines that take the "chaotic" OSM data 
> as of today and produce well-ordered standardised output for people "out 
> there" who cannot be bothered to keep up with OSM's tagging anarchy? It 
> would not be too hard.
>   
Isn't that just putting an artificial middle man into the equation?
Wouldn't it be better to have an organised, original data?
> And I'm not saying this because of my business (until now, keeping up 
> with changes and doing the standardisation takes more work than I get 
> paid for it so I would benefit from OSM itself being standardised); I 
> truly believe that the way things work in OSM, with "standards" being 
> un-enforceable and people constantly deviating from them (even if there 
> is a certain base consensus on many things) is the only way it *can* 
> work without degrading into some kind of Google Map Maker that does not 
> look for project members, but for worker ants.
>
> Bye
> Frederik
>
> _______________________________________________
> talk mailing list
> talk at openstreetmap.org
> http://lists.openstreetmap.org/listinfo/talk
>
>
>