When I say 'should', I mean 'should' in the sense of 'should if they want to make the widest use of data possible'. Obviously, there's a trade off between the amount of time/effort it takes to support more tags, and the extra data this gives you. In the case of yes/no vs true/false vs 1/0, I'm not sure how widespread the second two variations are (anecdotally, I rarely see them when editing the map), but on the other hand it's relatively trivial to add support for them. (although there's a third factor in play here, which is the amount of time and effort it takes to find out about the tag variations, and their various distributions, which is where we failed Aaron).
The old maxim of "be conservative in what you produce, be liberal in what you accept" comes to mind here. And whilst there has always been an argument that being conservative in what you accept encourages (or even forces) people to be conservative in what they produce, this rarely wins out - mainly because consumers/renderers have more of an obligation to their users (the consumers) than to producers. We can see how this has played out in the browser wars, where all browsers now accept tag soup, and the focus in now on standardising the interpretation of the soup (HTML5) rather than enforcing a strict adherance to the standards (XHTML4). Of course, for OSM the maxim should probably be "be liberal in what you accept, be liberal in what you produce, but with a community consensus where possible". And so far, this seems to more or less work, right? We've clearly got a way to go in communicating the consensus, and in giving guidelines to data-users, as this thread shows, but that's just a matter of incremental improvements. Frankie PS I'm off to go mapping with oneway=ja and building=kittens... ;-) On 30/09/2009, Dave F. <dave...@madasafish.com> wrote: > Frederik Ramm wrote: >> Hi, >> >> Dave F. wrote: >>>>> Who, within OSM & in their right sense of mind, would object to >>>>> being forced to use just Yes/No. I mean it's just an on/off switch! >> >>>> Says who? >> >>> Err... I do. >>> Could you expand on why you might think otherwise? >> >> Which tag were you talking about that you thought was simply a yes/no >> option? >> >> > Hi > I wasn't talking about a specific tag just their values. When I said > "Who would object to being forced to use just Yes/No" I meant in > comparison with True/False, 1/0; /not /totally exclusive to Y/N. > For the example of building=, I am aware that mappers may also define > their own values, which I agree with. > > I disagree with Frankie that "...renderers should probably accept all > three forms though". > I don't understand why others should have to spend time sorting out > OSM's vagaries purely because we can't decide. > > Cheers > Dave F. > > > > _______________________________________________ > talk mailing list > talk@openstreetmap.org > http://lists.openstreetmap.org/listinfo/talk > -- Frankie Roberto Experience Designer, Rattle 0114 2706977 http://www.rattlecentral.com _______________________________________________ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
