Re: [Tagging] Everybody is hiding?
Hi Ole, Am Donnerstag, 9. August 2012, 17:55:24 schrieb Ole Nielsen / osm: > First of all I actually approved the proposal but later realized that > having variable keys is less than ideal. then *please* tell me the reason why you believe this is the case, because I haven't seen any compelling counter-argument so far. What I have seen from different people: - allows for an almost infinite number of keys: existing tagging shows that keys tend to cluster, e.g. maxspeed:(22:00-06:00) is in use 395 (!) times with 6 different values (putting this into perspective: meagre 4494 occurences of maxspeed:backward). Those clustering effects become even stronger with increased usage. - kills PostgreSQL database performance: when you preprocess your routing data, you have to do a linear scan over all tag hstores anyway. - difficult because of special chars: the only situation where this actually matters is when you search inside your editor – and in that case the ':' already requires you to quote your key, at least in JOSM - difficult to parse for computers: every programmer can tell you in a second that this is plain wrong - difficult to parse for humans: so far, everybody I talked to about this was able to grasp the meaning of maxspeed:(22:00-06:00) = 100 in a split second And – of course – my favourite: - un-OSM-y, don't like it Eckhart ___ Tagging mailing list Tagging@openstreetmap.org http://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] Everybody is hiding?
Ole Nielsen / osm writes: > > http://wiki.openstreetmap.org/wiki/Proposed_features/Conditional_restrictions > > A short comment on the proposal: The actual conditions go into the tag > value. The transport mode (vehicle catagory) and the direction stay in the > key in accordance with current practice for access restrictions. > Ole / polderrunner > Good work there, a very good blend of expanding the key with already used information *:hgv:conditional=* and at the same time keeping more complex information in the values *=no:(12:00-18:00) It is good that the expansion of the key is only with things that are quite defined already: Maybe more could be added if they get commonly used. I am not sure if *:condition=* really needs to be added, but it has probably something to do with how the machines interprets keys. ___ Tagging mailing list Tagging@openstreetmap.org http://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] Everybody is hiding?
> Hi tagging list, > > the Extended Conditions proposal has been shot down by a majority, and > therefore there is still no "official" way of tagging quite a lot of > things. (As a side note, the Extended Conditions proposal is still the de > facto standard.) > > Therefore, I expected that those people who had voted against the proposal > came up with a well-designed alternative proposal â yet nothing > happened. Shall I conclude that all those people who voted against the > proposal did this just for the sake of voting against? First of all I actually approved the proposal but later realized that having variable keys is less than ideal. I am currently working on an alternative proposal and I was planning to announce it within a few days (I have only limited internet access the next couple of days). But here it is. http://wiki.openstreetmap.org/wiki/Proposed_features/Conditional_restrictions A short comment on the proposal: The actual conditions go into the tag value. The transport mode (vehicle catagory) and the direction stay in the key in accordance with current practice for access restrictions. Feel free to comment on it, preferably on the talk page. Ole / polderrunner ___ Tagging mailing list Tagging@openstreetmap.org http://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] Everybody is hiding?
Hi Frederik, Am Donnerstag, 9. August 2012, 14:36:40 schrieb Frederik Ramm: > You have to work on your expectations then. Has it occurred to you that > some people don't find extended conditions important enough at all? > > Personally, I think that most of the extended conditions that the > proposal tried to address were not worth having a tagging scheme for; > they were stuff that only a few perfectionists would want to map anyway. Yeah, e.g. those quixotic perfectionist geeks from Synyx. > And while I am not against perfectionists mapping stuff they like, I am > against elevating this to the state of an "accepted proposal" because > that would convey too much mindshare to such a marginal issue. In constrast to *really* important features like diet meals, clocks or fire hydrants… > The proposal is driven by a geek-y desire to convert every last bit of > information contained in a road sign into an OSM tag. But I don't think Okay, I repeat it one more time for you: this is not about some stuff geeks want to add to the database, this is serious stuff that some companies actually want to use (and other companies like MapQuest and Tele Atlas sell this kind of information). If you don't believe me then just have a look at GDF, which is an industrial standard that specifies exactly the same "geek-y" stuff (IIRC you can find some older versions of the standard on the internet). > that this is what people will usually want to do, and I fear that giving > this idea more mindshare will in the end lead to our editors being > burdened by special restriction composer preset tabs where you can > generate stuff like time and weather dependent speed limits for disabled > persons with children. Yeah, like the UI-cluttering turn restrictions plugin in JOSM… wait, what? Yes, it is a *plugin*. If you do not like it, just do not download it. > I don't think that the proposal is the "de facto standard" either. I > think some of its parts will probably be used - e.g. I could see > "maxspeed:wet" being of use. I think it is likely however that this will > be interpreted like a normal, fixed tag, […] I cannot find any wiki entry for - maxspeed:wet - maxspeed:hgv:forward - maxspeed:motorcycle - toll:hgv - toll:forward - access:hgv:forward (just to pick a few). If those are all fixed tags, then where are the wiki entries for them? On the other hand, the Extended Conditions proposal explains *all* of them, just in one page instead of thousand pages. > […] and I don't believe anyone will > actually implement a restriction parser that understands any combination > of restrictions on any tags. It's not that difficult to implement, trust me. > I have no problem whatsoever if the mapping of speed limits that only > apply to HGV at night happens by way of a "note" tag. It's just not > frequent enough to even discuss. For something that's not worth discussing, the discussion is quite lengthy. About that note tag proposal of yours: this is the most stupid proposal I have heard so far. I have a better one: why not stuff everything we ever want to tag into one big note tag, that would make all editors a *lot* simpler. (On the other hand, it would make using the data impossible, but as you already stated, the mapper is the only person that is important.) Eckhart ___ Tagging mailing list Tagging@openstreetmap.org http://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] Everybody is hiding?
Am 09.08.2012 um 14:36 schrieb Frederik Ramm : > I have no problem whatsoever if the mapping of speed limits that only apply > to HGV at night happens by way of a "note" tag. It's just not frequent enough > to even discuss. Actually time dependent speed limits, (e.g. to reduce noise at night time) or access conditions (e.g. for motorbikes) are not so rare, but as long as there is no agreed syntax (and no data consumer using this info) most mappers (myself included) will happily ignore these details. Cheers, Martin ___ Tagging mailing list Tagging@openstreetmap.org http://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] Everybody is hiding?
While I agree with Frederik almost completely, the absence of a tagging scheme for conditions will have some unwanted side effects. A current example in my opinion would be this great and completely intuitive n2/n3 tagging that was just invented. It was already documented in the german access article, giving the impression that this is an accepted tag. If we would have a tagging scheme for conditions this could have been prevented. Martin 2012/8/9 Frederik Ramm : > Hi, > > > On 08/09/2012 01:41 PM, Eckhart Wörner wrote: >> >> Therefore, I expected that those people who had voted against the >> proposal came up with a well-designed alternative proposal > > > You have to work on your expectations then. Has it occurred to you that some > people don't find extended conditions important enough at all? > > Personally, I think that most of the extended conditions that the proposal > tried to address were not worth having a tagging scheme for; they were stuff > that only a few perfectionists would want to map anyway. And while I am not > against perfectionists mapping stuff they like, I am against elevating this > to the state of an "accepted proposal" because that would convey too much > mindshare to such a marginal issue. > > The proposal is driven by a geek-y desire to convert every last bit of > information contained in a road sign into an OSM tag. But I don't think that > this is what people will usually want to do, and I fear that giving this > idea more mindshare will in the end lead to our editors being burdened by > special restriction composer preset tabs where you can generate stuff like > time and weather dependent speed limits for disabled persons with children. > > I don't think that the proposal is the "de facto standard" either. I think > some of its parts will probably be used - e.g. I could see "maxspeed:wet" > being of use. I think it is likely however that this will be interpreted > like a normal, fixed tag, and I don't believe anyone will actually implement > a restriction parser that understands any combination of restrictions on any > tags. > > I have no problem whatsoever if the mapping of speed limits that only apply > to HGV at night happens by way of a "note" tag. It's just not frequent > enough to even discuss. > > Bye > Frederik > > -- > Frederik Ramm ## eMail frede...@remote.org ## N49°00'09" E008°23'33" > > > ___ > Tagging mailing list > Tagging@openstreetmap.org > http://lists.openstreetmap.org/listinfo/tagging ___ Tagging mailing list Tagging@openstreetmap.org http://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] Everybody is hiding?
Hi, On 08/09/2012 01:41 PM, Eckhart Wörner wrote: Therefore, I expected that those people who had voted against the proposal came up with a well-designed alternative proposal You have to work on your expectations then. Has it occurred to you that some people don't find extended conditions important enough at all? Personally, I think that most of the extended conditions that the proposal tried to address were not worth having a tagging scheme for; they were stuff that only a few perfectionists would want to map anyway. And while I am not against perfectionists mapping stuff they like, I am against elevating this to the state of an "accepted proposal" because that would convey too much mindshare to such a marginal issue. The proposal is driven by a geek-y desire to convert every last bit of information contained in a road sign into an OSM tag. But I don't think that this is what people will usually want to do, and I fear that giving this idea more mindshare will in the end lead to our editors being burdened by special restriction composer preset tabs where you can generate stuff like time and weather dependent speed limits for disabled persons with children. I don't think that the proposal is the "de facto standard" either. I think some of its parts will probably be used - e.g. I could see "maxspeed:wet" being of use. I think it is likely however that this will be interpreted like a normal, fixed tag, and I don't believe anyone will actually implement a restriction parser that understands any combination of restrictions on any tags. I have no problem whatsoever if the mapping of speed limits that only apply to HGV at night happens by way of a "note" tag. It's just not frequent enough to even discuss. Bye Frederik -- Frederik Ramm ## eMail frede...@remote.org ## N49°00'09" E008°23'33" ___ Tagging mailing list Tagging@openstreetmap.org http://lists.openstreetmap.org/listinfo/tagging
[Tagging] Everybody is hiding?
Hi tagging list, the Extended Conditions proposal has been shot down by a majority, and therefore there is still no "official" way of tagging quite a lot of things. (As a side note, the Extended Conditions proposal is still the de facto standard.) Therefore, I expected that those people who had voted against the proposal came up with a well-designed alternative proposal – yet nothing happened. Shall I conclude that all those people who voted against the proposal did this just for the sake of voting against? Eckhart ___ Tagging mailing list Tagging@openstreetmap.org http://lists.openstreetmap.org/listinfo/tagging