Re: [Tagging] default value for oneway
On Fri, Aug 29, 2014 at 12:48 PM, ael law_ence@ntlworld.com wrote: To suggest that we now have to include every possible tag with an explicit value on every element is just ridiculous: the logical consequence of an explicit oneway on all ways. +1 The rule is and has always been in OSM the following : if oneway tag is missing, we assume it is not a oneway (excepted for roundabouts: and we had epic discussions about the default for motorways but finally it was said that we should add it explicitely also there). I can understand it is disturbing newcomers and people coming from GIS. But once you understand the concept, it's easy. Of course, adding explicitely that a street is not a oneway, especially in a dense area with many oneways make sens. It's just telling the other contributors that this particular attribut has been verified by someone. But don't ask others to make it mandatory. Pieren ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] default value for oneway
Am 28.08.2014 um 23:02 schrieb Xavier Noria: On Thu, Aug 28, 2014 at 10:39 PM, Peter Wendorff wendo...@uni-paderborn.de wrote: No, it isn't. The interpretation of the database, and the meaning, restricted to the fact of the streets oneway-ness is the same, but no value at all does not say this is no oneway street, it says nothing more than we don't know if it's oneway or not. That is the generic interpretation of a NULL value in programming (I am a programmer), absence of value. But your contract is that unset implies no for streets. So there you go. Got no value? I *have* to assume no. And since that's the case, the de facto usage pattern seems to be to leave oneway unset. The database has millions of NULLs for which the users mean an actual no. They didn't bother, but it is NULL for no. And that is a consequence of the design of the data model for that attribute. If this was 0-day of OSM and the attribute had possible values one-way, two-way, reversible, with an active default of two-way preselected in UIs, then you could in practice say NULL means unknown. +0.5, as UIs are decoupled from the data in OSM. You may write your own editor with a completely different UI, even one that doesn't know about oneway at all, so reasoning on UI preferences may help to get the best default, but not to derive rules from. regards Peter ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] default value for oneway
On Fri, Aug 29, 2014 at 9:43 AM, Peter Wendorff wendo...@uni-paderborn.de wrote: +0.5, as UIs are decoupled from the data in OSM. You may write your own editor with a completely different UI, even one that doesn't know about oneway at all, so reasoning on UI preferences may help to get the best default, but not to derive rules from. Agreed. That was a way to say: if you reset the values in the database to have no NULLs, and change the semantics of NULL to mean strictly unknown, not all is lost regarding convenience of data entry. iD (to put an example) can still choose to preselect no for streets. Nowadays it has a list of stuff for which yes is assumed: https://github.com/bhousel/iD/blob/master/js/id/core/oneway_tags.js And you increase certainty. If a UI does not include oneway and creates streets, then oneway remains unknown for them, fine. The user said nothing, so fine to write a NULL meaning don't know. Nevertheless, all this is a theoretical exercise just for the sake of thinking about the attribute. ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] default value for oneway
Am 29.08.2014 um 09:58 schrieb Xavier Noria: On Fri, Aug 29, 2014 at 9:43 AM, Peter Wendorff wendo...@uni-paderborn.de wrote: +0.5, as UIs are decoupled from the data in OSM. You may write your own editor with a completely different UI, even one that doesn't know about oneway at all, so reasoning on UI preferences may help to get the best default, but not to derive rules from. Agreed. That was a way to say: if you reset the values in the database to have no NULLs, and change the semantics of NULL to mean strictly unknown, not all is lost regarding convenience of data entry. iD (to put an example) can still choose to preselect no for streets. Nowadays it has a list of stuff for which yes is assumed: https://github.com/bhousel/iD/blob/master/js/id/core/oneway_tags.js And you increase certainty. I fear you don't. For the list you linked this is true as most ways are oneways, but in general it is dangerous. A user must be aware that he explicitly states this way is oneway (that is the case already due to the small arrows iD paints on the ways) or this way is no oneway (which is invisible, as it is the same visual appearance than no oneway tag at all). If you don't set the tag as a default, everything is fine as nobody accidentally sets it to a wrong value. If you set it to oneway=no by default, people will start to accidentally tag oneway=no, which results in wrong information in contrast to missing inforation we have no. Of course: if you could make people aware of the tag and make sure 99% of the ways are intentionally (!) tagged with yes or no, then it's fine, but I doubt it. regards Peter ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] default value for oneway
To suggest that we now have to include every possible tag with an explicit value on every element is just ridiculous: the logical consequence of an explicit oneway on all ways. Where there really is a need to remove ambiguity, surely something like an area or perhaps relation (less obvious to the casual mapper) is needed within which the default value(s) for a tag(s) is defined. Now I start worrying about new ways crossing the area, so maybe a relation is better after all. Just a suggestion which needs refinement if anyone here thinks this is sensible. But the current argument: shall we or not? isn't going anywhere and normal mappers just won't add explicit tags in normal circumstances. You need a different approach and maybe what I say above can start the ball rolling. ael ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
[Tagging] default value for oneway
Hi, The default value for oneway is no for most types of roads. That is, if the attribute has no value set, no is assumed. Which is the rationale for that default? In the European cities and towns I know the majority of streets are one-way. For example Barcelona, or Madrid, or Paris. In such areas, are contributors expected to annotate nos rather than leaving the tag blank? ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] default value for oneway
Hi Xavier, no is the default value of the oneway tag as it's the most correct assumption. First as in general most roads are not oneway roads (considering any road inside and outside of cities), and second as the other case around would be even worse: If yes, this is a oneway street would be assumed, you still wouldn't know which direction the traffic is allowed. You may add oneway=no to streets where you know they are not oneway, but you have to define the oneway tag to define the traffic direction, else there's still missing data on oneways. But as the base assumption of all applications I'm aware of is that streets is oneway=no, this is usually not necessary. If in some areas, where most highways are oneways, this leads to the default being less used than the non-default, this isn't bad either as the oneway tag is necessary to get complete information. regards Peter Am 28.08.2014 um 14:20 schrieb Xavier Noria: Hi, The default value for oneway is no for most types of roads. That is, if the attribute has no value set, no is assumed. Which is the rationale for that default? In the European cities and towns I know the majority of streets are one-way. For example Barcelona, or Madrid, or Paris. In such areas, are contributors expected to annotate nos rather than leaving the tag blank? ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] default value for oneway
Xavier Noria writes: In the European cities and towns I know the majority of streets are one-way. In not a single EU city I know of there is something close to a majority of streets being one-way. Even more. In most of the villages the roads are not one-way. Based on this it's a good rationale to assume a street is non one-way by default. This also applies to the cities in Asia i know of. Stephan ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] default value for oneway
I wish people in OSM would stop making things up, believing it makes their point of view stronger. On 28/08/2014 13:20, Xavier Noria wrote: In the European cities and towns I know the majority of streets are one-way. --- This email is free from viruses and malware because avast! Antivirus protection is active. http://www.avast.com ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] default value for oneway
On Thu, Aug 28, 2014 at 3:39 PM, Dave F. dave...@madasafish.com wrote: I wish people in OSM would stop making things up, believing it makes their point of view stronger. What? I am not assuming one-way would be a better default. Nor I am assuming anything about the world at large. What are you talking about? I only asked questions: 1) Which is the rationale? Not because I think it is wrong, but to understand it! 2) In cities and towns where two-way streets are exceptional like Barcelona or Madrid, are people expected to tag them no? The motivation for this question is that there seems to be the convention not to tag them, and therefore you cannot tell the confirmed ones from the untagged ones. ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] default value for oneway
Xavier Noria wrote, on 2014-08-28 15:45: 2) In cities and towns where two-way streets are exceptional like Barcelona or Madrid, are people expected to tag them no? The motivation for this question is that there seems to be the convention not to tag them, and therefore you cannot tell the confirmed ones from the untagged ones. As Peter pointed out, oneway=yes (or in the exeptional case of oneway against the vector of the OSM way, oneway=-1 ) is mandatory if the road is a oneway, for most categories. oneway=no can help other mappers to show it has been checked in areas of high oneway density. Like a concise note. Tom ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] default value for oneway
2014-08-28 14:45 GMT+01:00 Xavier Noria f...@hashref.com: On Thu, Aug 28, 2014 at 3:39 PM, Dave F. dave...@madasafish.com wrote: I wish people in OSM would stop making things up, believing it makes their point of view stronger. What? I am not assuming one-way would be a better default. Nor I am assuming anything about the world at large. What are you talking about? Let's put all that aside. I only asked questions: 1) Which is the rationale? Not because I think it is wrong, but to understand it! As Peter said, the default for services using OSM is always to assume a way is _not_ oneway unless tagged otherwise. 2) In cities and towns where two-way streets are exceptional like Barcelona or Madrid, are people expected to tag them no? The motivation for this question is that there seems to be the convention not to tag them, and therefore you cannot tell the confirmed ones from the untagged ones. No, I think more likely is that local mappers have not considered it a priority to add the oneway tags, or maybe there are so many that it's a difficult job that isn't finished yet. Best Dan ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] default value for oneway
On 2014-08-28 15:53, Dan S wrote: As Peter said, the default for services using OSM is always to assume a way is _not_ oneway unless tagged otherwise. Unless it is tagged as junction=roundabout ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] default value for oneway
On Thu, Aug 28, 2014 at 3:53 PM, Dan S danstowell+...@gmail.com wrote: 2) In cities and towns where two-way streets are exceptional like Barcelona or Madrid, are people expected to tag them no? The motivation for this question is that there seems to be the convention not to tag them, and therefore you cannot tell the confirmed ones from the untagged ones. No, I think more likely is that local mappers have not considered it a priority to add the oneway tags, or maybe there are so many that it's a difficult job that isn't finished yet. At least in the Spanish mailing list, with some exceptions people generally think leaving it blank is fine for any city, because it means no anyway, so why say anything, the argument goes. Of course, in Barcelona for example assuming no is impractical, since most streets are one-way. So in the case of Barcelona I don't think that is a good practice. And that seems to be also suggested in the wiki: oneway=no is used to confirm that (a part of) a street is NOT a oneway street. (Use only in order to avoid mapping errors in areas where e.g. oneway streets are common, or to override defaults.) ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] default value for oneway
For the sake of discussion, I believe the interface for setting this attribute could be different (I am a software developer). For example, in graphical interfaces like iD you could have no preselected as convenience. But if you send no, you are saying no. Otherwise, you could opt-out and leave the value as blank, that would mean unknown/unset. In APIs, the attribute would have no default. If you say no, it is no, if you send nothing, it is unset. That way you could distinguish nos from unsets. Right now you cannot because conventions promote saying nothing. I realize changing any of this may be impossible nowadays, and maybe you disagree with that proposal. But if there was a chance to revise this I know Ruby on Rails and could work on it. ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] default value for oneway
Right now, the oneway checkbox in iD cycles through “Yes” “No” and “Assumed to be No” (blank). There are a handful of situations that will switch this checkbox to say “Yes” “No” and “Assumed to be Yes” (blank). (for example, a `junction=roundabout` or `highway=motorway` tag) It sounds to me like iD is already working the way you describe. But definitely let us know if maybe there is an issue with translation or another way we could improve on it. I did work on this feature and am familiar with the code. Thanks, Bryan On Aug 28, 2014, at 10:32 AM, Xavier Noria f...@hashref.com wrote: For the sake of discussion, I believe the interface for setting this attribute could be different (I am a software developer). For example, in graphical interfaces like iD you could have no preselected as convenience. But if you send no, you are saying no. Otherwise, you could opt-out and leave the value as blank, that would mean unknown/unset. In APIs, the attribute would have no default. If you say no, it is no, if you send nothing, it is unset. That way you could distinguish nos from unsets. Right now you cannot because conventions promote saying nothing. I realize changing any of this may be impossible nowadays, and maybe you disagree with that proposal. But if there was a chance to revise this I know Ruby on Rails and could work on it. ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] default value for oneway
I believe that you haven't explicitly said so, but probably essentially want to be able to find streets that haven't been surveyed and potentially need a oneway tag and avoid false positives (aka such that are actually bi-directional). I don't believe you'll get any further with the oneway tag, but given that we have similar issues for example with name tags, you could consider something like proposed in http://wiki.openstreetmap.org/wiki/Proposed_features/Internal_quality , in your case validate:no_oneway Simon PS: the noname tag is actually substantially more popular than validate:no_name but if you are inventing something new, you might as well stick to the validate: scheme. Am 28.08.2014 16:32, schrieb Xavier Noria: For the sake of discussion, I believe the interface for setting this attribute could be different (I am a software developer). For example, in graphical interfaces like iD you could have no preselected as convenience. But if you send no, you are saying no. Otherwise, you could opt-out and leave the value as blank, that would mean unknown/unset. In APIs, the attribute would have no default. If you say no, it is no, if you send nothing, it is unset. That way you could distinguish nos from unsets. Right now you cannot because conventions promote saying nothing. I realize changing any of this may be impossible nowadays, and maybe you disagree with that proposal. But if there was a chance to revise this I know Ruby on Rails and could work on it. ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging signature.asc Description: OpenPGP digital signature ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] default value for oneway
On Thu, Aug 28, 2014 at 4:45 PM, Bryan Housel br...@7thposition.com wrote: Right now, the oneway checkbox in iD cycles through “Yes” “No” and “Assumed to be No” (blank). There are a handful of situations that will switch this checkbox to say “Yes” “No” and “Assumed to be Yes” (blank). (for example, a `junction=roundabout` or `highway=motorway` tag) It sounds to me like iD is already working the way you describe. But definitely let us know if maybe there is an issue with translation or another way we could improve on it. I did work on this feature and am familiar with the code. Awesome! Hi there! I think iD is coherent with the contract right now. Because the contract says that if you leave the value blank, that means no. So iD has the need for a three-valued check box to be able to say unset, since I suppose the database has plenty of NULLs. I was proposing a change in the way to interpret NULLs. If the value is unset, it is unset. The same way a name can be unset. Means unknown. Then the convenience default would be moved to the UI, no would be preselected. Equally helpful, but adds certainty to the data. Of course a change like this in meaning is major, it may be even impossible in practical terms, not familiar enough with the project ecosystem. ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] default value for oneway
On Thu, Aug 28, 2014 at 4:48 PM, Simon Poole si...@poole.ch wrote: I believe that you haven't explicitly said so, but probably essentially want to be able to find streets that haven't been surveyed and potentially need a oneway tag and avoid false positives (aka such that are actually bi-directional). You nailed it :). I am going out with my daughter to tag directions: https://twitter.com/fxn/status/503616557776646144 The closer I have been to distinguish that is this query: http://overpass-turbo.eu/s/4LR That makes me also wonder as a side-effect about the implication of the current contract and the usage patterns it promotes. Implications in particular for turn-by-turn indications, but that was secondary, my main motivation is the one above. I don't believe you'll get any further with the oneway tag, but given that we have similar issues for example with name tags, you could consider something like proposed in http://wiki.openstreetmap.org/wiki/Proposed_features/Internal_quality , in your case validate:no_oneway Simon PS: the noname tag is actually substantially more popular than validate:no_name but if you are inventing something new, you might as well stick to the validate: scheme. That's interesting, I'll take into account. ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] default value for oneway
Am 28.08.2014 17:07, schrieb Xavier Noria: ... That makes me also wonder as a side-effect about the implication of the current contract and the usage patterns it promotes. Implications in particular for turn-by-turn indications, but that was secondary, my main motivation is the one above. In any case there are roughly 45 million highway segments on which a oneway tag could make sense, vs. roughly 6 million oneway=yes and 1.5 million oneway=no. I suspect that it is really -far- too late to change the semantics of this specific attribute. Simon signature.asc Description: OpenPGP digital signature ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] default value for oneway
On Thu, Aug 28, 2014 at 5:21 PM, Simon Poole si...@poole.ch wrote: In any case there are roughly 45 million highway segments on which a oneway tag could make sense, vs. roughly 6 million oneway=yes and 1.5 million oneway=no. I suspect that it is really -far- too late to change the semantics of this specific attribute. Since nowadays NULL for a street means oneway=no a change in the semantics would be still be possible as far as the database is concerned. If you go today to the database and update all oneway attributes for streets which are blank to no, the meaning of the database is equivalent. Same for motorways, replace all NULLs with yes. Equivalent database. For a street, there is no practical difference nowadays between no and unset, which is a smell for me. Either way means no. I believe the default is useful for the UI, to preselect a value for example so that the user has to do nothing in the majority of street creations, less useful as a way to interpret NULLs because then you don't know what has been confirmed. ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] default value for oneway
For a street, there is no practical difference nowadays between no and unset, which is a smell for me. Either way means no. For the software? No, there isn't a difference. For the mapper? Yes, there is a difference. Since nowadays NULL for a street means oneway=no a change in the semantics would be still be possible as far as the database is concerned. If you go today to the database and update all oneway attributes for streets which are blank to no, the meaning of the database is equivalent. Theorically speaking, yes, you could add oneway=no to every street, and get a functionally equivalent database (from the software's POV). But, in practice, people most likely wouldn't agree with that (this change would be reverted). 2014-08-28 12:33 GMT-03:00 Xavier Noria f...@hashref.com: On Thu, Aug 28, 2014 at 5:21 PM, Simon Poole si...@poole.ch wrote: In any case there are roughly 45 million highway segments on which a oneway tag could make sense, vs. roughly 6 million oneway=yes and 1.5 million oneway=no. I suspect that it is really -far- too late to change the semantics of this specific attribute. Since nowadays NULL for a street means oneway=no a change in the semantics would be still be possible as far as the database is concerned. If you go today to the database and update all oneway attributes for streets which are blank to no, the meaning of the database is equivalent. Same for motorways, replace all NULLs with yes. Equivalent database. For a street, there is no practical difference nowadays between no and unset, which is a smell for me. Either way means no. I believe the default is useful for the UI, to preselect a value for example so that the user has to do nothing in the majority of street creations, less useful as a way to interpret NULLs because then you don't know what has been confirmed. ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] default value for oneway
On Thu, Aug 28, 2014 at 6:52 PM, John Packer john.pack...@gmail.com wrote: For a street, there is no practical difference nowadays between no and unset, which is a smell for me. Either way means no. For the software? No, there isn't a difference. For the mapper? Yes, there is a difference. The mapper can't tell unset from no either today. If unset mean unset (as in the hypothetical proposal I am outlining), then it would mean something: I am gonna volunteer this value, this value is unknown, etc. Problem is people at large are using unset to mean no. So nowadays you can't tell. Since nowadays NULL for a street means oneway=no a change in the semantics would be still be possible as far as the database is concerned. If you go today to the database and update all oneway attributes for streets which are blank to no, the meaning of the database is equivalent. Theorically speaking, yes, you could add oneway=no to every street, and get a functionally equivalent database (from the software's POV). But, in practice, people most likely wouldn't agree with that (this change would be reverted). Wouldn't agree with the equivalence of the database, or with the patch itself issuing a SQL statement? Such a patch wouldn't make any sense unless it was part of a bigger plan. I explained that only as a way to say as far as the database is concerned the data is equivalent. But for example, every single client software of OSM that is out of control of OSM is assuming that contract. That's what I believe makes a reset (no NULLs in the database) plus semantic change for NULLs would not be possible. No way to synchronize all that client code unless that was somehow coordinated as in a major upgrade. ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] default value for oneway
On Thu, 2014-08-28 at 13:52 -0300, John Packer wrote: For a street, there is no practical difference nowadays between no and unset, which is a smell for me. Either way means no. For the software? No, there isn't a difference. For the mapper? Yes, there is a difference. Since nowadays NULL for a street means oneway=no a change in the semantics would be still be possible as far as the database is concerned. If you go today to the database and update all oneway attributes for streets which are blank to no, the meaning of the database is equivalent. Theorically speaking, yes, you could add oneway=no to every street, and get a functionally equivalent database (from the software's POV). But, in practice, people most likely wouldn't agree with that (this change would be reverted). +1 To use add oneway=no in selected areas to confirm the road has been surveyed is fine, but not everywhere as that causes tag clutter and makes it difficult for a mapper to see the important tags. Phil (trigpoint) 2014-08-28 12:33 GMT-03:00 Xavier Noria f...@hashref.com: On Thu, Aug 28, 2014 at 5:21 PM, Simon Poole si...@poole.ch wrote: In any case there are roughly 45 million highway segments on which a oneway tag could make sense, vs. roughly 6 million oneway=yes and 1.5 million oneway=no. I suspect that it is really -far- too late to change the semantics of this specific attribute. Since nowadays NULL for a street means oneway=no a change in the semantics would be still be possible as far as the database is concerned. If you go today to the database and update all oneway attributes for streets which are blank to no, the meaning of the database is equivalent. Same for motorways, replace all NULLs with yes. Equivalent database. For a street, there is no practical difference nowadays between no and unset, which is a smell for me. Either way means no. I believe the default is useful for the UI, to preselect a value for example so that the user has to do nothing in the majority of street creations, less useful as a way to interpret NULLs because then you don't know what has been confirmed. ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] default value for oneway
On Thu, Aug 28, 2014 at 7:13 PM, Philip Barnes p...@trigpoint.me.uk wrote: +1 To use add oneway=no in selected areas to confirm the road has been surveyed is fine, but not everywhere as that causes tag clutter and makes it difficult for a mapper to see the important tags. Which tools does a hard-core mapper use to edit? ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] default value for oneway
Am 28.08.2014 19:10, schrieb Xavier Noria: ... But for example, every single client software of OSM that is out of control of OSM is assuming that contract. That's what I believe makes a reset (no NULLs in the database) plus semantic change for NULLs would not be possible. No way to synchronize all that client code unless that was somehow coordinated as in a major upgrade. While just a technicality it probably is worth pointing out that in reality the change you propose would require adding 40 million or so oneway=no tags to the DB and making a new version of each of the effected objects. We have worse things in the DB, but still it would need more than just uneasiness with the current situation (which is for largish parts of the world not an issue) to justify the change Simon signature.asc Description: OpenPGP digital signature ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] default value for oneway
On Thu, 2014-08-28 at 19:16 +0200, Xavier Noria wrote: On Thu, Aug 28, 2014 at 7:13 PM, Philip Barnes p...@trigpoint.me.uk wrote: +1 To use add oneway=no in selected areas to confirm the road has been surveyed is fine, but not everywhere as that causes tag clutter and makes it difficult for a mapper to see the important tags. Which tools does a hard-core mapper use to edit? In my case I mainly use josm. Phil (trigpoint) ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] default value for oneway
Since I see the characteristics of Barcelona (and other cities/towns I know) are exceptional for most of you guys, let me share a couple of maps to explain where I am coming from. This is a typical sector of Barcelona: https://www.dropbox.com/s/k7o32zbneoi8y6q/barcelona_sample.png?dl=0 As you see, almost everything is one-way. Of course there are some important streets here and there that are two-way, but the ratio is really small. So for example, in this map from OSM that has streets with no oneway set highlighted in blue: https://www.dropbox.com/s/yjosl8lwyk3xkme/barcelona_bad_sector.png?dl=0 that area in the center with many blue lines... almost all of them are wrong. You cannot rely on that default in Barcelona at all. ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] default value for oneway
2014-08-28 22:31 GMT+02:00 Xavier Noria f...@hashref.com: that area in the center with many blue lines... almost all of them are wrong. You cannot rely on that default in Barcelona at all. And in this really rare situation it is reasonable to use oneway=no. ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] default value for oneway
Am 28.08.2014 um 19:10 schrieb Xavier Noria: On Thu, Aug 28, 2014 at 6:52 PM, John Packer john.pack...@gmail.com wrote: For a street, there is no practical difference nowadays between no and unset, which is a smell for me. Either way means no. For the software? No, there isn't a difference. For the mapper? Yes, there is a difference. The mapper can't tell unset from no either today. If unset mean unset (as in the hypothetical proposal I am outlining), then it would mean something: I am gonna volunteer this value, this value is unknown, etc. Problem is people at large are using unset to mean no. So nowadays you can't tell. Since nowadays NULL for a street means oneway=no a change in the semantics would be still be possible as far as the database is concerned. If you go today to the database and update all oneway attributes for streets which are blank to no, the meaning of the database is equivalent. No, it isn't. The interpretation of the database, and the meaning, restricted to the fact of the streets oneway-ness is the same, but no value at all does not say this is no oneway street, it says nothing more than we don't know if it's oneway or not. If somewhere oneway=no is tagged, it says (or should say): Someone verified this street to be oneway. For software there's a problem for any unknown value, so some assumption has to be made. Assuming oneway=no is the best effort assumption as it causes least harm (allows the right direction correctly, allows the wrong correction incorrectly for untagged oneways; and works for untagged non-oneways). Adding oneway=no to any highway that is untagged currently removes the information of the highway being verified, and adds no value with the given assumption. Adding oneway=no by hand on a case by case basis where you know a highway is not a oneway may be a solution; but I don't think many people would do that. Your UI proposals for a default value no to be actively unchecked by the user is dangerous as often people keep the defaults without knowledge or thinking about it, so better keep unknown/unset as the default - just as we have it right now. Theorically speaking, yes, you could add oneway=no to every street, and get a functionally equivalent database (from the software's POV). But, in practice, people most likely wouldn't agree with that (this change would be reverted). Wouldn't agree with the equivalence of the database, or with the patch itself issuing a SQL statement? Such a patch wouldn't make any sense unless it was part of a bigger plan. I explained that only as a way to say as far as the database is concerned the data is equivalent. -1, as explained above. regards Peter ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] default value for oneway
Am 28.08.2014 um 22:35 schrieb Mateusz Konieczny: 2014-08-28 22:31 GMT+02:00 Xavier Noria f...@hashref.com: that area in the center with many blue lines... almost all of them are wrong. You cannot rely on that default in Barcelona at all. And in this really rare situation it is reasonable to use oneway=no. Well, yes - but only while adding oneway=yes where it needs to be set, so even here oneway=no is more like a way checked., adding oneway=no does not help very much, as long as oneway=yes is that incomplete. regards Peter ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] default value for oneway
On Thu, Aug 28, 2014 at 10:49 PM, Peter Wendorff wendo...@uni-paderborn.de wrote: Am 28.08.2014 um 22:35 schrieb Mateusz Konieczny: 2014-08-28 22:31 GMT+02:00 Xavier Noria f...@hashref.com: that area in the center with many blue lines... almost all of them are wrong. You cannot rely on that default in Barcelona at all. And in this really rare situation it is reasonable to use oneway=no. Well, yes - but only while adding oneway=yes where it needs to be set, so even here oneway=no is more like a way checked., adding oneway=no does not help very much, as long as oneway=yes is that incomplete. Exactly. I interpreted Mateusz meant in Barcelona it is reasonable to set explicit nos rather than leaving defaults. Basically, the default in Barcelona is useless. Here you should tag every street. yes because it is needed in the majority of streets, and no because you cannot trust the default as a consequence. Which is fine of course, the default has to be the best one given the world-wide scope of OSM. ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] default value for oneway
On Thu, Aug 28, 2014 at 10:39 PM, Peter Wendorff wendo...@uni-paderborn.de wrote: No, it isn't. The interpretation of the database, and the meaning, restricted to the fact of the streets oneway-ness is the same, but no value at all does not say this is no oneway street, it says nothing more than we don't know if it's oneway or not. That is the generic interpretation of a NULL value in programming (I am a programmer), absence of value. But your contract is that unset implies no for streets. So there you go. Got no value? I *have* to assume no. And since that's the case, the de facto usage pattern seems to be to leave oneway unset. The database has millions of NULLs for which the users mean an actual no. They didn't bother, but it is NULL for no. And that is a consequence of the design of the data model for that attribute. If this was 0-day of OSM and the attribute had possible values one-way, two-way, reversible, with an active default of two-way preselected in UIs, then you could in practice say NULL means unknown. ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
