[OSM-talk] Edit review: intermittent waters
Hi, I would like to support two ideas that appeared in this thread: 1. Peter Wendorff's about documenting discouraged/old tagging schemes in wiki as such - with link to correct schemes. 2. Worst Fixers general idea of merging multiple tags describing the same thing into one (possibly globally). That is if they really mean the same. Lukáš Matějka (LM_1) 2012/5/31 Russ Nelson nel...@crynwr.com: Worst Fixer writes: Persuade people to map just one way, THEN once they're doing that, go back and get rid of the old way. Sane people use type= for relation types. They use water= tag to express whether it is lake, pond, river or stream. Not how often it flows. You seem not to understand. Perhaps German is not your first language? Nobody is talking about the sanity or lack of sanity of editors except perhaps you. I'm talking about how people *actually* map. I think it's great that you're starting up a conversation on how we should interpret data not documented in the wiki. I'm NOT sure that we want to be *changing* data not documented in the wiki. Not sure at all. In fact, I'm pretty sure that we *shouldn't* be changing it. Sure that *you* shouldn't be changing it. Y'see, once you've made that change, one and only one interpretation of this undocumented data is available to everyone -- YOUR interpretation. You might be right, you might be wrong, but YOUR voice will prevail. Whereas, if we documented this data, and said Don't map like this -- map like that, then we accomplish two goals: 1) we let data users know what is the standard interpretation of this data, and 2) we encourage editors to stop editing like this. You know ... without calling them insane. So yeah, you should stop making these edits, and if you won't stop, I support taking action to stop you. -- --my blog is at http://blog.russnelson.com Crynwr supports open source software 521 Pleasant Valley Rd. | +1 315-600-8815 Potsdam, NY 13676-3213 | Sheepdog ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Edit review: intermittent waters
Am 30.05.2012 02:40, schrieb Russ Nelson:
Frederik Ramm writes:
2. I don't think you should continue to make mass edits under the
username WorstFixer because that implies that before you fixed
things they were among the worst which has the potential to offend
people.
I always thought it meant that he was the worst person to be doing
this fixing, or doing the worst job at fixing things.
The problem with his worst fixing, is that unless he starts {bugging /
bothering / teaching / correcting / annoying / discouraging} mappers
who are making these edits he disagrees with, they're going to keep
making these edits.
I'd prefer to see a consensus among editors that we don't do that
anymore, with the old way not being in the OSM wiki anywhere,
-1
It should be in the wiki anywhere, but absolutely clearly marked as
this is how NOT to do it, together with a reference to the new or
better scheme.
Often trivial ways to tag something are the fields where problems are
visible later with some practice.
When tagging something I never mapped before I often search for the
keywords which come in mind, and often these are the trivial keys
being obsolete by a new tagging scheme.
If there's nothing (as you suggest) and I search for alternatives,
everything is fine.
If there's nothing and I give up, tagging my own idea, it's more or less
the old variant again.
Therefore I would not delete the old variants, but clearly mark them, so
that everyone can find it
1) to read about past tagging practice and
2) as a link anchor to the new tagging style.
regards
Peter
___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Edit review: intermittent waters
Because if people are still editing like this, then I don't see the problem that WorstFixer is fixing as actually being fixed. Users of the map will still have to deal with both the format that WorstFixer doesn't like AND the format he does like. Less of the former and more of the latter, but still some of both. You seem to not understand. 93% of this tags come from imports. Ich think 7% comes from import, but edited by hand after, so current database not holds imporer user name. Persuade people to map just one way, THEN once they're doing that, go back and get rid of the old way. Sane people use type= for relation types. They use water= tag to express whether it is lake, pond, river or stream. Not how often it flows. That is agreed on wiki. Other ways of tagging not documented on wiki at all. -- WorstFixer, twitter: http://twitter.com/WorstFixer ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Edit review: intermittent waters
Hello Paul, You use HTML letters in mailing list. Using blue color shows your power. 2012/5/30 Paul Norman penor...@mac.com: Could you post the code used to generate the changesets? Ich may not. Some parts of it are not open source. I think of clean rewrite. That would be the easiest way for some of us to review your proposed changes. It wouldn’t remove the need to explain it to non-programmers, but it would be much easier for some of us if we could look at the code. Why bother posting code if it means Ich need to write long posts to people who can not analyze more than their neighbour hood, but send their mail fastly? - You provide no information on why you are removing the keys that you are Some people asked for another column, edit reason. That is hard but do able. Ich will create database of bad tags with comments. - A couple of other people have expressed concern over the message sent by having a mechanical edit from the WorstFixer username. I share them. It is not enough to dismiss this as an objection that is not “valid” Ok. I will do the same way as DWG does. Ich will create separate account WorstFixer_repair to distinguish it from my main account. - A significant number of these ways appear to be from US NHD data. You should also consult specifically with the US community and develop a consensus there that the edit is worth doing, in addition to the global community. US ist on the globe. They forget that from time to time, but must remember. Consulting global community is enough, according to current policy. - The mapping from NHD FCode to OSM tags used for some of these imports may of not been ideal. I have been working on a better one but have not finished. I ask you to finish it and show. If you are not able to, Ich will have to. I believe it would be best to exclude the US from this edit and later on (post-rebuild likely) propose an edit which includes changing tagging on untouched objects. This edit consists mostly of US import clean up. You know that. You just lazy to fix it. Three years already. Post-rebuild ist way of telling fuck off forever. Some believe in Jesus second coming, some believe in rebuild. I not share both belifs. Both not have any effect on what Ich doing. -- WorstFixer, twitter: http://twitter.com/WorstFixer ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Edit review: intermittent waters
From: Worst Fixer [mailto:worstfi...@gmail.com] Subject: Re: [OSM-talk] Edit review: intermittent waters Hello Paul, You use HTML letters in mailing list. Using blue color shows your power. My client uses HTML to reply to HTML messages. Since your message was HTML, I followed up in HTML. 2012/5/30 Paul Norman penor...@mac.com: Could you post the code used to generate the changesets? Ich may not. Some parts of it are not open source. I think of clean rewrite. The logic that you have created is what I'm interested in. That would be the easiest way for some of us to review your proposed changes. It wouldn’t remove the need to explain it to non-programmers, but it would be much easier for some of us if we could look at the code. Why bother posting code if it means Ich need to write long posts to people who can not analyze more than their neighbour hood, but send their mail fastly? The onus is on you to communicate the reasons for your edit. You are proposing large mechanical edits touching a wide variety of tags. The consultation is therefore likely to be complicated, to go along with the complicated edit. - You provide no information on why you are removing the keys that you are Some people asked for another column, edit reason. That is hard but do able. Ich will create database of bad tags with comments. To review your proposed edit you really need to explain why you are making each change. If we don't know why you're making a change, how can it be reviewed? - A couple of other people have expressed concern over the message sent by having a mechanical edit from the WorstFixer username. I share them. It is not enough to dismiss this as an objection that is not “valid” Ok. I will do the same way as DWG does. Ich will create separate account WorstFixer_repair to distinguish it from my main account. The concern was with the WorstFixer part of the name. If WorstFixer is your main account then I have concerns about such an inexperienced user making mechanical edits. - A significant number of these ways appear to be from US NHD data. You should also consult specifically with the US community and develop a consensus there that the edit is worth doing, in addition to the global community. US ist on the globe. They forget that from time to time, but must remember. Consulting global community is enough, according to current policy. The consultation in the policies is a bare minimum. It may come out in consultation that an edit requires additional consultation with other groups. This edit, as you say, consists mainly of US import clean-up. It is not unreasonable to expect you to then consult with talk-us. - The mapping from NHD FCode to OSM tags used for some of these imports may of not been ideal. I have been working on a better one but have not finished. I ask you to finish it and show. If you are not able to, Ich will have to. No - you do not have to. Another option is to wait. The current logic is at https://github.com/pnorman/ogr2osm-translations/blob/us_nhd/us_nhd.py but this has not yet been reviewed by the US community. I believe it would be best to exclude the US from this edit and later on (post-rebuild likely) propose an edit which includes changing tagging on untouched objects. This edit consists mostly of US import clean up. You know that. You just lazy to fix it. Three years already. Post-rebuild ist way of telling fuck off forever. Some believe in Jesus second coming, some believe in rebuild. I not share both belifs. Both not have any effect on what Ich doing. My work on the NHD translations continues. I foresee it finishing either during the rebuild or after it, based on current progress. As there are to be no mechanical edits when the rebuild is going on, that places it as after. If I finish before the rebuild then I'd propose something then, but it will take some time for anything I propose to be reviewed since it would be fairly extensive. ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Edit review: intermittent waters
Worst Fixer writes: Persuade people to map just one way, THEN once they're doing that, go back and get rid of the old way. Sane people use type= for relation types. They use water= tag to express whether it is lake, pond, river or stream. Not how often it flows. You seem not to understand. Perhaps German is not your first language? Nobody is talking about the sanity or lack of sanity of editors except perhaps you. I'm talking about how people *actually* map. I think it's great that you're starting up a conversation on how we should interpret data not documented in the wiki. I'm NOT sure that we want to be *changing* data not documented in the wiki. Not sure at all. In fact, I'm pretty sure that we *shouldn't* be changing it. Sure that *you* shouldn't be changing it. Y'see, once you've made that change, one and only one interpretation of this undocumented data is available to everyone -- YOUR interpretation. You might be right, you might be wrong, but YOUR voice will prevail. Whereas, if we documented this data, and said Don't map like this -- map like that, then we accomplish two goals: 1) we let data users know what is the standard interpretation of this data, and 2) we encourage editors to stop editing like this. You know ... without calling them insane. So yeah, you should stop making these edits, and if you won't stop, I support taking action to stop you. -- --my blog is athttp://blog.russnelson.com Crynwr supports open source software 521 Pleasant Valley Rd. | +1 315-600-8815 Potsdam, NY 13676-3213 | Sheepdog ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Edit review: intermittent waters
On 29 May 2012, at 12:51, Worst Fixer wrote: Hello. I ask you to review my planned edit. There are lot of ways to tag intermittent water feature found in database. Most popular is intermittent=yes. All others come from different old imports. Date ist 2009-2010 year. I countiered ~350 000 features tagged in different such ways. Most is done by 10 users. I ask users iandees and SK53 join discussion, as most of such tags were imported by them. Others welcome too. I propose unification of tagging in all this imports. Following tags converted to intermittent=yes: frequency=intermittent occurrence=intermittent stream=intermittent water=intermittent type=intermittent Following tags converted to intermittent=no: frequency=perennial stream=perennial stream=ephemeral converted to intermittent=ephemeral. Just removed stream=fixme. Converted fdate field from NHD imports in iso8601 date, moved to check_date tag. Removed all id-like tags. If no valid objections will be raised, I upload this change on 2012-06-12. Here is overview: Short, to get the idea: http://worstfixer.000a.biz/04-intermittent/overview-short.html Long, for exact analysis: http://worstfixer.000a.biz/04-intermittent/overview-full.html.gz I currently looking for place to upload exact .osm.gz for a preview. Suggestions welcome. Hi WorstFixer, I think this one might need a little more thought – what happens to something previously tagged water=intermittent... it becomes intermittent=yes... intermittent what? I doubt there's a nice way of predicting what water= should become to make it correctly tagged. Bob___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Edit review: intermittent waters
Hi, On 05/29/12 13:51, Worst Fixer wrote: There are lot of ways to tag intermittent water feature found in database. Most popular is intermittent=yes. All others come from different old imports. Date ist 2009-2010 year. I countiered ~350 000 features tagged in different such ways. Most is done by 10 users. 1. Is there any benefit? Does anybody use that tag at all, or is it just you feverishly looking for things you could fix? 2. I don't think you should continue to make mass edits under the username WorstFixer because that implies that before you fixed things they were among the worst which has the potential to offend people. I currently looking for place to upload exact .osm.gz for a preview. You could get an account on the dev server and put things there. Bye Frederik -- Frederik Ramm ## eMail frede...@remote.org ## N49°00'09 E008°23'33 ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Edit review: intermittent waters
2. I don't think you should continue to make mass edits under the username WorstFixer because that implies that before you fixed things they were among the worst which has the potential to offend people. I've not been following what has and hasn't been done, or is proposed, but the username WorstFixer looked to me like someone had changed their name after having had it pointed out that after their fixes things were worse than before, and that anyone else could have fixed things in a better way. So yes, not a good username choice. Ed ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Edit review: intermittent waters
Worst Fixer wrote: I ask you to review my planned edit. There are lot of ways to tag intermittent water feature found in database. Most popular is intermittent=yes. All others come from different old imports. Date ist 2009-2010 year. I countiered ~350 000 features tagged in different such ways. Most is done by 10 users. Following tags converted to intermittent=yes: frequency=intermittent occurrence=intermittent stream=intermittent water=intermittent type=intermittent Following tags converted to intermittent=no: frequency=perennial stream=perennial While that doesn't sound unreasonable... Short, to get the idea: http://worstfixer.000a.biz/04-intermittent/overview-short.html ..this page actually lists many more tags that you propose to change. You're going to remove 9413 ele tags; why? No doubt at some point someone decided that lake:shore_length:miles=2 was a useful thing to record, and you want to remove it. Why? I have a suggestion: break your proposed edits into smaller chunks. For example, in this case, propose an edit that *only* makes the 'intermittent' change that I've quoted above. Then propose separate edits that make each of the other changes. Regards, Phil. ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Edit review: intermittent waters
2012/5/29 Phil Endecott spam_from_osm_t...@chezphil.org: You're going to remove 9413 ele tags; why? No doubt at some point someone decided that lake:shore_length:miles=2 was a useful thing to record, and you want to remove it. Why? because there is no such thing as a shore length, it depends on the resolution. Aside from this there is also no tradition in OSM to record units in the key. So this particular key really doesn't seem to make any sense (the length of the shore in OSM-precision is already in the data geometry). See also here: http://en.wikipedia.org/wiki/Coastline_paradox I have a suggestion: break your proposed edits into smaller chunks. For example, in this case, propose an edit that *only* makes the 'intermittent' change that I've quoted above. Then propose separate edits that make each of the other changes. +1 cheers, Martin ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Edit review: intermittent waters
Martin Koppenhoefer wrote: 2012/5/29 Phil Endecott spam_from_osm_t...@chezphil.org: No doubt at some point someone decided that lake:shore_length:miles=2 was a useful thing to record, and you want to remove it. Ã Why? because there is no such thing as a shore length, it depends on the resolution. Aside from this there is also no tradition in OSM to record units in the key. So this particular key really doesn't seem to make any sense (the length of the shore in OSM-precision is already in the data geometry). See also here: http://en.wikipedia.org/wiki/Coastline_paradox This is of course true, and Worst Fixer's table at http://worstfixer.000a.biz/04-intermittent/overview-short.html would benefit from a column giving a whole series of justifications like this for each of the proposed changes. Regards, Phil. ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Edit review: intermittent waters
Hello. I used reply to instead of reply to all in my mail agent. We had a small thread with Thomas. Here is major result we achived. Thomas expressed opinion that not 100% of water=intermittent have other tags, so we have no way count them as water. In my sub-extract of water=intermittent: 127310 become intermittent=yes; 124417 have already natural tag; 749 get natural=wetland because NHD:FType=Inundation Area; 2123 have waterway tag; 21 has landuse tag. 127310-124417-749-2123-21=0. Check sum passed. This tag came from imports only, that is why it kann be cleaned up perfectly. -- WorstFixer, twitter: http://twitter.com/WorstFixer ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Edit review: intermittent waters
Actually, the conclusion, while it involved that, also involved that there are potential other uses (e.g. on river=intermittent; stream=intermittent etc) that need to be checked too, and that this seems like an arbitrary renaming of tags that doesn't gain anything, but may destroy data. Thanks Tom Davie On 29 May 2012, at 20:08, Worst Fixer wrote: Hello. I used reply to instead of reply to all in my mail agent. We had a small thread with Thomas. Here is major result we achived. Thomas expressed opinion that not 100% of water=intermittent have other tags, so we have no way count them as water. In my sub-extract of water=intermittent: 127310 become intermittent=yes; 124417 have already natural tag; 749 get natural=wetland because NHD:FType=Inundation Area; 2123 have waterway tag; 21 has landuse tag. 127310-124417-749-2123-21=0. Check sum passed. This tag came from imports only, that is why it kann be cleaned up perfectly. -- WorstFixer, twitter: http://twitter.com/WorstFixer ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Edit review: intermittent waters
Worst Fixer writes: Just removed stream=fixme. Why? Removed all id-like tags. Why? If no valid objections will be raised, I upload this change on 2012-06-12. Don't. -- --my blog is athttp://blog.russnelson.com Crynwr supports open source software 521 Pleasant Valley Rd. | +1 315-600-8815 Potsdam, NY 13676-3213 | Sheepdog ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Edit review: intermittent waters
Frederik Ramm writes:
2. I don't think you should continue to make mass edits under the
username WorstFixer because that implies that before you fixed
things they were among the worst which has the potential to offend people.
I always thought it meant that he was the worst person to be doing
this fixing, or doing the worst job at fixing things.
The problem with his worst fixing, is that unless he starts {bugging /
bothering / teaching / correcting / annoying / discouraging} mappers
who are making these edits he disagrees with, they're going to keep
making these edits.
I'd prefer to see a consensus among editors that we don't do that
anymore, with the old way not being in the OSM wiki anywhere,
backed up with an analysis of the timing of these edits as not
recently made, *BEFORE* any of these mass edits are made.
Because if people are still editing like this, then I don't see the
problem that WorstFixer is fixing as actually being fixed. Users of
the map will still have to deal with both the format that WorstFixer
doesn't like AND the format he does like. Less of the former and more
of the latter, but still some of both.
Persuade people to map just one way, THEN once they're doing that, go
back and get rid of the old way.
--
--my blog is athttp://blog.russnelson.com
Crynwr supports open source software
521 Pleasant Valley Rd. | +1 315-600-8815
Potsdam, NY 13676-3213 | Sheepdog
___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Edit review: intermittent waters
Could you post the code used to generate the changesets? That would be the easiest way for some of us to review your proposed changes. It wouldn’t remove the need to explain it to non-programmers, but it would be much easier for some of us if we could look at the code. That being said, a few concerns jump out at me - You provide no information on why you are removing the keys that you are - A couple of other people have expressed concern over the message sent by having a mechanical edit from the WorstFixer username. I share them. It is not enough to dismiss this as an objection that is not “valid” - A significant number of these ways appear to be from US NHD data. You should also consult specifically with the US community and develop a consensus there that the edit is worth doing, in addition to the global community. - The mapping from NHD FCode to OSM tags used for some of these imports may of not been ideal. I have been working on a better one but have not finished. I believe it would be best to exclude the US from this edit and later on (post-rebuild likely) propose an edit which includes changing tagging on untouched objects. From: Worst Fixer [mailto:worstfi...@gmail.com] Sent: Tuesday, May 29, 2012 4:51 AM To: talk@openstreetmap.org Subject: [OSM-talk] Edit review: intermittent waters Hello. I ask you to review my planned edit. There are lot of ways to tag intermittent water feature found in database. Most popular is intermittent=yes. All others come from different old imports. Date ist 2009-2010 year. I countiered ~350 000 features tagged in different such ways. Most is done by 10 users. I ask users iandees and SK53 join discussion, as most of such tags were imported by them. Others welcome too. I propose unification of tagging in all this imports. Following tags converted to intermittent=yes: frequency=intermittent occurrence=intermittent stream=intermittent water=intermittent type=intermittent Following tags converted to intermittent=no: frequency=perennial stream=perennial stream=ephemeral converted to intermittent=ephemeral. Just removed stream=fixme. Converted fdate field from NHD imports in iso8601 date, moved to check_date tag. Removed all id-like tags. If no valid objections will be raised, I upload this change on 2012-06-12. Here is overview: Short, to get the idea: http://worstfixer.000a.biz/04-intermittent/overview-short.html Long, for exact analysis: http://worstfixer.000a.biz/04-intermittent/overview-full.html.gz I currently looking for place to upload exact .osm.gz for a preview. Suggestions welcome. -- WorstFixer, twitter: http://twitter.com/WorstFixer ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
