Re: [OSM-talk] Mechanically Cleaning Up FIXME Tags
On 25/02/2015 08:51, Tobias Knerr wrote: On 25.02.2015 02:58, Bryce Nesbitt wrote: It is apparent that a number of imports have left tens of thousands of fixme notes that have a low chance of ever getting addressed. Pick your favorite from the lists above: set␣better␣denotation is my mine. That's from a mechanical edit that should never have happened in the first place. The edit was basically done in order to establish the denotation tag for trees, which was almost nonexistent before. The denotation values were not pulled from an external source, but based on guesses of the kind another tree within x meters = must be a cluster of trees. In my opinion, it could make sense to also remove the denotation keys on trees with set␣better␣denotation. After all, the continuing existence of that fixme shows that no human ever verified these. I also agree with the general goal to get rid of pointless fixme values. Just for a bit of background on this specific issue, for the lucky people who missed out on it last time around, the mechanical edit that added those values was discussed here: https://lists.openstreetmap.org/pipermail/tagging/2010-September/thread.html#4297 and there's some discussion (a couple of years after the event) on the German forum here: http://forum.openstreetmap.org/viewtopic.php?pid=309562 The discussion on the GB list lead to a revert there: https://lists.openstreetmap.org/pipermail/talk-gb/2010-November/010492.html More comment from the Netherlands: http://forum.openstreetmap.org/viewtopic.php?pid=121302#p121302 Cheers, Andy ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Mechanically Cleaning Up FIXME Tags
brycenesbitt wrote about mechanically reversing: fixme=set␣better␣denotation denotation=cluster Agreed on this one. And with 222k objects, that is a really big clean up. - -- sly, contact direct : sylvain /a\ letuffe o r g http://wiki.openstreetmap.org/wiki/User:Sletuffe -- View this message in context: http://gis.19327.n5.nabble.com/Mechanically-Cleaning-Up-FIXME-Tags-tp5834869p5835359.html Sent from the General Discussion mailing list archive at Nabble.com. ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Mechanically Cleaning Up FIXME Tags
Le 26.02.2015 19:25, Paul Johnson a écrit : Now that we have an anointed notes system, how about an automated move to notes, with the owner of the note being the person who originated the FIXME? Please, no. On http://resultmaps.neis-one.org/osm-notes-overview, [1] I prefer the first graph, showing how the notes db is already getting clustered -- too many openers, not enough closers. It seems that the OpenStreetBug experience is already getting reproduced (the negative part, not the beginning of it). JB. Links: -- [1] http://resultmaps.neis-one.org/osm-notes-overview, ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Mechanically Cleaning Up FIXME Tags
On 27 February 2015 at 14:30, JB jb...@mailoo.org wrote: On http://resultmaps.neis-one.org/osm-notes-overview, I prefer the first graph, showing how the notes db is already getting clustered I find the page in fact a bit hard to read (but apart from that very useful). What is the difference between the first and fourth graph? In my opinion, the main performance indicator should be the monthly number of notes closed. As long as the monthly number of closed notes is increasing, I think it doesn't really matter if the proportion of close-to-open actions is increasing too. Is there any way in which I can read the monthly number of close actions from the graphs? -- Matthijs ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Mechanically Cleaning Up FIXME Tags
On Fri, Feb 27, 2015 at 8:17 AM, Greg Troxel g...@ir.bbn.com wrote: I am still not sure I am following. A number of the top fixme values came from a single user, import, or manual (usually JOSM) edit. The proposal is a (semi) mechanical delete to check bounds, other tagging errors, then remove the keys. In other words, the process would start by loading the tagged objects and looking at them to see what's there. If someone wants to revert a changeset or delete data instead, that should be proposed first: http://taginfo.openstreetmap.org/search?q=FIXME ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Mechanically Cleaning Up FIXME Tags
On Wed, Feb 25, 2015 at 12:51 AM, Tobias Knerr o...@tobias-knerr.de wrote: On 25.02.2015 02:58, Bryce Nesbitt wrote: It is apparent that a number of imports have left tens of thousands of fixme notes that have a low chance of ever getting addressed. Pick your favorite from the lists above: set␣better␣denotation is my mine. That's from a mechanical edit that should never have happened in the first place. The edit was basically done in order to establish the denotation tag for trees, which was almost nonexistent before. The denotation values were not pulled from an external source, but based on guesses of the kind another tree within x meters = must be a cluster of trees. In my opinion, it could make sense to also remove the denotation keys on trees with set␣better␣denotation. After all, the continuing existence of that fixme shows that no human ever verified these. I also agree with the general goal to get rid of pointless fixme values. I've sent a message to the user, to open a discussion about mechanically reversing: fixme=set␣better␣denotation denotation=cluster Because these tags are scattered all over Europe, they are likely to eat mapper time during routine editing and fixme cleanup. The same tag appears on a limited number of seamarks http://www.openstreetmap.org/node/2456524071 ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Mechanically Cleaning Up FIXME Tags
On Fri, Feb 27, 2015 at 8:30 AM, JB jb...@mailoo.org wrote: Le 26.02.2015 19:25, Paul Johnson a écrit : Now that we have an anointed notes system, how about an automated move to notes, with the owner of the note being the person who originated the FIXME? Please, no. On http://resultmaps.neis-one.org/osm-notes-overview, I prefer the first graph, showing how the notes db is already getting clustered — too many openers, not enough closers. It seems that the OpenStreetBug experience is already getting reproduced (the negative part, not the beginning of it). So let's explore why people aren't working notes rather than complain about moving some objects that are essentially obsolete now that notes Is A Thing™. ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Mechanically Cleaning Up FIXME Tags
Bryce Nesbitt bry...@obviously.com writes: On Wed, Feb 25, 2015 at 1:10 AM, SomeoneElse li...@atownsend.org.uk wrote: On 25/02/2015 05:00, Bryce Nesbitt wrote: Any fixme in wide use I'm not interested in deleting. I'd strongly oppose the mechanical deletion of low volume fixme values. Mappers local to me often use individually worded fixmes describing something that needs investigation. By definition these values are not in wide use, but definitely should be kept. If I'm going to be in an area I always load the local notes and fixmes onto the Garmin so that if I'm near something that needs some attibute checking, I know about it. Hold on, you may have misunderstood. The only fixme tags proposed for deletion are the mechanically added ones on thousands of nodes. Any onesey twosey value would of course stay. Any value like continue that's has high counts, but edited by hundreds of unique users, would stay. I am still not sure I am following. There is a difference between: 1) identify foo as a fixme= value that is often associated with mechanical edit 2) therefore, remove all fixme=foo and 3) identify foo as a fixme= value that is often associated with mechanical edit 4) remove all fixme=foo tags *which were actually added by a mechanical edit* Do you mean 1/2 or do you mean 3/4? If 3/4, I think this is fine. If 1/2, I think it's going to remove hand-mapper fixme tags, which seems not ok. pgpF6KpEHBMw_.pgp Description: PGP signature ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Mechanically Cleaning Up FIXME Tags
On 2/26/2015 10:49 AM, Jonathan Bennett wrote: If the problem is in an area where there's no-one to survey, then so what? Fixmes don't show up on any end-user (as opposed to mapping QA) rendering, they don't mess up routing, they don't affect geocoding or have any other negative consequences for consumers of the data. So just leave them be until someone can get to the area to survey. J. I am strongly in this camp. I have not seen any actual harm or problem presented for 1.3 million fixme tags yet. But there is the potential for problems if removed. Even fixme=yes tags convey information: Someone felt something was in question about that node/way/polygon. That is not insignificant information. ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Mechanically Cleaning Up FIXME Tags
I agree. In most cases, a FIXME should be left until someone on-site can verify what is correct. -- John F. Eldredge -- j...@jfeldredge.com Darkness cannot drive out darkness; only light can do that. Hate cannot drive out hate; only love can do that. Dr. Martin Luther King, Jr. On February 26, 2015 3:49:58 AM Jonathan Bennett jonobenn...@gmail.com wrote: On 26/02/2015 08:43, Andreas Labres wrote: This confirmation of course could be automated: show the user the object with the tags on some areal imagery background and she/he can decide (in most cases, I'd say). No -- the aerial imagery could be out of date, and it may not be possible to tell if the problem has been fixed (or even existed in the first place) *only* from aerial images. Confirmation by survey would reliable. If the problem is in an area where there's no-one to survey, then so what? Fixmes don't show up on any end-user (as opposed to mapping QA) rendering, they don't mess up routing, they don't affect geocoding or have any other negative consequences for consumers of the data. So just leave them be until someone can get to the area to survey. J. ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Mechanically Cleaning Up FIXME Tags
Now that we have an anointed notes system, how about an automated move to notes, with the owner of the note being the person who originated the FIXME? On Thu, Feb 26, 2015 at 10:38 AM, John F. Eldredge j...@jfeldredge.com wrote: I agree. In most cases, a FIXME should be left until someone on-site can verify what is correct. -- John F. Eldredge -- j...@jfeldredge.com Darkness cannot drive out darkness; only light can do that. Hate cannot drive out hate; only love can do that. Dr. Martin Luther King, Jr. On February 26, 2015 3:49:58 AM Jonathan Bennett jonobenn...@gmail.com wrote: On 26/02/2015 08:43, Andreas Labres wrote: This confirmation of course could be automated: show the user the object with the tags on some areal imagery background and she/he can decide (in most cases, I'd say). No -- the aerial imagery could be out of date, and it may not be possible to tell if the problem has been fixed (or even existed in the first place) *only* from aerial images. Confirmation by survey would reliable. If the problem is in an area where there's no-one to survey, then so what? Fixmes don't show up on any end-user (as opposed to mapping QA) rendering, they don't mess up routing, they don't affect geocoding or have any other negative consequences for consumers of the data. So just leave them be until someone can get to the area to survey. J. ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Mechanically Cleaning Up FIXME Tags
Here's an example semi-bulk FIXME cleanup just done. This was manual, not script based: http://www.openstreetmap.org/changeset/29107328 Clearly this was a simple mistake (a JOSM user doing select all and getting nodes in addition to the ways they wanted to target). The original changeset was: http://www.openstreetmap.org/changeset/26696654 When it comes to cleanup, I don't view a fixme object as inherently more important than any other needed cleanup. If a fixme note adds information I don't already have, great. Else the fixme note is just eating time and mapping energy that could be better spent. - An example intentional tag I'd like to clear is: fixme=stream␣attibutes␣missing stream=fixme Along with any fixme that simply indicates missing data (since the lack of that data is obvious). ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Mechanically Cleaning Up FIXME Tags
On 25.02.15 08:01, Frederik Ramm wrote: I think in many cases the proper action to perform on an object with a FIXME tag that has a low chance of ever getting addressed is deletion. -1 You can't really tell if the problem wasn't fixed of if it was fixed and the user fixing it forgot to delete the fixme tag. Only if you can determine that the fixme tag definitively was from an import and this importing user has created (maybe last edited) that object, then it could eventually be deleted. But I'd prefer a visual confirmation by a human on any instance. This confirmation of course could be automated: show the user the object with the tags on some areal imagery background and she/he can decide (in most cases, I'd say). /al ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Mechanically Cleaning Up FIXME Tags
On 26/02/2015, Bryce Nesbitt bry...@obviously.com wrote: *One)* We have a fixme system where human mappers are encouraged to pay extra attention to particular areas or objects. *Two) *There is an issue of mapper fatigue: each mapper will look at only so many such tags in a lifetime of mapping. *Three)* The fixme system is not self-cleaning. Certain conditions result in fixme tags that are unlikely to be acted on. There are some 1.3 million open fixme tagged items, more than half from mechanical tagging. *Four)* In some cases the fixme tags happen to be associated with poor quality imports. But this is not universal: some poor data has fixme tags, other poor data does not. +1 on all that, except that 4) is barely relevant. If an import is so bad that it needs to be undone, I really hope that the presence of a fixme tag is not the only way to detect said import. - How about a two step process: *Step One ) * People who wish to delete a particular import look through the FIXME tagged items, and propose specific deletions. For example there's a bus stop import that looks to be of bad quality. If that data is removed, the fixme tag will go with it. Problem solved. *Make a specific proposal showing why the fixme tag is needed in order to clean the data.* Fair enough, but note that the problem being solved is the bad import, not the distracting fixme tags. *Step Two ) *Remaining fixme values with a count above 1 are reviewed. If they are deemed to add value, or if they come from many hand mapping efforts, they stay. The rest are mechanically trimmed. The usual find a frequently-used tag that ought to be deleted an maybe its associated data fixed process then. Not really specific to fixme tags, until you point out a particular fixme value that deserves the treatment. ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Mechanically Cleaning Up FIXME Tags
On 26/02/2015, Andreas Labres l...@lab.at wrote: On 25.02.15 08:01, Frederik Ramm wrote: I think in many cases the proper action to perform on an object with a FIXME tag that has a low chance of ever getting addressed is deletion. highway=service access=private surface=asphalt fixme=the surface type changes here according to imagery Clearly that fixme is unlikely to be resolved anytime soon (survey needed but impossible). And your suggested action is to delete the actual way object ?? I wouldn't even remove the fixme. Sorry for only offering an anecdote against your in many cases, but it reflects my general feeling. Deleting an unfixable fixme tag can make sense; deleting the underlying object doesn't. The fixme tag may be a hint that the underlying object is so bad that it needs to be deleted, but the tag itself is no reason to delete. An interesting version of don't shoot the messenger :p ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Mechanically Cleaning Up FIXME Tags
On 26/02/2015 08:43, Andreas Labres wrote: This confirmation of course could be automated: show the user the object with the tags on some areal imagery background and she/he can decide (in most cases, I'd say). No -- the aerial imagery could be out of date, and it may not be possible to tell if the problem has been fixed (or even existed in the first place) *only* from aerial images. Confirmation by survey would reliable. If the problem is in an area where there's no-one to survey, then so what? Fixmes don't show up on any end-user (as opposed to mapping QA) rendering, they don't mess up routing, they don't affect geocoding or have any other negative consequences for consumers of the data. So just leave them be until someone can get to the area to survey. J. ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Mechanically Cleaning Up FIXME Tags
I think that it may be a good idea to considering deleting fixmes that: - were added by mechanical mass edit - were added to existing objects - are completely useless - are not indicating low quality of data/tags For example set␣better␣denotation is not fitting - it seems that in this case also denotation tag should be removed, together with fixme. Obviously these restrictions apply only to an automatic edit. 2015-02-25 2:58 GMT+01:00 Bryce Nesbitt bry...@obviously.com: I'm opening a discussion about a potential mechanical edit to FIXME tags: http://taginfo.openstreetmap.org/keys/fixme#values http://taginfo.openstreetmap.org/keys/FIXME#values It is apparent that a number of imports have left tens of thousands of fixme notes that have a low chance of ever getting addressed. Pick your favorite from the lists above: set␣better␣denotation is my mine. The goal would be to reduce the pain felt by anyone with fixme warnings turned on in their editing tool. ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Mechanically Cleaning Up FIXME Tags
On 25/02/2015, Bryce Nesbitt bry...@obviously.com wrote: fixme=yes is an interesting one socially. It's a bit like tiger:reviewed=no Yes. I'm also willing to bet that in manny cases fixme=yes was added by mistake, without the mapper understanding what it means. If there's an obvious problem, I might feel confident to fix the issue and clear the tag out. Sure. But for most nodes I might be unsure what's wrong, or not be confident I know 100% about the object. Thus the fixme=yes sits there forever, for future generations of mappers to look at, make the same conclusion, and leave the tag for the next next generation of mapper. IMHO if you're an experienced mapper in this area and you couldn't make head or tails of a fixme, nobody will. So shamelessly remove the distracting tag. It's no worse than the aging process you suggest in another mail. ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Mechanically Cleaning Up FIXME Tags
I think that it may be a good idea to considering automated deleting fixmes that: - were added by mechanical mass edit - were added to existing objects - are completely useless - are not indicating low quality of data/tags For example set␣better␣denotation is not fitting - it seems that in this case also denotation tag should be removed, together with fixme. 2015-02-25 2:58 GMT+01:00 Bryce Nesbitt bry...@obviously.com: I'm opening a discussion about a potential mechanical edit to FIXME tags: http://taginfo.openstreetmap.org/keys/fixme#values http://taginfo.openstreetmap.org/keys/FIXME#values It is apparent that a number of imports have left tens of thousands of fixme notes that have a low chance of ever getting addressed. Pick your favorite from the lists above: set␣better␣denotation is my mine. The goal would be to reduce the pain felt by anyone with fixme warnings turned on in their editing tool. ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Mechanically Cleaning Up FIXME Tags
On 26/02/2015 18:25, Paul Johnson wrote: Now that we have an anointed notes system, how about an automated move to notes, with the owner of the note being the person who originated the FIXME? Personally I'd rather keep any FIXMEs on the objects that they relate to. -- Steve --- This email has been checked for viruses by Avast antivirus software. http://www.avast.com ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Mechanically Cleaning Up FIXME Tags
To make this simpler, for now I propose to mechanically delete the tags: fixme=stream␣attibutes␣missing stream=fixme From several stream imports in the USA. Does anyone have comment or considerations for that proposal (beyond the usual mechanical edit policy)? ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [Talk-us] [OSM-talk] Mechanically Cleaning Up FIXME Tags
To make this simpler, for now I propose to mechanically delete the tags: fixme=stream␣attibutes␣missing stream=fixme From several stream imports in the USA. Does anyone have comment or considerations for that proposal (beyond the usual mechanical edit policy)? ___ Talk-us mailing list Talk-us@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-us
Re: [OSM-talk] Mechanically Cleaning Up FIXME Tags
On Thu, Feb 26, 2015 at 3:55 PM, Blake Girardot bgirar...@gmail.com wrote: I am strongly in this camp. I have not seen any actual harm or problem presented for 1.3 million fixme tags yet. But there is the potential for problems if removed. Even fixme=yes tags convey information: Someone felt something was in question about that node/way/polygon. That is not insignificant information. At no point has a proposal been made to remove fixme=yes. -- As for harm, any user of a tool that shows fixme tags: http://keepright.ipax.at/report_map.php?zoom=12lat=39.95356lon=-75.12364 Is directing energy to both useful fixme tags, and fixme tags that at best are markers of a poorly considered import. A bulk purge cold save a lot of time. I do encourage people to remove fixme keys when they edit: many times the underlying problem gets fixed, but the fixme tag stays. ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Mechanically Cleaning Up FIXME Tags
But mass-removing that import's fixmes I meant : mass-removing the import's objects -- sly, direct contact : sylv...@letuffe.org http://wiki.openstreetmap.org/wiki/User:Sletuffe - -- sly, contact direct : sylvain /a\ letuffe o r g http://wiki.openstreetmap.org/wiki/User:Sletuffe -- View this message in context: http://gis.19327.n5.nabble.com/Mechanically-Cleaning-Up-FIXME-Tags-tp5834869p5834934.html Sent from the General Discussion mailing list archive at Nabble.com.___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Mechanically Cleaning Up FIXME Tags
On 25/02/2015, Martin Koppenhoefer dieterdre...@gmail.com wrote: 2015-02-25 11:07 GMT+01:00 moltonel 3x Combo molto...@gmail.com: * fixme=yes Provides no information -1, it indicates a certain overaverage probability of some (not further specified) problem It can also indicate a tagging error (as in adding this fixme was unintentional). In cases where it isn't it still falls IMHO under the usefulness threshold. There's probably an object right beside it that has a bigger issue but isn't tagged with fixme. I don't particularly care about mass-deleting fixemes to begin with, but if I happen to be editing an object tagged fixme=yes I'll certainly remove the tag. ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Mechanically Cleaning Up FIXME Tags
On 25/02/2015, sly (sylvain letuffe) lis...@letuffe.org wrote: I do also agree with Frederic, imports of external data not conflated added with some fixme=please fix my bad import by surveying it on the ground should be remove alltogether. Good integration should be done at import time and should'nt rely on others by spamming the fixmes. An import that tagged its object with fixmes is arguably a bad quality import. But mass-removing that import's fixmes (assuming those fixmes carry usefull info) degrades data quality further. Two bads don't make a good. If you're talking about fixmes that don't carry usefull info, then it doesn't matter wether they were created by imports or not. For all the other manually entered fixme, a lot of care and discussion should be performed before attempting anything automatically. Even the rather useless fixme=yes should be handeled with care. The mapper might have added information that he knows to be perfectible and wanted to enlight that. My guess is that manually-entered fixmes will not lend themselves to automated fixing at all. You'd spend more time fixing your script's heuristics than you'd spend manually doing the fixes. ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Mechanically Cleaning Up FIXME Tags
On 25/02/2015, sly (sylvain letuffe) lis...@letuffe.org wrote: On mercredi 25 février 2015, you wrote: +1, in any case it seems more likely that the problems gets fixed with a fixme tag than without. -1 I do think this is is only true in a short term view. Push that reasoning to the extrem, should we add fixme=* tags for every QA tool report, people will get lazy about it and just stop fixing things. The initial goal of more fixme, more chances to get fixed might be lost That extreme goes against the if existing QA tools can spot the issue already, then a fixme is not needed rule, so it won't happen. A bloated fixme/notes db is indeed no fun. But removing usefull notes is like a company canceling contracts to solve server scaling issues. We just need to agree on the usefullness threshold, which of course is a subjective thing. ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Mechanically Cleaning Up FIXME Tags
brycenesbitt wrote It's also possible to turn some of those like could_be_dunes_or_beach into notes, rather than FIXME. -1 Don't push dust under another carpet - -- sly, contact direct : sylvain /a\ letuffe o r g http://wiki.openstreetmap.org/wiki/User:Sletuffe -- View this message in context: http://gis.19327.n5.nabble.com/Mechanically-Cleaning-Up-FIXME-Tags-tp5834869p5834935.html Sent from the General Discussion mailing list archive at Nabble.com. ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Mechanically Cleaning Up FIXME Tags
On Wed, Feb 25, 2015 at 1:10 AM, SomeoneElse li...@atownsend.org.uk wrote: On 25/02/2015 05:00, Bryce Nesbitt wrote: Any fixme in wide use I'm not interested in deleting. I'd strongly oppose the mechanical deletion of low volume fixme values. Mappers local to me often use individually worded fixmes describing something that needs investigation. By definition these values are not in wide use, but definitely should be kept. If I'm going to be in an area I always load the local notes and fixmes onto the Garmin so that if I'm near something that needs some attibute checking, I know about it. Hold on, you may have misunderstood. The only fixme tags proposed for deletion are the mechanically added ones on thousands of nodes. Any onesey twosey value would of course stay. Any value like continue that's has high counts, but edited by hundreds of unique users, would stay. --- The goal is to delete only the real junk. The goal is to cut the volume down to something human mappers can manage. ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Mechanically Cleaning Up FIXME Tags
On 25/02/2015 17:23, Bryce Nesbitt wrote: Hold on, you may have misunderstood. Actually I think you've misunderstood: You've said these are Junk Tags, and I think everyone has agreed with you on that. However people have also pointed out that they are probably attached to Junk Data. This means deleting the tags mechanically won't do anything to solve the problem of junk data, only make it harder to find for the average mapper. Therefore we shouldn't mechanically delete any tags, because the net result will be to make OSM *as a whole* worse. Hopefully that makes sense to you. J. ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Mechanically Cleaning Up FIXME Tags
On 25/02/2015, Jonathan Bennett jonobenn...@gmail.com wrote: On 25/02/2015 17:23, Bryce Nesbitt wrote: Hold on, you may have misunderstood. Actually I think you've misunderstood: Being the thread starter, I doubt that Bryce has misunderstood the point of the thread. You've said these are Junk Tags, and I think everyone has agreed with you on that. No, not everybody agrees that all fixme tags from imports are junk tags. Some may be junk. Some may indicate a badly-done import that should have fixed the issue before import. Some are reasonable use of fixme during an import (such as tiger's dual_carriageway). However people have also pointed out that they are probably attached to Junk Data. Please give specific examples. For some cases it may be true, but there's no sense in drawing a general conclusion from maybes and anecdotes. This means deleting the tags mechanically won't do anything to solve the problem of junk data, only make it harder to find for the average mapper. Therefore we shouldn't mechanically delete any tags, because the net result will be to make OSM *as a whole* worse. Yes, altough I argued earlyer that in some cases just removing fixmes (mechanically or not) is the right thing to do. Just please pretty please stop making blanket statements about all fixme tags. ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Mechanically Cleaning Up FIXME Tags
fixme=yes is an interesting one socially. It's a bit like tiger:reviewed=no If there's an obvious problem, I might feel confident to fix the issue and clear the tag out. But for most nodes I might be unsure what's wrong, or not be confident I know 100% about the object. Thus the fixme=yes sits there forever, for future generations of mappers to look at, make the same conclusion, and leave the tag for the next next generation of mapper. --- Tools such as Keepright could sidestep some of the fixme clutter by aging out older notes. As a given fixme ages, the percentage chance of it getting fixed goes down. With JOSM at least reading the last edit date is a separate and somewhat slow operation for the mapper. ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Mechanically Cleaning Up FIXME Tags
On Wed, Feb 25, 2015 at 9:57 AM, Jonathan Bennett jonobenn...@gmail.com wrote: Actually I think you've misunderstood: You've said these are Junk Tags, and I think everyone has agreed with you on that. However people have also pointed out that they are probably attached to Junk Data. How about if I summarize the discussion so far, and see if there is misunderstanding: *One)* We have a fixme system where human mappers are encouraged to pay extra attention to particular areas or objects. *Two) *There is an issue of mapper fatigue: each mapper will look at only so many such tags in a lifetime of mapping. *Three)* The fixme system is not self-cleaning. Certain conditions result in fixme tags that are unlikely to be acted on. There are some 1.3 million open fixme tagged items, more than half from mechanical tagging. *Four)* In some cases the fixme tags happen to be associated with poor quality imports. But this is not universal: some poor data has fixme tags, other poor data does not. Similarly some mechanically added fixme tags are valuable, some are (ahem) less so. - How about a two step process: *Step One ) * People who wish to delete a particular import look through the FIXME tagged items, and propose specific deletions. For example there's a bus stop import that looks to be of bad quality. If that data is removed, the fixme tag will go with it. Problem solved. *Make a specific proposal showing why the fixme tag is needed in order to clean the data.* *Step Two ) *Remaining fixme values with a count above 1 are reviewed. If they are deemed to add value, or if they come from many hand mapping efforts, they stay. The rest are mechanically trimmed. ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Mechanically Cleaning Up FIXME Tags
To me, fixme tags always come with the implied 'I don't know, it needs surveying from a more knowledgeable local mapper' so I don't think a widespread mechanical edit will help. As others have said: What's the problem with these tags? As they're not harming the database. How will deleting them improve the quality of the database? Keep Right has the option to ignore fixmes that are irrelevant to a user. From what I've seen information added to 'Notes' is worse than fixmes. For a really useful mass edit, remove all the created_by=* tags that are attached to thousands (millions?) of nodes. Dave F. On 25/02/2015 02:37, Hans De Kryger wrote: Sounds like a great idea. My favorites to work on is dual␣carriageway http://taginfo.openstreetmap.org/tags/FIXME=dual%20carriageway not␣dual␣carriageway http://taginfo.openstreetmap.org/tags/FIXME=not%20dual%20carriageway. *Regards,** * *Hans* * * *http://www.openstreetmap.org/user/TheDutchMan13 * * * On Tue, Feb 24, 2015 at 6:58 PM, Bryce Nesbitt bry...@obviously.com mailto:bry...@obviously.com wrote: I'm opening a discussion about a potential mechanical edit to FIXME tags: http://taginfo.openstreetmap.org/keys/fixme#values http://taginfo.openstreetmap.org/keys/FIXME#values It is apparent that a number of imports have left tens of thousands of fixme notes that have a low chance of ever getting addressed. Pick your favorite from the lists above: set␣better␣denotation is my mine. The goal would be to reduce the pain felt by anyone with fixme warnings turned on in their editing tool. ___ talk mailing list talk@openstreetmap.org mailto:talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk --- This email has been checked for viruses by Avast antivirus software. http://www.avast.com ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Mechanically Cleaning Up FIXME Tags
I agree with Tordanik that automated addition of fixme=* tags to previously existing objects (just like set better denotation on natural=tree) is counter productive, and not only displays warning on JOSM, but make my days on the ground worst as my osmand is showing non human asked corrections. (Better build an external reporting tool for that) I do also agree with Frederic, imports of external data not conflated added with some fixme=please fix my bad import by surveying it on the ground should be remove alltogether. Good integration should be done at import time and should'nt rely on others by spamming the fixmes. For all the other manually entered fixme, a lot of care and discussion should be performed before attempting anything automatically. Even the rather useless fixme=yes should be handeled with care. The mapper might have added information that he knows to be perfectible and wanted to enlight that. Tordanik wrote On 25.02.2015 02:58, Bryce Nesbitt wrote: It is apparent that a number of imports have left tens of thousands of fixme notes that have a low chance of ever getting addressed. Pick your favorite from the lists above: set␣better␣denotation is my mine. That's from a mechanical edit that should never have happened in the first place. The edit was basically done in order to establish the denotation tag for trees, which was almost nonexistent before. The denotation values were not pulled from an external source, but based on guesses of the kind another tree within x meters = must be a cluster of trees. In my opinion, it could make sense to also remove the denotation keys on trees with set␣better␣denotation. After all, the continuing existence of that fixme shows that no human ever verified these. I also agree with the general goal to get rid of pointless fixme values. ___ talk mailing list talk@ https://lists.openstreetmap.org/listinfo/talk - -- sly, contact direct : sylvain /a\ letuffe o r g http://wiki.openstreetmap.org/wiki/User:Sletuffe -- View this message in context: http://gis.19327.n5.nabble.com/Mechanically-Cleaning-Up-FIXME-Tags-tp5834869p5834907.html Sent from the General Discussion mailing list archive at Nabble.com. ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Mechanically Cleaning Up FIXME Tags
2015-02-25 11:07 GMT+01:00 moltonel 3x Combo molto...@gmail.com: How will removing any fixme tag make the actual _data_ in OSM better? It'll just make it harder for people editing it to determine what is good data and what isn't. +1 I'm willing to believe that a particular fixme value is useless (either because it was fixed without removing the tag or because it is now deemed not a problem or because QA tools do a better job of pointing out the error), but you'll have to make the case for each individual value. None of the values in your get rid list are obviously useless to me, please explain. +1 The fact that something has a low chance to get fixed is irrelevant. +1, in any case it seems more likely that the problems gets fixed with a fixme tag than without. Now here's my shortlist of fixme tags that can be simply deleted, with explanation : * fixme=yes Provides no information -1, it indicates a certain overaverage probability of some (not further specified) problem * fixme=name and variants of add_full_address when QA tools agree that a name or address is missing The QA tools do a better job of pointing those out. But watch out for fixme=name that really mean the tagged name is wrong for example. +1 cheers, Martin ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Mechanically Cleaning Up FIXME Tags
On mercredi 25 février 2015, you wrote: +1, in any case it seems more likely that the problems gets fixed with a fixme tag than without. -1 I do think this is is only true in a short term view. Push that reasoning to the extrem, should we add fixme=* tags for every QA tool report, people will get lazy about it and just stop fixing things. The initial goal of more fixme, more chances to get fixed might be lost I can't resist re-linking to my old (hopefully failed?) prediction about Notes (there is a paragraph about fixme): http://gis.19327.n5.nabble.com/Death-prediction-of-notes-on-osm-org-as-it- seams-to-be-planned-tt5737341.html -- sly qui suis-je : http://sly.letuffe.org - -- sly, contact direct : sylvain /a\ letuffe o r g http://wiki.openstreetmap.org/wiki/User:Sletuffe -- View this message in context: http://gis.19327.n5.nabble.com/Mechanically-Cleaning-Up-FIXME-Tags-tp5834869p5834913.html Sent from the General Discussion mailing list archive at Nabble.com.___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Mechanically Cleaning Up FIXME Tags
On 25/02/2015 07:01, Frederik Ramm wrote: I think in many cases the proper action to perform on an object with a FIXME tag that has a low chance of ever getting addressed is deletion. +1 Though just as blind mechanical imports should not be done, so blind mechanical deletes should also not be done. There are some 11000 seamark:fixme=please_fix_position objects that are from just such an import that was done over 4 years ago. I have been slowly working my way through these and have found that there are many that have been moved, but the fixme tag has not been removed. Here I remove the tag. On the others I check to see if there is any recent active editing of these objects nearby - if so, I leave them untouched. Only in areas where no such activity is evident, do I perform a mass delete. If encouraged by the community, I will speed up this process, though at a measured pace so as to not overload the seamark renderer. ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Mechanically Cleaning Up FIXME Tags
Unsure how that will resolve any of the problems. On 25/02/2015 05:02, Bryce Nesbitt wrote: It's also possible to turn some of those like could_be_dunes_or_beach into notes, rather than FIXME. ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk --- This email has been checked for viruses by Avast antivirus software. http://www.avast.com ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Mechanically Cleaning Up FIXME Tags
On 25/02/2015, SomeoneElse li...@atownsend.org.uk wrote: On 25/02/2015 05:00, Bryce Nesbitt wrote: Get rid: fixme=check/adjust␣position␣and/or␣merge␣with␣existing␣stop␣if␣exists fixme=type_of_palm fixme=imported_to_be_checked FIXME=stream␣attribute␣data␣missing Keep: fixme=continue fixme=position fixme=resurvey fixme=dual_carriageway I may be missing something here, but what actually is the benefit of this? Even in the situation where a problematical import brought in lots of Palm Trees but not their species (or whatever) the fixme tag is still serving a useful purpose - in this case this data was imported by a problematical import. How will removing any fixme tag make the actual _data_ in OSM better? It'll just make it harder for people editing it to determine what is good data and what isn't. +1 I'm willing to believe that a particular fixme value is useless (either because it was fixed without removing the tag or because it is now deemed not a problem or because QA tools do a better job of pointing out the error), but you'll have to make the case for each individual value. None of the values in your get rid list are obviously useless to me, please explain. The fact that something has a low chance to get fixed is irrelevant. There's an infinite amount of OSM data that needs fixing but have a low chance to be. Somebody made the effort to point out particular things with fixme tags, please respect that. The problem is more likely to get fixed with the tag than without. I'm also not a fan of converting fixmes to notes. I use both fixmes and notes, to me they fit different workflows and audiences. And unleashing 0.5M notes will make browsers unhappy. Again, if you think that a particular fixme value needs to be mass-converted to a note, please explain individual cases. Now here's my shortlist of fixme tags that can be simply deleted, with explanation : * fixme=yes Provides no information * fixme=name and variants of add_full_address when QA tools agree that a name or address is missing The QA tools do a better job of pointing those out. But watch out for fixme=name that really mean the tagged name is wrong for example. * Is there a QA tool pointing out the equivalent of stream_attribute_missing ? ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Mechanically Cleaning Up FIXME Tags
On 25.02.2015 02:58, Bryce Nesbitt wrote: It is apparent that a number of imports have left tens of thousands of fixme notes that have a low chance of ever getting addressed. Pick your favorite from the lists above: set␣better␣denotation is my mine. That's from a mechanical edit that should never have happened in the first place. The edit was basically done in order to establish the denotation tag for trees, which was almost nonexistent before. The denotation values were not pulled from an external source, but based on guesses of the kind another tree within x meters = must be a cluster of trees. In my opinion, it could make sense to also remove the denotation keys on trees with set␣better␣denotation. After all, the continuing existence of that fixme shows that no human ever verified these. I also agree with the general goal to get rid of pointless fixme values. ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Mechanically Cleaning Up FIXME Tags
On 25/02/2015 05:00, Bryce Nesbitt wrote: Any fixme in wide use I'm not interested in deleting. I'd strongly oppose the mechanical deletion of low volume fixme values. Mappers local to me often use individually worded fixmes describing something that needs investigation. By definition these values are not in wide use, but definitely should be kept. If I'm going to be in an area I always load the local notes and fixmes onto the Garmin so that if I'm near something that needs some attibute checking, I know about it. Get rid: fixme=check/adjust␣position␣and/or␣merge␣with␣existing␣stop␣if␣exists fixme=type_of_palm fixme=imported_to_be_checked FIXME=stream␣attribute␣data␣missing Keep: fixme=continue fixme=position fixme=resurvey fixme=dual_carriageway I may be missing something here, but what actually is the benefit of this? Even in the situation where a problematical import brought in lots of Palm Trees but not their species (or whatever) the fixme tag is still serving a useful purpose - in this case this data was imported by a problematical import. How will removing any fixme tag make the actual _data_ in OSM better? It'll just make it harder for people editing it to determine what is good data and what isn't. Cheers, Andy ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Mechanically Cleaning Up FIXME Tags
On 25/02/2015 07:01, Frederik Ramm wrote: +1 to all that. ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Mechanically Cleaning Up FIXME Tags
Hi, I think in many cases the proper action to perform on an object with a FIXME tag that has a low chance of ever getting addressed is deletion. For example, if something is tagged fixme=check/adjust␣position␣and/or␣merge␣with␣existing␣stop␣if␣exists then that's clearly a failed import (as it should have established a procedure to conflate) and if nobody values the data enough to bring it into shape, then the least we want to do is clean the fixme tag and keep the broken data. Or am I misunderstanding something? Bye Frederik -- Frederik Ramm ## eMail frede...@remote.org ## N49°00'09 E008°23'33 ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
[OSM-talk] Mechanically Cleaning Up FIXME Tags
I'm opening a discussion about a potential mechanical edit to FIXME tags: http://taginfo.openstreetmap.org/keys/fixme#values http://taginfo.openstreetmap.org/keys/FIXME#values It is apparent that a number of imports have left tens of thousands of fixme notes that have a low chance of ever getting addressed. Pick your favorite from the lists above: set␣better␣denotation is my mine. The goal would be to reduce the pain felt by anyone with fixme warnings turned on in their editing tool. ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Mechanically Cleaning Up FIXME Tags
Sounds like a great idea. My favorites to work on is dual␣carriageway http://taginfo.openstreetmap.org/tags/FIXME=dual%20carriageway not␣dual␣carriageway http://taginfo.openstreetmap.org/tags/FIXME=not%20dual%20carriageway. *Regards,* *Hans* *http://www.openstreetmap.org/user/TheDutchMan13 http://www.openstreetmap.org/user/TheDutchMan13* On Tue, Feb 24, 2015 at 6:58 PM, Bryce Nesbitt bry...@obviously.com wrote: I'm opening a discussion about a potential mechanical edit to FIXME tags: http://taginfo.openstreetmap.org/keys/fixme#values http://taginfo.openstreetmap.org/keys/FIXME#values It is apparent that a number of imports have left tens of thousands of fixme notes that have a low chance of ever getting addressed. Pick your favorite from the lists above: set␣better␣denotation is my mine. The goal would be to reduce the pain felt by anyone with fixme warnings turned on in their editing tool. ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Mechanically Cleaning Up FIXME Tags
It's also possible to turn some of those like could_be_dunes_or_beach into notes, rather than FIXME. ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Mechanically Cleaning Up FIXME Tags
Any fixme in wide use I'm not interested in deleting. Get rid: fixme=check/adjust␣position␣and/or␣merge␣with␣existing␣stop␣if␣exists fixme=type_of_palm fixme=imported_to_be_checked FIXME=stream␣attribute␣data␣missing Keep: fixme=continue fixme=position fixme=resurvey fixme=dual_carriageway Speaking of which: is there a way to get the number of unique editor ID's for fixme=continue? ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk