[HOT] JOSM people validating / mapping in Nepal, please run JOSM's validator before finishing a tile
kusala9 at googlemail.com
Thu May 7 13:21:11 UTC 2015
An interesting thread - I work for a national mapping agency and we collate
vector data from all over the world. My personal view is that a vector
database can either be consistent or complete and, when over certain size,
is rarely both. We run complex validation post-compilation across millions
of datasets and do some corrections automatically. Fixing them at source is
impossible - it's preferable to support the community with encoding guides
and comprehensive tools (some "inconsistencies" are just personal, national
or cartographic preferences). My observation on OSM is that the same
dynamic applies but the benefits far far outweigh the inconsistencies. as
the project grows so its organisation and tools must evolve to adapt to the
larger population.... jon.
On Thu, May 7, 2015 at 1:10 PM, Dave Corley <davecorley at gmail.com> wrote:
> Hi all,
> Just to clarify, josm & id both now validate but it's possible newbies
> don't know how to handle the returned errors and just ignore them.
> The untagged link in my original email was given as a sample only. There
> are also many, many issues with routing, boundaries, waterways, etc etc
> Granted there may be some false positives, but those would be the
> On 7 May 2015 12:27, "Florian Lohoff" <f at zz.de> wrote:
>> On Thu, May 07, 2015 at 12:57:06AM -0700, Springfield Harrison wrote:
>> > Hello Dave,
>> > This is amazing to see the vast number of invalid tags. This
>> > calls into question the integrity of the database. Do you have much
>> > getting people to run the validator?
>> In OSM there is no such thing as an invalid tag. Everybody is allowed to
>> invent new tags although your expectation should not be that there is a
>> single data consumer who will make use of it.
>> The common set of tags has been selected by a "do-ocracy" - You see
>> data consumers making use of a certain tag and you start tagging to
>> influence these data consumers.
>> > I am baffled that the data validation does not take place right
>> at the
>> > data entry stage. This is very common in database applications. All
>> > critical fields have validation rules so that the operator can neither
>> skip the
>> > critical fields nor enter data that is not applicable to that field.
>> If JOSM,
>> > complex as it is, is lacking input data validation, that is a serious
>> > in my opinion. For this type of mission, complete and accurate data is
>> > critical. You do not have the luxury of time hoping that people will
>> > with a post entry validation process.
>> Both ways are perfectly valid - validating on input or on data
>> OSM has gone the way of validating or using data on the consumer side
>> which i am very happy about. I have invented a tag myself for hinting a
>> routing instance for infrastructure e.g. Telecoms cables. If there would
>> have been an input validation i could not have done so.
>> Florian Lohoff f at zz.de
>> We need to self-defense - GnuPG/PGP enable your email today!
>> -----BEGIN PGP SIGNATURE-----
>> Version: GnuPG v1.4.10 (GNU/Linux)
>> -----END PGP SIGNATURE-----
> HOT mailing list
> HOT at openstreetmap.org
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the HOT