Re: [Tagging] Hotel dataset import? / Re: Baby-sitting

[Martin Koppenhoefer]

sent from a phone

> On 13. Mar 2019, at 21:30, Sergio Manzi  wrote:
> 
> As far as regards Giovanni's import, my position is that I'm pretty sure that 
> the work he has done is of high quality,
> 

the thing is, Giovanni can do the best work in the world, the resulting data 
will still completely depend on the external dataset.



> probably much better than many "locally sourced" information (you probably 
> have an idea of the quantity of "locally sourced" crap I see here in Venice, 
> mainly self-defined hotels and  B which are... nothing
> 

the situation in tourist hotspots is generally different from other areas (much 
more activity in general, and more SEO activity), the Venice map is not 
representative for most of OSM, it is representative for tourist areas.


> if I want an Hotel or a room I can go to Tripadvisor or to Airbnb (not 
> endorsing those in any way, just an example...), see if it accepts my pet and 
> my credit card and then use OSM to navigate to the address.
> 

Remote areas apart, I would not use OSM to search for a place to stay either. 
Usually the main criteria are location, standard, availability and price, of 
which only location can be answered with OSM data.
Still I believe the location of hotels is an interesting geographic 
information. You can do more with the hotel information than searching for a 
hotel to stay.

Cheers, Martin ___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Hotel dataset import? / Re: Baby-sitting

[Sergio Manzi]

On 2019-03-13 21:30, Sergio Manzi wrote:
> I also think that, as you correctly pointed out, both me and *Peter*, (/I 
> guess.../) forgot about "armchair mapping" which is neither "import" nor 
> "locally sourced" and probably represent a huge amount of data in OSM.


My bad! Please read:

I also think that, as you correctly pointed out, both me and *Tom*, (/I 
guess.../) forgot about "armchair mapping" which is neither "import" nor 
"locally sourced" and probably represent a huge amount of data in OSM.



smime.p7s
Description: S/MIME Cryptographic Signature
___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Hotel dataset import? / Re: Baby-sitting

[Sergio Manzi]

On 2019-03-13 19:04, Volker Schmidt wrote:
> On Wed, 13 Mar 2019 at 15:45, Sergio Manzi mailto:s...@smz.it>> 
> wrote:
>
> Have mappers walked along the whole world coastlines? Have they descended 
> all world's rivers by canoe?
>
> No, these are examples where imports make sense as these are feature that 
> change slowly (normally) 
>
> Have all Mumbay, New York, Rome, Shanghai, etc., streets and alleys been 
> walked by volunteers mapping their location and names?
>
> No, but many have been photographed and the photographs are available with 
> open license (Mapillary and OpenStreetCam) and many raods have been recorded 
> with GPX tracks. This allows armchair mapping with high quality results, 
> where the mapper does not need to be the photographer.
> New York on OpenStreetCam 
> , 
> Amsterdam on Mapillary 
> 
>
> All the peaks escalated?
>
> Again classical data that change slowly, where imports are justified.
>
> But when you start importing trees or buildings or hotels or petrol stations 
> or shops or crops than you are importing  perishable data which are 
> maintained outside OSM, and hence need frequent updates, with the obvious 
> cost that they need to be updated in the external source database and in the 
> OSM database in synchronism.
>
> Volker


Fair enough, Volker, I agree on all your points. But let's recap, if you don't 
mind.

My objection was to Tom Pfeifer assertion "/I think you misunderstand. OSM is 
*based **on locally sourced, handcrafted data*. That creates the high quality./"

To me "/based on/" means the bulk of data which constitues OSM foundation.

I also have to admit that as far as concern myself, I'm very much more 
interested in what you call "/data that changes slowly/" and I see that as the 
OSM foundation:

  * natural terrain with all its features like hills, mountains, rivers and 
lakes, forests and land cover in general
  * man made features like highways, canals, buildings, bridges, etc.
  * persistent and generally useful information like street numbers

The above, slowly changing information, is what I think represent OSM "base" 
information, and I don't believe the bulk of the above is "locally sourced, 
handcrafted data", but I'm ready to change my mind if proved wrong.

I also think that, as you correctly pointed out, both me and Peter, (/I 
guess.../) forgot about "armchair mapping" which is neither "import" nor 
"locally sourced" and probably represent a huge amount of data in OSM.



As far as regards Giovanni's import, my position is that I'm pretty sure that 
the work he has done is of high quality, probably much better than many 
"locally sourced" information (/you probably have an idea of the quantity of 
"locally sourced" crap I see here in Venice, mainly self-defined hotels and  
B which are... nothing... just the address of people renting their 
rooms/apartment/ and stick a name on it).

Personally I would be much happier if OSM would limit itself to its "base 
layer" and let other data aggregators handle different kind of information: if 
I want an Hotel or a room I can go to Tripadvisor or to Airbnb (/not endorsing 
those //_in any way_//, just an example.../), see if it accepts my pet and my 
credit card and then use OSM to navigate to the address.

Unhappily things are not like this and we are mapping hotels. In this regard I 
strongly tend to trust Giovanni import much more than the "locally sourced" 
data.

Cheers,

Sergio






smime.p7s
Description: S/MIME Cryptographic Signature
___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Hotel dataset import? / Re: Baby-sitting

[Sergio Manzi]

On 2019-03-13 20:22, Martin Koppenhoefer wrote:
>
> On 13. Mar 2019, at 16:34, Sergio Manzi mailto:s...@smz.it>> 
> wrote:
>
>> ... an example of massively imported data (/which I agree with, btw.../): 
>> https://blogs.bing.com/maps/2018-06/microsoft-releases-125-million-building-footprints-in-the-us-as-open-data
>
>
> This data wasn’t generally imported AFAIK (I believe a small part of it was 
> imported). Here’s the wiki page: 
> https://wiki.openstreetmap.org/wiki/Microsoft_Building_Footprint_Data
>
> here an import:
> https://wiki.openstreetmap.org/wiki/Microsoft_Building_Import_-_North_Central_Texas
>
>
> Cheers, Martin


Yes, right: it wasn't "/massively imported data/" but "/a massive amount of 
data which is being imported/".

Thanks for pointing out.

Sergio



smime.p7s
Description: S/MIME Cryptographic Signature
___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Hotel dataset import? / Re: Baby-sitting

[Martin Koppenhoefer]

sent from a phone

> On 13. Mar 2019, at 16:34, Sergio Manzi  wrote:
> 
> ... an example of massively imported data (which I agree with, btw...): 
> https://blogs.bing.com/maps/2018-06/microsoft-releases-125-million-building-footprints-in-the-us-as-open-data


This data wasn’t generally imported AFAIK (I believe a small part of it was 
imported). Here’s the wiki page: 
https://wiki.openstreetmap.org/wiki/Microsoft_Building_Footprint_Data

here an import:
https://wiki.openstreetmap.org/wiki/Microsoft_Building_Import_-_North_Central_Texas


Cheers, Martin 


___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Hotel dataset import? / Re: Baby-sitting

[Martin Koppenhoefer]

sent from a phone

> On 13. Mar 2019, at 15:44, Sergio Manzi  wrote:
> 
> Have mappers walked along the whole world coastlines? Have they descended all 
> world's rivers by canoe? Have all Mumbay, New York, Rome, Shanghai, etc., 
> streets and alleys been walked by volunteers mapping their location and 
> names? All the peaks escalated?


I don’t know about New York or Shanghai, but in Rome the mappers have indeed 
walked the streets, usually multiple times since 2004. Definitely not based on 
imported data. Neither are rivers or the coastline, these features are derived 
from aerial imagery (not an import).

Ciao, Martin 
___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Hotel dataset import? / Re: Baby-sitting

[Volker Schmidt]

On Wed, 13 Mar 2019 at 15:45, Sergio Manzi  wrote:

> Have mappers walked along the whole world coastlines? Have they descended
> all world's rivers by canoe?
>
No, these are examples where imports make sense as these are feature that
change slowly (normally)

Have all Mumbay, New York, Rome, Shanghai, etc., streets and alleys been
> walked by volunteers mapping their location and names?
>
No, but many have been photographed and the photographs are available with
open license (Mapillary and OpenStreetCam) and many raods have been
recorded with GPX tracks. This allows armchair mapping with high quality
results, where the mapper does not need to be the photographer.
New York on OpenStreetCam
,
Amsterdam
on Mapillary


All the peaks escalated?
>
Again classical data that change slowly, where imports are justified.

But when you start importing trees or buildings or hotels or petrol
stations or shops or crops than you are importing  perishable data which
are maintained outside OSM, and hence need frequent updates, with the
obvious cost that they need to be updated in the external source database
and in the OSM database in synchronism.

Volker

>
>
___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Hotel dataset import? / Re: Baby-sitting

[Sergio Manzi]

On 2019-03-13 15:44, Sergio Manzi wrote:
> On 2019-03-13 15:27, Mateusz Konieczny wrote:
>>
>>
>>
>> Mar 13, 2019, 12:53 PM by s...@smz.it:
>>
>> On 2019-03-13 12:15, Tom Pfeifer wrote:
>>> I think you misunderstand. OSM is based on locally sourced, handcrafted 
>>> data. That creates the high quality.
>>
>> That's totally inaccurate.
>>
>> The reality is that OSM is based on imported data, augmented by locally 
>> sourced information (/of ///sometimes /questionable quality/).
>>
>> Sergio
>>
>> [citation needed]
>>
>> It may be true in some rare regions, in many cases real mapping requires 
>> deletion of substandard
>> imports dumped into OSM (see TIGER mess in USA, low quality landcover in 
>> Slovakia making
>> mapping a horrible experience).
>>
>> Describing OSM as "imported data, augmented by" is as far as I known 
>> complete misrepresenting the
>> situation.
>
> I haven't said that OSM *is* imported data, but that it is *based on* 
> imported data: that's different. And I never said that the "augmentation" 
> contributed by mappers is of irrelevant importance (/even if sometime the 
> quality is sub-par/).
>
> Have mappers walked along the whole world coastlines? Have they descended all 
> world's rivers by canoe? Have all Mumbay, New York, Rome, Shanghai, etc., 
> streets and alleys been walked by volunteers mapping their location and 
> names? All the peaks escalated?
>
> Sergio
>
... an example of massively imported data (/which I agree with, btw.../): 
https://blogs.bing.com/maps/2018-06/microsoft-releases-125-million-building-footprints-in-the-us-as-open-data



smime.p7s
Description: S/MIME Cryptographic Signature
___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Hotel dataset import? / Re: Baby-sitting

[Sergio Manzi]

On 2019-03-13 15:27, Mateusz Konieczny wrote:
>
>
>
> Mar 13, 2019, 12:53 PM by s...@smz.it:
>
> On 2019-03-13 12:15, Tom Pfeifer wrote:
>> I think you misunderstand. OSM is based on locally sourced, handcrafted 
>> data. That creates the high quality.
>
> That's totally inaccurate.
>
> The reality is that OSM is based on imported data, augmented by locally 
> sourced information (/of ///sometimes /questionable quality/).
>
> Sergio
>
> [citation needed]
>
> It may be true in some rare regions, in many cases real mapping requires 
> deletion of substandard
> imports dumped into OSM (see TIGER mess in USA, low quality landcover in 
> Slovakia making
> mapping a horrible experience).
>
> Describing OSM as "imported data, augmented by" is as far as I known complete 
> misrepresenting the
> situation.

I haven't said that OSM *is* imported data, but that it is *based on* imported 
data: that's different. And I never said that the "augmentation" contributed by 
mappers is of irrelevant importance (/even if sometime the quality is sub-par/).

Have mappers walked along the whole world coastlines? Have they descended all 
world's rivers by canoe? Have all Mumbay, New York, Rome, Shanghai, etc., 
streets and alleys been walked by volunteers mapping their location and names? 
All the peaks escalated?

Sergio




smime.p7s
Description: S/MIME Cryptographic Signature
___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Hotel dataset import? / Re: Baby-sitting

[Mateusz Konieczny]

Mar 13, 2019, 2:34 PM by cascaf...@gmail.com:

> > Il giorno mer 13 mar 2019 alle ore 12:16 Tom Pfeifer <> 
> > t.pfei...@computer.org > > ha scritto:
>
> > I think you misunderstand. OSM is based on locally sourced, handcrafted 
> > data. That creates the high quality
> Have youi ever seen hot tasks results? Huge building polys, mixed 
> misalignments, nonxistent tags... lots of poor quality stuff and, worst of 
> all, everything mixed to (few) high quality surveys... but it seems they are 
> useful...of course they are: later you can improve them.
> > Import means data are copied from another database into OSM. That means, 
> > somebody else collected the data. This collector has made his own mistakes. 
> > All databases contain errors, or outdated items, even if they are labelled 
> > official, governmental etc.
>
Extremely low quality of HOT mapping (in part, maybe mostly remote mapping, not 
local surveys)
is other big problem and I would not lump HOT organized mapping with local 
mappers.
___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Hotel dataset import? / Re: Baby-sitting

[Volker Schmidt]

Just my five €cents:
My experience with the massive imports in Veneto is, under many aspects,
negative.
I have tried in vain to stop or revert them.
There are a number of problems, just to list a few basic ones:

   - we import perisahbale data which are at best remotely related to what
   should be in a geographical database
   - we are importing data without verifyiing them
   - we do not have the manpower to maintain imported data
   - and the most basic one is that data imports are in 100% contrast with
   the basic database rule of keeping and updating data only in one place.
   Importing perishable data from other sources into OSM is simply wrong.
   External data should be linked, but never copied.


On Wed, 13 Mar 2019 at 14:59, Sergio Manzi  wrote:

> On 2019-03-13 14:45, Martin Koppenhoefer wrote:
>
>
> Am Mi., 13. März 2019 um 12:54 Uhr schrieb Sergio Manzi :
>
>> On 2019-03-13 12:15, Tom Pfeifer wrote:
>>
>> I think you misunderstand. OSM is based on locally sourced, handcrafted
>> data. That creates the high quality.
>>
>> That's totally inaccurate.
>>
>> The reality is that OSM is based on imported data, augmented by locally
>> sourced information (*of **sometimes questionable quality*).
>>
>
>
> There are examples for this like the US, Friuli Venezia Giulia
> (buildings), maybe France (landcover + cadastre) and the Netherlands, but
> this is still not globally the case, and I would hope it will not get to
> this point. You are relatively new here Sergio and may not know the history
> of OSM, also in Italy where you map, most data was mapped by mappers.
> Notable Italian exceptions may be Friuli Venezia Giulia, administrative
> boundaries by ISTAT, fuel station import last year, some locally confined
> imports in some municipalities.
>
> Cheers,
> Martin
>
> Was this https://www.openstreetmap.org/node/288179924 mapped by Reinhold
> Messner carrying an high accuracy GPS in is backback? As I think he is one
> of the very few who put foot there...
>
> Sergio
> ___
> 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] Hotel dataset import? / Re: Baby-sitting

[Mateusz Konieczny]

Mar 13, 2019, 12:53 PM by s...@smz.it:

> On 2019-03-13 12:15, Tom Pfeifer wrote:
>
>> I  think you misunderstand. OSM is based on locally sourced,  
>> handcrafted data. That creates the high quality. 
>>
>
> That's totally inaccurate.
>
>
> The reality is that OSM is based on imported data, augmented by  locally 
> sourced information (> of > sometimes > questionablequality> ).
>
>
> Sergio
>
>
[citation needed]

It may be true in some rare regions, in many cases real mapping requires 
deletion of substandard
imports dumped into OSM (see TIGER mess in USA, low quality landcover in 
Slovakia making 
mapping a horrible experience).

Describing OSM as "imported data, augmented by" is as far as I known complete 
misrepresenting the
situation.
___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Hotel dataset import? / Re: Baby-sitting

[Sergio Manzi]

On 2019-03-13 14:45, Martin Koppenhoefer wrote:
>
> Am Mi., 13. März 2019 um 12:54 Uhr schrieb Sergio Manzi  >:
>
> On 2019-03-13 12:15, Tom Pfeifer wrote:
>> I think you misunderstand. OSM is based on locally sourced, handcrafted 
>> data. That creates the high quality.
>
> That's totally inaccurate.
>
> The reality is that OSM is based on imported data, augmented by locally 
> sourced information (/of ///sometimes /questionable quality/).
>
>
>
> There are examples for this like the US, Friuli Venezia Giulia (buildings), 
> maybe France (landcover + cadastre) and the Netherlands, but this is still 
> not globally the case, and I would hope it will not get to this point. You 
> are relatively new here Sergio and may not know the history of OSM, also in 
> Italy where you map, most data was mapped by mappers. Notable Italian 
> exceptions may be Friuli Venezia Giulia, administrative boundaries by ISTAT, 
> fuel station import last year, some locally confined imports in some 
> municipalities.
>
> Cheers,
> Martin
>
Was this https://www.openstreetmap.org/node/288179924 mapped by Reinhold 
Messner carrying an high accuracy GPS in is backback? As I think he is one of 
the very few who put foot there...

Sergio



smime.p7s
Description: S/MIME Cryptographic Signature
___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Hotel dataset import? / Re: Baby-sitting

[Martin Koppenhoefer]

Am Mi., 13. März 2019 um 14:36 Uhr schrieb Cascafico Giovanni <
cascaf...@gmail.com>:

> > Il giorno mer 13 mar 2019 alle ore 12:16 Tom Pfeifer <
> t.pfei...@computer.org> ha scritto:
>
> > I think you misunderstand. OSM is based on locally sourced, handcrafted
> data. That creates the high quality
>
> Have youi ever seen hot tasks results? Huge building polys, mixed
> misalignments, nonxistent tags... lots of poor quality stuff and, worst of
> all, everything mixed to (few) high quality surveys... but it seems they
> are useful...of course they are: later you can improve them.
>


if you see issues in other fields, you should raise them (in the end, HOT
tasks aren't imports though, so they follow different requirements). They
cannot be used to justify importing other data.




>
> > Import means data are copied from another database into OSM. That means,
> somebody else collected the data. This collector has made his own mistakes.
> All databases contain errors, or outdated items, even if they are labelled
> official, governmental etc.
>
> The collector, whoever he is, is not error free (he is human). OSM mappers
> are humans.
>


yes, but we have our own type of error, and we have personal responsability
for the things we are entering.

Cheers,
Martin
___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Hotel dataset import? / Re: Baby-sitting

[Andy Townsend]

On 13/03/2019 13:34, Cascafico Giovanni wrote:


Have you ever seen hot tasks results? Huge building polys, mixed 
misalignments, nonxistent tags... lots of poor quality stuff and, 
worst of all, everything mixed to (few) high quality surveys... but it 
seems they are useful...of course they are: later you can improve them.


Just because someone else makes a pigs ear of adding data to OSM doesn't 
mean you should too :)


This sort of thing has been a problem in OSM for a _very_ long time. 
https://blog.gravitystorm.co.uk/2009/11/10/the-pottery-club/ was written 
nearly 10 years ago now!


Best Regards,

Andy



___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Hotel dataset import? / Re: Baby-sitting

[Martin Koppenhoefer]

Am Mi., 13. März 2019 um 12:54 Uhr schrieb Sergio Manzi :

> On 2019-03-13 12:15, Tom Pfeifer wrote:
>
> I think you misunderstand. OSM is based on locally sourced, handcrafted
> data. That creates the high quality.
>
> That's totally inaccurate.
>
> The reality is that OSM is based on imported data, augmented by locally
> sourced information (*of **sometimes questionable quality*).
>


There are examples for this like the US, Friuli Venezia Giulia (buildings),
maybe France (landcover + cadastre) and the Netherlands, but this is still
not globally the case, and I would hope it will not get to this point. You
are relatively new here Sergio and may not know the history of OSM, also in
Italy where you map, most data was mapped by mappers. Notable Italian
exceptions may be Friuli Venezia Giulia, administrative boundaries by
ISTAT, fuel station import last year, some locally confined imports in some
municipalities.

Cheers,
Martin
___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Hotel dataset import? / Re: Baby-sitting

[Cascafico Giovanni]

> Il giorno mer 13 mar 2019 alle ore 12:16 Tom Pfeifer <
t.pfei...@computer.org> ha scritto:

> I think you misunderstand. OSM is based on locally sourced, handcrafted
data. That creates the high quality

Have youi ever seen hot tasks results? Huge building polys, mixed
misalignments, nonxistent tags... lots of poor quality stuff and, worst of
all, everything mixed to (few) high quality surveys... but it seems they
are useful...of course they are: later you can improve them.

> Import means data are copied from another database into OSM. That means,
somebody else collected the data. This collector has made his own mistakes.
All databases contain errors, or outdated items, even if they are labelled
official, governmental etc.

The collector, whoever he is, is not error free (he is human). OSM mappers
are humans.

The 53 POIs were not simply copied: they were geocoded, (possibly, tag
mapped (if this ML gather some verdicts), location checked, conflated. If
you write "Import means data are copied from another database into OSM",
you are saying mine is not an import. Hence I think we lack a proper,
better import definition, rather then the extremely generic version updated
yesterday.

> IIRC the dataset you are using is several years old, last updated 2017.
The items in your Umap do
> not say when the database entry was last checked.
> Thus when copying these data into OSM, they appear as freshly updated,
March 2019, although the
> information is much older and the hotel might have closed or changed
ownership/pet policies/wifi in
> 2018.

Again, "freshly" word is subjectiive. Is a 2 y/o address "fresh"? Is a 3 mo
old payment method "fresh"?Probably you are right, there could be some
hotel diisused, less likely dismantled, several may have canceled visa
payments,,,but I think that no data in worse.

In sept 2015, in the same region, an import was performed involving >400k
address nodes. Same hotel source, same method: dataset built collecting
gathered municipality surveys; data published in may 2014, survey dates
unknown. Until today nobody complained about that import, even if some
addresses were void, even if some we discovered be replaced. AFAIK those
addresses are anyway in use and are a good base for geocoding: should we
had abort that import?

> Thus your task is to discuss with the Italian community if your import,
i.e. copying the other data,
> improves or decreases the OSM data quality. The changeset size is less
important, though I would
> prefer an community-approved import to be traceable in not too many
changesets.

Due to the nature of the umap editing, this "import" couldn't be performed
in one shot. neither uploaded by a dedicated account. What I was aiming,
when I started asking suggestions n this ML, was to set a tagging plan for
a commnity based, umap supported "import". If this doesn't fit anymore the
official import definition, let's forget about it.

Sincerely, thank for your feedback.
___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Hotel dataset import? / Re: Baby-sitting

[Sergio Manzi]

On 2019-03-13 12:15, Tom Pfeifer wrote:
> I think you misunderstand. OSM is based on locally sourced, handcrafted data. 
> That creates the high quality.

That's totally inaccurate.

The reality is that OSM is based on imported data, augmented by locally sourced 
information (/of ///sometimes /questionable quality/).

Sergio



smime.p7s
Description: S/MIME Cryptographic Signature
___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Hotel dataset import? / Re: Baby-sitting

[marc marc]

Hello,

Le 13.03.19 à 10:50, Cascafico Giovanni a écrit :
> - since there is one (1) ML unaswered doubt if "01 Jan - 31 Dec" should 
> be replaced by 24/7: import is anyway bad

I find that they are 2 totally different things: one informs that the 
poi is open all year round, the other all day. for the majority of poi, 
you can't deduct from each other (a gas station can be open all year 
round but only from 8:00 am to 8:00 pm because located for example
in a shopping centre that closes access during the night.)

> - pets and childcare information are "volatile" (payment methods aren't? 
> and yet tag exists): import is anyway bad

from what I see, a common cause of import failure is when the proposer 
refuses to take into account the local community concerned. if the 
latter considers that the information of pets dating from 2 years ago is 
too volatile, the best way to achieve failure is to persist in wanting 
to import them with everything else. the best way to progress is to 
propose a first step without this data (split out data that provoke 
criticism), it is always possible to propose an addition (divide
to avoid "all or nothing").

Regards,
Marc
___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Hotel dataset import? / Re: Baby-sitting

[Tom Pfeifer]

On 13.03.2019 10:50, Cascafico Giovanni wrote:

The lattest is great and leads to grotesque consequences: when a colleague hands me 12 guidepost 
locations to geocode, I must message planet import ML, local ML, write an import page and ask my 
fellow a consent form with certified signature. Is this what we aim?
I think setting such a binding requirements on small size "imports" is putting a gravestone on many 
potential small, local sources.


I think you misunderstand. OSM is based on locally sourced, handcrafted data. That creates the high 
quality.


Import means data are copied from another database into OSM. That means, somebody else collected the 
data. This collector has made his own mistakes. All databases contain errors, or outdated items, 
even if they are labelled official, governmental etc.


IIRC the dataset you are using is several years old, last updated 2017. The items in your Umap do 
not say when the database entry was last checked.


Thus when copying these data into OSM, they appear as freshly updated, March 2019, although the 
information is much older and the hotel might have closed or changed ownership/pet policies/wifi in 
2018. In particular in a "stealth" import, one node at a time, I would assume the OSM data to be 
individually checked and current.


Thus your task is to discuss with the Italian community if your import, i.e. copying the other data, 
improves or decreases the OSM data quality. The changeset size is less important, though I would 
prefer an community-approved import to be traceable in not too many changesets.


Your colleagues data of 12 guideposts are fine if they are freshly collected 
from the ground.

Hope that helps to understand why we are careful with imports.

tom

___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Hotel dataset import? / Re: Baby-sitting

[Cascafico Giovanni]

>
> Il giorno lun 11 mar 2019 alle ore 14:11 Mateusz Konieczny <
> matkoni...@tutanota.com> ha scritto:
> > Fixed in
> https://wiki.openstreetmap.org/w/index.php?title=Import=1820348=1551233
>
>
I took a 600+ dataset, geocoded them, manually conflate 53 of them, put
source references on umap, put source institution name on changeset, ask
for some tags clarification (yet didn't receive an agreed feedback).
Am I missing something?

The only (sequential) results of this discussion are:

- I declare <100 nodes but Volker noticed source dataset has 600+, but then:
- I'm asked where the <100 was mentioned (cannot remember), conclusion: I'm
malicious: it's an import!
- since there is one (1) ML unaswered doubt if "01 Jan - 31 Dec" should be
replaced by 24/7: import is anyway bad
- pets and childcare information are "volatile" (payment methods aren't?
and yet tag exists): import is anyway bad
- "bulk" word removed from main import wiki [1]: hence it's obviously an
import

The lattest is great and leads to grotesque consequences: when a colleague
hands me 12 guidepost locations to geocode, I must message planet import
ML, local ML, write an import page and ask my fellow a consent form with
certified signature. Is this what we aim?
I think setting such a binding requirements on small size "imports" is
putting a gravestone on many potential small, local sources.

Please, forgive my sarcasm, but I'm really discouraged. I published the 53
changeset id's [2]. Feel free to revert.
If anybody wants to help me, please feed back on the import doc I wrote
on-the-fly [3]

[1]
https://wiki.openstreetmap.org/w/index.php?title=Import=revision=1820348=1551233
[2] https://ethercalc.org/hpil8syy4rgz.csv
[3] https://wiki.openstreetmap.org/wiki/Import/Catalogue/Hotels-RAFVG-umap
___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Hotel dataset import? / Re: Baby-sitting

[Cascafico Giovanni]

Il giorno lun 11 mar 2019 alle ore 12:20 Cascafico Giovanni <
cascaf...@gmail.com> ha scritto:

> > You were criticized for stretching the opening_hours syntax to describe
> seasonal operations ("Jan 01
> > - Dec 31"), but did not respond nor adjust your tagging.
>
> Sorry for that. Is it wrong? I did test it with JOSM OpeningHoursEditor
> [2] plugin. If anybopdy in this ML prefers 24/7 I will change the values. I
> used "Jan 01 - Dec 31" just to set a value consistent to overall hotel
> dataset opening periods.
>



> > Your import focuses on soft business policies, such as allowing pets or
> supervising kids. Such
> > policies can change even more rapidly, and are better shown in separate
> datasets and not OSM itself.
>
> This consideration raises a huge question about italian fuel stations,
> since operators, brands and names come and go.
> In case of a future regular import, I'll remove pets and childcare,
> inserting just phisical elements like gymn, garage, air-conditioning,
> welness etc.
>

I'm preparing in import wiki. Any thought about the above points?
___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Hotel dataset import? / Re: Baby-sitting

[Mateusz Konieczny]

Mar 11, 2019, 12:20 PM by cascaf...@gmail.com:

> Il giorno sab 9 mar 2019 alle ore 22:15 Tom Pfeifer <> t.pfei...@computer.org 
> > > ha scritto:
>
> > I have severe problems with your process. First, yes it is an import. You 
> > called it an import
> > yourself ("manually importing") here, and on the Italien list where you 
> > first asked about the
> > tagging.
> > Where did you see a "100 nodes" limit documented? You are copying from one 
> > database
> > into another, and if you do just one node a day, it is a slow import.
>
> As stated in the main wiki page [1], importing is "(also known as Bulk 
> Importing)". 
>
Fixed in 
https://wiki.openstreetmap.org/w/index.php?title=Import=1820348=1551233

___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Hotel dataset import? / Re: Baby-sitting

[marc marc]

Le 11.03.19 à 12:20, Cascafico Giovanni a écrit :
> limiting by myself the number not to trigger it as an import.
> 
> If anybody in this ML considers that this activity is different from 
> getting informations from hotel websites and putting them in OSM, I will 
> immediately abort the insertions.

splitting a import into several changeset to "be under the radar" is 
still an import and is indeed not the same as getting info from the poi 
it-self
___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Hotel dataset import? / Re: Baby-sitting

[Cascafico Giovanni]

Il giorno sab 9 mar 2019 alle ore 22:15 Tom Pfeifer 
ha scritto:

> I have severe problems with your process. First, yes it is an import. You
called it an import
> yourself ("manually importing") here, and on the Italien list where you
first asked about the
> tagging.
> Where did you see a "100 nodes" limit documented? You are copying from
one database
> into another, and if you do just one node a day, it is a slow import.

As stated in the main wiki page [1], importing is "(also known as Bulk
Importing)". Maybe I misunderstood this wider acceptation when I started a
manual insertion of POIs, limiting by myself the number not to trigger it
as an import.

If anybody in this ML considers that this activity is different from
getting informations from hotel websites and putting them in OSM, I will
immediately abort the insertions.

> You were criticized for stretching the opening_hours syntax to describe
seasonal operations ("Jan 01
> - Dec 31"), but did not respond nor adjust your tagging.

Sorry for that. Is it wrong? I did test it with JOSM OpeningHoursEditor [2]
plugin. If anybopdy in this ML prefers 24/7 I will change the values. I
used "Jan 01 - Dec 31" just to set a value consistent to overall hotel
dataset opening periods.

> Your import focuses on soft business policies, such as allowing pets or
supervising kids. Such
> policies can change even more rapidly, and are better shown in separate
datasets and not OSM itself.

This consideration raises a huge question about italian fuel stations,
since operators, brands and names come and go.
In case of a future regular import, I'll remove pets and childcare,
inserting just phisical elements like gymn, garage, air-conditioning,
welness etc.

> Your import does not include any check, how current or old the data in
the imported set are. In the
> hotel business, things can change very fast. Hotels open and close, and
change ownership and
> operations.>

In umap, dataset layer you can read  RAFVG opendata, ottobre [october]
2017. Anyway terms like "very fast" are subjective: some time ago I aborted
a draft for Bed& Breakfast import bacause surveys detected some business
shut downs. So who's in charge of the "very" fast evaluations? If we demand
that every entry in possible source dataaset is right what do you think
will be the situation in OSM today?

> The link on your Umap site leads to [5], which is licensed under the
Italian Open Data License,
> linked here [6]. It requires attribution, machine-translated: "On
condition of: indicate the source
> of the Information and the name of the Licensor, including, if possible,
a copy of this license or a
> link (link) to it."
> You have not attributed correctly. You changesets, e.g. [7], give in the
CS comment "RAFVG source",
> which is an incomprehensible acronym if you don't know the context. The
CS has no source tag at all
> (although the editor you use has a mechanism for it),

You are right, but I can't find that mechanism in Level0

>thus you do not name the source correctly, you
> do not name the Licensor, and you do not include a link although
possible. You have also not checked
> if the attribution on the changeset only would be sufficient.> You use
and advertise in your umap the use of Level0 as an editor. This tool is
excellent for
> quickly fixing a tag, but I would find it error-prone to upload mass
changes without a validation step.
> Thus I conclude: Visualising the dataset in your Umap approach [1] is an
excellent idea, unreviewed
> copying of the data into OSM is not.

Anyway, If you consider the activity is definitely an import, I'll revert
involved changesets,  taking in account  the above disapproval points.


[1] https://wiki.openstreetmap.org/wiki/Import
[2] https://wiki.openstreetmap.org/wiki/JOSM/Plugins/OpeningHoursEditor
___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Hotel dataset import? / Re: Baby-sitting

[Tom Pfeifer]

Giovanni,

I have severe problems with your process. First, yes it is an import. You called it an import 
yourself ("manually importing") here, and on the Italien list where you first asked about the 
tagging [2]. Where did you see a "100 nodes" limit documented? You are copying from one database 
into another, and if you do just one node a day, it is a slow import.


Imports require the Import Guidelines to be followed [4]. I cannot see any discussion with the 
community. I cannot see any check of license compatibility. There is no documentation. There is no 
entry in the import catalogue.


You were criticized for stretching the opening_hours syntax to describe seasonal operations ("Jan 01 
- Dec 31"), but did not respond nor adjust your tagging.[3]


The link on your Umap site leads to [5], which is licensed under the Italian Open Data License, 
linked here [6]. It requires attribution, machine-translated: "On condition of: indicate the source 
of the Information and the name of the Licensor, including, if possible, a copy of this license or a 
link (link) to it."


You have not attributed correctly. You changesets, e.g. [7], give in the CS comment "RAFVG source", 
which is an incomprehensible acronym if you don't know the context. The CS has no source tag at all 
(although the editor you use has a mechanism for it), thus you do not name the source correctly, you 
do not name the Licensor, and you do not include a link although possible. You have also not checked 
if the attribution on the changeset only would be sufficient.


Your import does not include any check, how current or old the data in the imported set are. In the 
hotel business, things can change very fast. Hotels open and close, and change ownership and 
operations.


Your import focuses on soft business policies, such as allowing pets or supervising kids. Such 
policies can change even more rapidly, and are better shown in separate datasets and not OSM itself.


You use and advertise in your umap the use of Level0 as an editor. This tool is excellent for 
quickly fixing a tag, but I would find it error-prone to upload mass changes without a validation step.


Thus I conclude: Visualising the dataset in your Umap approach [1] is an excellent idea, unreviewed 
copying of the data into OSM is not.


[2] https://lists.openstreetmap.org/pipermail/talk-it/2019-February/065837.html
[3] https://lists.openstreetmap.org/pipermail/talk-it/2019-February/065839.html
[4] https://wiki.openstreetmap.org/wiki/Import/Guidelines
[5] 
https://www.dati.friuliveneziagiulia.it/Istruzione-cultura-e-sport/Strutture-Alberghiere-e-RTA/fiiw-i5su

[6] https://www.dati.gov.it/content/italian-open-data-license-v20
[7] https://www.openstreetmap.org/node/2088900846/history

Tom

On 09.03.2019 12:19, Cascafico Giovanni wrote:

[1] http://umap.openstreetmap.fr/it/map/turismo-fvg_295722

Il sab 9 mar 2019, 12:18 Cascafico Giovanni mailto:cascaf...@gmail.com>> ha 
scritto:


In short (I'm on phone)...
Please find here [1] umap used for manual conflation. For source data and 
license, follow footer
info link.

AFAIK <100 nodes don't fall in import category.

Il sab 9 mar 2019, 11:46 Tom Pfeifer mailto:t.pfei...@computer.org>> ha
scritto:

On 09.03.2019 10:10, Cascafico Giovanni wrote:
 > Well, hotels dataset I'm manually importing has a boolean 
baby-sitting field (as for
"pets" in other
 > ML thread).
 > I think that a generic info is better than no info, particularly if 
hotel features
childcare=yes and
 > whatever contact tag.

Giovanni,

can you please let us know, which hotels dataset you are importing,
under what license this dataset is published,
and where you discussed the import with the OSM community?

Please not that an import is an import even if you do just one hotel 
per day.

tom


___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Hotel dataset import? / Re: Baby-sitting

[Volker Schmidt]

AFAIK <100 nodes don't fall in import category.
>>
>
I see 586 blue elements on the umap, or am I missing something?
___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Hotel dataset import? / Re: Baby-sitting

[Cascafico Giovanni]

[1] http://umap.openstreetmap.fr/it/map/turismo-fvg_295722

Il sab 9 mar 2019, 12:18 Cascafico Giovanni  ha
scritto:

> In short (I'm on phone)...
> Please find here [1] umap used for manual conflation. For source data and
> license, follow footer info link.
>
> AFAIK <100 nodes don't fall in import category.
>
> Il sab 9 mar 2019, 11:46 Tom Pfeifer  ha scritto:
>
>> On 09.03.2019 10:10, Cascafico Giovanni wrote:
>> > Well, hotels dataset I'm manually importing has a boolean baby-sitting
>> field (as for "pets" in other
>> > ML thread).
>> > I think that a generic info is better than no info, particularly if
>> hotel features childcare=yes and
>> > whatever contact tag.
>>
>> Giovanni,
>>
>> can you please let us know, which hotels dataset you are importing,
>> under what license this dataset is published,
>> and where you discussed the import with the OSM community?
>>
>> Please not that an import is an import even if you do just one hotel per
>> day.
>>
>> tom
>>
>> ___
>> 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] Hotel dataset import? / Re: Baby-sitting

[Cascafico Giovanni]

In short (I'm on phone)...
Please find here [1] umap used for manual conflation. For source data and
license, follow footer info link.

AFAIK <100 nodes don't fall in import category.

Il sab 9 mar 2019, 11:46 Tom Pfeifer  ha scritto:

> On 09.03.2019 10:10, Cascafico Giovanni wrote:
> > Well, hotels dataset I'm manually importing has a boolean baby-sitting
> field (as for "pets" in other
> > ML thread).
> > I think that a generic info is better than no info, particularly if
> hotel features childcare=yes and
> > whatever contact tag.
>
> Giovanni,
>
> can you please let us know, which hotels dataset you are importing,
> under what license this dataset is published,
> and where you discussed the import with the OSM community?
>
> Please not that an import is an import even if you do just one hotel per
> day.
>
> tom
>
> ___
> Tagging mailing list
> Tagging@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/tagging
>
___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging