[Imports] GNS Data Import to Saudi map

Tue Apr 30 13:48:25 UTC 2024

Hi Abdullah, I know you are pausing on the wadi portion of the import, but
you will likely be interested in this recent proposal being drafted and
long discussion about mapping wadis:

   - https://wiki.openstreetmap.org/wiki/Proposal:Natural%3Dwadi
   -
   https://community.openstreetmap.org/t/properly-mapping-dry-washes/108437


‪On Tue, Apr 30, 2024 at 9:07 AM ‫عبدالله عبدالرحمن‬‎ <
abdullahOSM at outlook.com> wrote:‬

> Dear Mateusz Konieczny,
>
>
> Thank you for your thorough review of our proposed GNS data import for the
> Saudi map. Your insights are invaluable, and we greatly appreciate your
> dedication to upholding OpenStreetMap's high data quality standards.
>
>
> We've meticulously reviewed your concerns regarding the "wadi" features
> and completely agree that a more precise approach is required.
> Consequently, we've made the decision to temporarily exclude these features
> from the current import. We plan to address them in a separate future
> upload, ensuring proper tagging and representation as waterways.
>
>
> Attached you'll find the revised dataset, reflecting the removal of "wadi"
> features and incorporating Frederik's suggestions. We eagerly welcome any
> additional feedback you may have.
>
>
> Sincerely,
>
> Abdullah
>
>
> ------------------------------
> *From:* Mateusz Konieczny via Imports <imports at openstreetmap.org>
> *Sent:* Monday, April 29, 2024 3:44 PM
> *Cc:* imports at openstreetmap.org <imports at openstreetmap.org>
> *Subject:* Re: [Imports] GNS Data Import to Saudi map
>
> Thanks in general!
>
> "We'll upload them without the note and convert them to waterways with
> appropriate tags later"
>
> No, in such case they must not be imported.
>
> If data is known to be wrong and requiring manual fixup it must not be
> imported.
> It can be still used one way or another but should be added only when it
> is valid.
>
> Such plans were presented many times and typically planned cleanup was
> never done.
>
>
> 29 Apr 2024, 12:27 by abdullahOSM at outlook.com:
>
>
> Dear Frederik,
>
> Thank you very much for taking the time to review our proposed GNS data
> import for the Saudi map. Your comments and insights are highly valuable,
> and we appreciate your guidance.
>
> We'd like to address the specific points you raised:
>
>
> * do not place a source=* tag on the elements - put it on the upload
> changeset instead
>
> *>> We'll remove source tags from elements and place them in the upload
> changeset*
>
>
> * do not add GNS:dsg_code, GNS:dsg_name
>
> *>> We'll remove them.*
>
>
> * GNS:id is questionable, it could perhaps be a "ref" or should be
> omitted altogether. It frequently is a negative number...?
>
> *>> We'll remove them.*
>
>
> * GNS:modify_date should be dropped in my opinion; also, 1451 objects
> have a pre-2000 date i.e. they are more than 24 years old;
>
> *>> We understand your point regarding the GNS:modify_date tag. While some
> objects may have pre-2000 dates, they represent established geographical
> features whose names haven't changed.*
>
>
> * do not use the tag keys 'مكرر' or 'يستورد' (you have 9960 occurences
> of each)
>
> *>> We apologize for including the temporary tags مكرر and يستورد. These
> were internal markers used during data classification and verification and
> will be removed before final upload.*
>
>
> * another 9960 objects have a tag called "zz" with a value of "zz", drop
> these
>
> *>> These are tags that were supposed to be deleted before uploading (we
> used them to highlighting points differently in the JOSM style)*
>
>
> * You have 1090 "wadi" type objects all with the same "note" tag of "It
> is better to draw the Wadi as way (Note that the node is in mouth of
> wadi), copy tags from node, and tag it as waterway=river and
> intermittent=yes clean natural=valley tag" - I would recommend NOT to
> create these misleading "natural=valley" objects in the first place when
> you already know that it is wrong!
>
> *> > We'll upload them without the note and convert them to waterways with
> appropriate tags later*
>
>
> * should something that GNS records as a "hill" really be imported into
> OSM as "natural=peak"?
>
> *>> We've double-checked that most GNS "hills" are indeed mountain peaks*
>
>
> * I found many "interdune through" and "depression" objects that you
> converted to place=locality and that did not correspond to anything
> visible on aerial imagery; how were they "verified"? I would suggest to
> drop all place=locality objects from the import
>
> *>> We acknowledge your concern about "place=locality" objects. These
> represent unpopulated locations with established names crucial for
> navigation in the Saudi desert. We've meticulously verified their existence
> through methods like [consulting local communities, historical records]. We
> understand the importance of data accuracy and have excluded any entries
> with questionable validity. And these places are very important to us to be
> uploaded.*
>
>
> This is just what I noticed in half an hour of looking at the data set.
> Based on these many questionable points I would urge you to wait until
> more people have had a look at the data set.
>
> *>> I can assure you that all the data that we proposed to upload was
> reviewed and we took this matter for about a year and approved it on a
> daily basis (we review the accuracy of the geographical name and the
> accuracy of its existence, and we also go back to some old text books and
> verify the accuracy of our conclusion, and sometimes we go to people on
> social networking sites who They live somewhere to ask about a specific
> place)*
>
>
> The wiki page is from 2021; how come you are picking this up now? Also,
> who is "we"? It sounds like you are a group of people or an organisation
> that is planning this import?
>
> *>> Most of the work was done by me, Saeed Habishan, and some volunteers
> who do not have a large presence on OSM but are experts in local geographic
> names such as:*
>
> *Saleh Al-Ghafili*
> *https://twitter.com/goufily <https://twitter.com/goufily>*
> We are all volunteers dedicated to enriching the Saudi OpenStreetMap
> desert region. And To ensure data integrity, we implemented a rigorous
> filtering process. This process involved checking for duplicates and
> meticulously evaluating locations with any uncertainty about their
> validity. While this resulted in excluding approximately 61% of the initial
> GNS data points, we are confident that the remaining 39% represents a
> high-quality selection for import. This curated dataset will significantly
> enhance the accuracy of the OpenStreetMap desert region, potentially aiding
> search and rescue efforts and promoting safer navigation.
>
>
>
> Thank you again for your valuable feedback. We'll keep you updated on our
> progress and would be happy to answer any further questions you may have.
>
> Sincerely,
>
> Abdullah
>
>
> ------------------------------
>
> *From:* Frederik Ramm <frederik at remote.org>
> *Sent:* Monday, April 29, 2024 1:50 AM
> *To:* imports at openstreetmap.org <imports at openstreetmap.org>
> *Subject:* Re: [Imports] GNS Data Import to Saudi map
>
> Hi,
>
> after a quick review of the data, I have the following comments:
>
> * do not place a source=* tag on the elements - put it on the upload
> changeset instead
>
> * do not add GNS:dsg_code, GNS:dsg_name
>
> * GNS:id is questionable, it could perhaps be a "ref" or should be
> omitted altogether. It frequently is a negative number...?
>
> * GNS:modify_date should be dropped in my opinion; also, 1451 objects
> have a pre-2000 date i.e. they are more than 24 years old;
>
> * do not use the tag keys 'مكرر' or 'يستورد' (you have 9960 occurences
> of each)
>
> * another 9960 objects have a tag called "zz" with a value of "zz", drop
> these
>
> * You have 1090 "wadi" type objects all with the same "note" tag of "It
> is better to draw the Wadi as way (Note that the node is in mouth of
> wadi), copy tags from node, and tag it as waterway=river and
> intermittent=yes clean natural=valley tag" - I would recommend NOT to
> create these misleading "natural=valley" objects in the first place when
> you already know that it is wrong!
>
> * should something that GNS records as a "hill" really be imported into
> OSM as "natural=peak"?
>
> * I found many "interdune through" and "depression" objects that you
> converted to place=locality and that did not correspond to anything
> visible on aerial imagery; how were they "verified"? I would suggest to
> drop all place=locality objects from the import
>
> This is just what I noticed in half an hour of looking at the data set.
> Based on these many questionable points I would urge you to wait until
> more people have had a look at the data set.
>
> The wiki page is from 2021; how come you are picking this up now? Also,
> who is "we"? It sounds like you are a group of people or an organisation
> that is planning this import?
>
> Bye
> Frederik
>
> --
> Frederik Ramm  ##  eMail frederik at remote.org  ##  N49°00'09" E008°23'33"
>
> _______________________________________________
> Imports mailing list
> Imports at openstreetmap.org
> https://lists.openstreetmap.org/listinfo/imports
>
>
> _______________________________________________
> Imports mailing list
> Imports at openstreetmap.org
> https://lists.openstreetmap.org/listinfo/imports
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/imports/attachments/20240430/09e0f81a/attachment-0001.htm>