<html>
  <head>
    <meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
  </head>
  <body text="#000000" bgcolor="#FFFFFF">
    <p>Dear fellow-mappers,</p>
    <p>In The Netherlands we went through this process 4 years ago. This
      means we can talk from experience.</p>
    <p>After the preparation period, we started the real import process
      around march 2014 with a group of some 30 volunteers. At the end
      of 2014, the initial import was finished. At the time, we decided
      to import the building id's, but not the address id's. The address
      id's we not available in the WFS data source we used and we
      reasoned that the addresses can be identified by the address tags.<br>
    </p>
    <p>Since then, we are in the process of maintaining the imported
      buildings and address. This concerns around 11.000 new buildings
      and the same number of new addresses every month.<br>
      The tooling we use in this process depends highly on the building
      id's. Not having have the address id's seriously complicates the
      maintenance of addresses, Even though the combination of postcode
      and house number are unique in The Netherlands. Being official
      government id's, the building and address id's in The Netherlands
      get more and more official applications day by day. They are
      already required in notarial acts and used by the chamber of
      commerce. Leaving them out would not be an option in my opinion.</p>
    <p>We have no issue with mappers being afraid to touch buildings
      because of the building id's. Most people know where to find the
      id, or file an import request on the forum for a building or a
      block of buildings. Some complex buildings like Utrecht Central
      station keep being mapped manually because the official data is
      too complex.<br>
      The main issue we have is with addresses being deleted, because
      people delete a node instead of the poi tags. This is nothing new,
      but it is detected more often because of the continuous comparison
      between OSM and official building and address data.</p>
    <p>Gertjan Idema<br>
    </p>
    <p><br>
    </p>
    <div class="moz-cite-prefix">On 01/11/2018 15:39, Pieter Vander
      Vennet wrote:<br>
    </div>
    <blockquote type="cite"
cite="mid:CAM-CmFA5X3p-Ej1FeiWSTrsRjoA7mXwD99Gf2_wwM1rED=U+-g@mail.gmail.com">
      <meta http-equiv="content-type" content="text/html; charset=UTF-8">
      <div dir="ltr">
        <div id="gmail-magicdomid3" class="gmail-ace-line"><span
            class="gmail-author-a-wz87zz90zz66zz82zt3itz84z5r8z80z4z69z">Hello
            everyone</span><span
            class="gmail-author-a-dz78zz77zz84z80jwz67z68t65f7">,</span></div>
        <div class="gmail-ace-line"><span
            class="gmail-author-a-dz78zz77zz84z80jwz67z68t65f7">Original
            Poster here.<br>
          </span></div>
        <div id="gmail-magicdomid4" class="gmail-ace-line"><br>
        </div>
        <div id="gmail-magicdomid5" class="gmail-ace-line"><span
            class="gmail-author-a-wz87zz90zz66zz82zt3itz84z5r8z80z4z69z">Thanks
            for your remarks. A lot of people</span><span
            class="gmail-author-a-dz78zz77zz84z80jwz67z68t65f7"> active</span><span
            class="gmail-author-a-wz87zz90zz66zz82zt3itz84z5r8z80z4z69z">
            in the Belgian community are following this discussion</span><span
class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99"> with
            interest and some bewilderment</span><span
            class="gmail-author-a-wz87zz90zz66zz82zt3itz84z5r8z80z4z69z">.</span><span
class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99"> Like
            the first message, this is a reply that has been written by
            several </span><span
            class="gmail-author-a-dz78zz77zz84z80jwz67z68t65f7">members</span><span
class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99">
            together. </span></div>
        <div id="gmail-magicdomid6" class="gmail-ace-line"><span
            class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99">We
            did have a hard time filtering the actual discussion about
            our import from the thread. We would kindly invite everyone
            to take the discussion of other imports and the more philos</span><span
            class="gmail-author-a-dz78zz77zz84z80jwz67z68t65f7">o</span><span
class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99">phical
            points  to a more appropriate place (e.g. the talk-list) and
            stick to the topic of the Flemish bu</span><span
            class="gmail-author-a-dz78zz77zz84z80jwz67z68t65f7">i</span><span
class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99">ldings
            import.</span></div>
        <div id="gmail-magicdomid7" class="gmail-ace-line"><br>
        </div>
        <div id="gmail-magicdomid8" class="gmail-ace-line"><span
            class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99">We
            have been working on this for two years and don't mind
            working on it for a few more </span><span
            class="gmail-author-a-dz78zz77zz84z80jwz67z68t65f7">months,
            before running the actual integration</span><span
            class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99">.
            We</span><span
            class="gmail-author-a-dz78zz77zz84z80jwz67z68t65f7"> are, in
            the first place, looking for advice and guidance so as to
            further improve our methods - together with with</span><span
class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99"> a
            go-ahead </span><span
            class="gmail-author-a-dz78zz77zz84z80jwz67z68t65f7">once all
            issues are resolved.</span><span
            class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99"> </span></div>
        <div id="gmail-magicdomid9" class="gmail-ace-line"><br>
        </div>
        <div id="gmail-magicdomid10" class="gmail-ace-line"><span
            class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99">In
            case this was not clear befor</span><span
            class="gmail-author-a-wz87zz90zz66zz82zt3itz84z5r8z80z4z69z">e:</span><span
class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99"> 
            this is _not_ a flat import where all the data is dumped
            into OSM automatically. This is an effort of the Belgian
            community, where mappers select a subset of the data they
            choose to integrate piecemeal, usually one street at a time.
            The data is loaded in JOSM and checked against the aerial
            imagery involving the mapper's common sense.</span></div>
        <div id="gmail-magicdomid11" class="gmail-ace-line"><span
            class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99">As
            an example, this is a screenshot of the tool in action: </span><span
class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99
            gmail-url"><a
href="https://matrix.org/_matrix/media/v1/download/matrix.org/WmemnfTQZOSfoKoIlByFBHmv"
              moz-do-not-send="true">https://matrix.org/_matrix/media/v1/download/matrix.org/WmemnfTQZOSfoKoIlByFBHmv</a></span><span
class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99"> .
            You can see that the tool reuses tags from OSM and is </span><span
class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99
            gmail-comment gmail-c-byY8GXqDvvFz0dvd">offering the mapper
            the choice</span><span
            class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99">
            which geometry should be used</span><span
            class="gmail-author-a-dz78zz77zz84z80jwz67z68t65f7"> (by not
            saving the import)</span><span
            class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99">.
            In combination with aerial imagery, this should cause little
            to no problems</span><span
            class="gmail-author-a-dz78zz77zz84z80jwz67z68t65f7"> in
            regard with the original OSM-data</span><span
            class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99">.</span></div>
        <div id="gmail-magicdomid12" class="gmail-ace-line"><br>
        </div>
        <div id="gmail-magicdomid13" class="gmail-ace-line"><span
            class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99">Mateusz
            and Frederik, your points regarding documentation are being
            addressed as we speak. </span></div>
        <div id="gmail-magicdomid15" class="gmail-ace-line"><br>
        </div>
        <div id="gmail-magicdomid16" class="gmail-ace-line"><span
            class="gmail-author-a-dz78zz77zz84z80jwz67z68t65f7">#
            External IDs</span></div>
        <div id="gmail-magicdomid17" class="gmail-ace-line"><br>
        </div>
        <div id="gmail-magicdomid297" class="gmail-ace-line"><span
            class="gmail-author-a-wz87zz90zz66zz82zt3itz84z5r8z80z4z69z">The
            most discussed point seems to be the IDs we wanted to
            include. </span></div>
        <div id="gmail-magicdomid19" class="gmail-ace-line"><span
            class="gmail-author-a-dz78zz77zz84z80jwz67z68t65f7">We
            believe the</span><span
            class="gmail-author-a-wz87zz90zz66zz82zt3itz84z5r8z80z4z69z">
            IDs</span><span
            class="gmail-author-a-dz78zz77zz84z80jwz67z68t65f7"> will
            significantly ease updates to our buildings when the GRB is
            updated</span><span
            class="gmail-author-a-wz87zz90zz66zz82zt3itz84z5r8z80z4z69z">,</span><span
            class="gmail-author-a-dz78zz77zz84z80jwz67z68t65f7"> and
            make the whole process more robust.</span></div>
        <div id="gmail-magicdomid20" class="gmail-ace-line"><span
            class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99
            gmail-comment gmail-c-W8FTpN63EZVvaZMX">They</span><span
            class="gmail-author-a-dz78zz77zz84z80jwz67z68t65f7
            gmail-comment gmail-c-W8FTpN63EZVvaZMX"> provide a way to
            update and cross</span><span
            class="gmail-author-a-z80ziz85za1pez66zo0z71zz72zjz66zz65zz78z
            gmail-comment gmail-c-W8FTpN63EZVvaZMX"> </span><span
            class="gmail-author-a-dz78zz77zz84z80jwz67z68t65f7
            gmail-comment gmail-c-W8FTpN63EZVvaZMX">reference the data </span><span
class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99
            gmail-comment gmail-c-W8FTpN63EZVvaZMX">now and in the </span><span
            class="gmail-author-a-dz78zz77zz84z80jwz67z68t65f7
            gmail-comment gmail-c-W8FTpN63EZVvaZMX">future.</span><span
            class="gmail-author-a-dz78zz77zz84z80jwz67z68t65f7"> Exactly
            by keeping th</span><span
            class="gmail-author-a-z80ziz85za1pez66zo0z71zz72zjz66zz65zz78z">e</span><span
            class="gmail-author-a-dz78zz77zz84z80jwz67z68t65f7">s</span><span
class="gmail-author-a-z80ziz85za1pez66zo0z71zz72zjz66zz65zz78z">e</span><span
            class="gmail-author-a-dz78zz77zz84z80jwz67z68t65f7"> IDs,
            updating data down the road will be smoother and prevent
            later changes </span><span
            class="gmail-author-a-z80ziz85za1pez66zo0z71zz72zjz66zz65zz78z">from
            OSM contributors </span><span
            class="gmail-author-a-dz78zz77zz84z80jwz67z68t65f7">to be
            overwritten. As </span><span
            class="gmail-author-a-wz87zz90zz66zz82zt3itz84z5r8z80z4z69z">an
          </span><span
            class="gmail-author-a-dz78zz77zz84z80jwz67z68t65f7">added
            benefit, keeping ID</span><span
            class="gmail-author-a-z80ziz85za1pez66zo0z71zz72zjz66zz65zz78z">s</span><span
            class="gmail-author-a-dz78zz77zz84z80jwz67z68t65f7"> makes
            the import tool-agnostic</span><span
            class="gmail-author-a-wz75zz89zaz68zz88ziz87z7z122zz83z95z83zwh">. 
          </span><span
            class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99">It
            will also make it much easier to flag errors in the source
            data.</span></div>
        <div id="gmail-magicdomid21" class="gmail-ace-line"><br>
        </div>
        <div id="gmail-magicdomid22" class="gmail-ace-line"><span
            class="gmail-author-a-dz78zz77zz84z80jwz67z68t65f7">We
            aren't afraid that people will refrain from editing
            source-tagged objects: new users (using </span><span
            class="gmail-author-a-wz87zz90zz66zz82zt3itz84z5r8z80z4z69z">the</span><span
class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99"> iD
            editor</span><span
            class="gmail-author-a-dz78zz77zz84z80jwz67z68t65f7">) will
            probably not notice those tags in the first place; advanced
            users will know about them. And intermediate users will
            probably look them up. Merging and splitting are rather rare
            operations </span><span
            class="gmail-author-a-wz87zz90zz66zz82zt3itz84z5r8z80z4z69z">–</span><span
            class="gmail-author-a-dz78zz77zz84z80jwz67z68t65f7"> once
            the geometry is accurate, they should be</span><span
            class="gmail-author-a-z80ziz85za1pez66zo0z71zz72zjz66zz65zz78z">come</span><span
            class="gmail-author-a-dz78zz77zz84z80jwz67z68t65f7">
            unnecessary operations. </span><span
            class="gmail-author-a-wz87zz90zz66zz82zt3itz84z5r8z80z4z69z">Finally</span><span
            class="gmail-author-a-dz78zz77zz84z80jwz67z68t65f7">, as the
            data set is available for free under an open license, anyone
            can verify the data independently (though </span><span
            class="gmail-author-a-wz87zz90zz66zz82zt3itz84z5r8z80z4z69z">not</span><span
            class="gmail-author-a-dz78zz77zz84z80jwz67z68t65f7"> </span><span
            class="gmail-author-a-wz87zz90zz66zz82zt3itz84z5r8z80z4z69z">on
            the ground</span><span
            class="gmail-author-a-dz78zz77zz84z80jwz67z68t65f7">, </span><span
            class="gmail-author-a-wz87zz90zz66zz82zt3itz84z5r8z80z4z69z">of
            course</span><span
            class="gmail-author-a-dz78zz77zz84z80jwz67z68t65f7">).</span></div>
        <div id="gmail-magicdomid23" class="gmail-ace-line"><br>
        </div>
        <div id="gmail-magicdomid24" class="gmail-ace-line"><span
            class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99">We
            will address the worries about the external IDs on a point
            by point basis below:</span></div>
        <div id="gmail-magicdomid25" class="gmail-ace-line"><br>
        </div>
        <div id="gmail-magicdomid27" class="gmail-ace-line"><span
            class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99">From
          </span><span
            class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99
            gmail-b"><b>Frederik Ramm:</b></span></div>
        <div id="gmail-magicdomid29" class="gmail-ace-line">
          <ul class="gmail-list-indent1">
            <li><span
                class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99
                gmail-i"><i>If you delete a building</i></span><span
                class="gmail-author-a-wz75zz89zaz68zz88ziz87z7z122zz83z95z83zwh
                gmail-i"><i> </i></span><span
                class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99
                gmail-i"><i>that has such an ID, how will you ensure i</i></span><span
class="gmail-author-a-z80zz86zz87zz77zikorz83zz78zwn61z89zj gmail-i"><i>t</i></span><span
class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99
                gmail-i"><i> isn't brought in again</i></span><span
                class="gmail-author-a-wz75zz89zaz68zz88ziz87z7z122zz83z95z83zwh
                gmail-i"><i> </i></span><span
                class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99
                gmail-i"><i>through a later "update import"? Etc.</i></span><span
class="gmail-author-a-wz75zz89zaz68zz88ziz87z7z122zz83z95z83zwh gmail-i"><i>"</i></span></li>
          </ul>
        </div>
        <div id="gmail-magicdomid31" class="gmail-ace-line"><br>
        </div>
        <div id="gmail-magicdomid32" class="gmail-ace-line"><span
            class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99">There
            will not be "an update import". The tool is built for
            continuous use. It will improve the geometry of existing
            buildings and already looks at the source tags. The tool is
            buil</span><span
            class="gmail-author-a-dz78zz77zz84z80jwz67z68t65f7">t</span><span
class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99"> for
            heavy mappers who will be monitored by our closely knit
            community. Importing and updating will be done street by
            street</span><span
            class="gmail-author-a-dz78zz77zz84z80jwz67z68t65f7"> or </span><span
class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99">block
            by block</span><span
            class="gmail-author-a-dz78zz77zz84z80jwz67z68t65f7"> basis</span><span
class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99">. We
            believe we can</span><span
            class="gmail-author-a-dz78zz77zz84z80jwz67z68t65f7"> trust</span><span
class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99">
            these mappers to analy</span><span
            class="gmail-author-a-dz78zz77zz84z80jwz67z68t65f7">z</span><span
class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99">e
            situations like this on a case by case basis. In this case,
            whether of not the deleted building had an ID does not
            matter much.</span></div>
        <div id="gmail-magicdomid33" class="gmail-ace-line"><br>
        </div>
        <div id="gmail-magicdomid34" class="gmail-ace-line"><br>
        </div>
        <div id="gmail-magicdomid35" class="gmail-ace-line"><span
            class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99">These
            points, respectively by Frederik Ramm, Christoph Hormann and
            Mateusz Konieczny are quite similar. </span><span
            class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99
            gmail-b"><b>We'll answer them together.</b></span></div>
        <div id="gmail-magicdomid36" class="gmail-ace-line">
          <ul class="gmail-list-indent1">
            <li><span
                class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99
                gmail-i"><i>The idea of having an "audit trail" for
                  every single geometry by way of an Id for that
                  individual geometry is interesting, but I think that
                  it is totally sufficient if a changeset carries the
                  information that this changeset has been imported from
                  XYZ data source at time stamp T; everything else can
                  be researched down the line if the need should</i></span></li>
            <li><span
                class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99
                gmail-i"><i> For this purpose it is completely
                  unnecessary to bother the OSM  community with external
                  IDs.  If you want to check if the data has been 
                  unchanged since you added it then do exactly that -
                  check if there are  any newer versions of the objects
                  that have originally been added in  the import.</i></span></li>
          </ul>
        </div>
        <div id="gmail-magicdomid48" class="gmail-ace-line">
          <ul class="gmail-list-indent1">
            <li><span
                class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99
                gmail-i"><i>* What is the point of adding tags like
                  source:geometry:entity given that like any other tags
                  they may be edited once added to OSM?</i></span></li>
          </ul>
        </div>
        <div id="gmail-magicdomid49" class="gmail-ace-line"><br>
        </div>
        <div id="gmail-magicdomid50" class="gmail-ace-line"><span
            class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99">It
            is not, and has never been, the intention to fully automate
            this. As we are trying to make clear, this info is needed
            and used all the time, and not at some vague "maybe we do an
            update sometime". We are not using the IDs to make a direct
            (full database) comparison to see what exists in one and not
            in the other on a gigantic scale. We are not looking for an
            'auto update OSM with all new buildings added to GRB'. We
            use the IDs, on a SINGLE BUILDING comparison basis only to
            see what's changed (geometry touch-ups, or entirely
            replacing a building).</span></div>
        <div id="gmail-magicdomid51" class="gmail-ace-line"><br>
        </div>
        <div id="gmail-magicdomid52" class="gmail-ace-line"><span
            class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99">"THEY
            WON'T BE STABLE ANYWAY"</span></div>
        <div id="gmail-magicdomid53" class="gmail-ace-line"><span
            class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99">If
            the external references are edited, the tool will flag the
            building as needing an update. The only thing a (deliberate
            or accidental) unneeded change of the ID would do, is alert
            the tool's users that something doesn't add up. Either the
            building has been replaced in reality (and thus having a new
            UIDN), and you can improve the geometry of the new building.
            If no apparent change to the physical building can be
            traced, the UIDN can be restored.  The amount of 'false
            positives' due to unintended edits of the IDs is expect not
            to come remotely close to the useful flaggings.</span></div>
        <div id="gmail-magicdomid54" class="gmail-ace-line"><span
            class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99">Without
            the tags, it's hard to tell which buildings have been
            imported: you would need complicated spatial heuristics
            because we don't blindly copy buildings, we improve them
            through other sources. Once mapped, the geometry changing
            would happen more often than the tags changing, so we'd have
            a lot more false positives. <br>
          </span></div>
        <div class="gmail-ace-line"><span
            class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99"><br>
          </span></div>
        <div class="gmail-ace-line"><span
            class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99"><br>
          </span></div>
        <div id="gmail-magicdomid56" class="gmail-ace-line"><span
            class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99">"WHY
            NOT JUST CREATE A DATABASE OF LINKS EXTERNALLY?"</span></div>
        <div id="gmail-magicdomid57" class="gmail-ace-line"><span
            class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99">In
            theory it would be possible to have the tool keep a register
            of which OSM ID maps to which ID in the GRB or to evaluate
            changesets to get similar info. Some issues with this
            approach:</span></div>
        <div id="gmail-magicdomid451" class="gmail-ace-line">
          <ul class="gmail-list-bullet1">
            <li><span
                class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99">Because
                we aren't doing an automatic import but manually add the
                buildings through JOSM, this would require us to copy
                the OSM ID of each building manually, or rely on
                heuristics.</span></li>
          </ul>
        </div>
        <div id="gmail-magicdomid452" class="gmail-ace-line">
          <ul class="gmail-list-bullet1">
            <li><span
                class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99">We
                don't want to centralize the link between OSM and GRB.
                There isn't one single person doing the import, it's
                something several people in the community</span><span
                class="gmail-author-a-wz75zz89zaz68zz88ziz87z7z122zz83z95z83zwh">
                work on</span><span
                class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99">.
                Anyone could host the tool if its current maintainer
                disappears. </span></li>
          </ul>
        </div>
        <div id="gmail-magicdomid453" class="gmail-ace-line">
          <ul class="gmail-list-bullet1">
            <li><span
                class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99">Pushing
                the analysis of what exactly has happened towards a
                changeset is an unnecessary burden on a mapper. The
                changeset will not just contain "here be new buildings",
                but also "in this case, we just used part of the
                geometry of the GRB building, but we left part of the
                building geometry intact because that was better in
                OSM". After having analysed the changeset, the mapper
                would still have to look up in an import database what
                exactly the relationship between the OSM and GRB objects
                was at the time of import</span><span
                class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99
                gmail-comment gmail-c-aprzMNVsJDa6ka9t">.</span></li>
          </ul>
        </div>
        <div id="gmail-magicdomid454" class="gmail-ace-line">
          <ul class="gmail-list-bullet1">
            <li><span
                class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99">An
                external database providing the link between OSM and GRB
                objects would be outdated after a day and almost
                impossible to update with changes on the OSM side</span></li>
          </ul>
        </div>
        <div id="gmail-magicdomid456" class="gmail-ace-line"><br>
        </div>
        <div id="gmail-magicdomid62" class="gmail-ace-line"><br>
        </div>
        <div id="gmail-magicdomid63" class="gmail-ace-line"><span
            class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99">"WHY
            NOT JUST COMPARE GEOMETRIES?"</span></div>
        <div id="gmail-magicdomid64" class="gmail-ace-line"><span
            class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99">Doing
            a geometrical analysis to analyze differences would be
            impractical, not just because this would be computationally
            heavy, but also because it would lead to too much false
            positives. For example because of tiny changes, but also
            because we do not blindly use source geometries since we
            first address overnoding.</span></div>
        <div class="gmail-ace-line"><span
            class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99"><br>
          </span></div>
        <div class="gmail-ace-line"><span
            class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99"><br>
          </span></div>
        <div class="gmail-ace-line"><span
            class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99"><br>
          </span></div>
        <div class="gmail-ace-line"><span
            class="gmail-author-a-cz67zz90zo5jz86zz83zz75zz85zz87zz82zz80zh99">I
            hope that all confusion is cleared now and that we can move
            this issue forward.<br>
          </span></div>
        <div>
          <div dir="ltr" class="gmail_signature"
            data-smartmail="gmail_signature">
            <div dir="ltr"><br>
            </div>
            <div>With Regards,</div>
            <div>The Belgian Mappers,</div>
            <div>Pieter Vander Vennet<br>
            </div>
          </div>
        </div>
      </div>
      <br>
      <fieldset class="mimeAttachmentHeader"></fieldset>
      <pre class="moz-quote-pre" wrap="">_______________________________________________
Imports mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Imports@openstreetmap.org">Imports@openstreetmap.org</a>
<a class="moz-txt-link-freetext" href="https://lists.openstreetmap.org/listinfo/imports">https://lists.openstreetmap.org/listinfo/imports</a>
</pre>
    </blockquote>
  </body>
</html>