Loading ...
Sorry, an error occurred while loading the content.

Re: Things we can agree on?

Expand Messages
  • davewissenbach@yahoo.com
    ... I think that we re all XML novices here, but I ll write a DTD and validate output (yours and mine) just as soon as we can code some samples. I actually had
    Message 1 of 24 , Nov 2, 2001
    View Source
    • 0 Attachment
      --- In gpsxml@y..., Dan Foster <egroups@t...> wrote:

      I think that we're all XML novices here, but I'll write a DTD and
      validate output (yours and mine) just as soon as we can code some
      samples. I actually had finished converting my program to write your
      last sample formats, without the DTD, but I'm going to spend a little
      time to code the optional attributes as elements, per the most recent
      discussions here.

      One thing that won't be covered by the DTD is a design decision that
      I had to make--that I would only separately output waypoints which
      are not part of routes.

      The Garmin GPSs give us the waypoints twice, but I see no reason to
      encode all of these twice, once as a routepoint, and once as a
      waypoint.
      >
      > As I've said before, DTDs aren't my strong suit, so I hope you'll
      help
      > define the DTD for GPX once we start putting the pieces together.
      >
      > --
      > Dan Foster
      > TopoGrafix - GPS Software, Waypoints, and Maps
      > http://www.topografix.com - mailto:egroups@t...
    • Dan Foster
      Hello, Here s a sample GPX file containing some waypoints and a route from a hike I took recently. Comments? Anyone else have sample data at this point?
      Message 2 of 24 , Nov 7, 2001
      View Source
      • 0 Attachment
        Hello,

        Here's a sample GPX file containing some waypoints and a route from a
        hike I took recently. Comments? Anyone else have sample data at this
        point? Eventually, we should have a set of GPX files to validate
        against.

        Hike.gpx

        <?xml version="1.0"?>
        <gpx version="1.0" src="EasyGPS">
        <wpt lat="42.723617" lon="-71.912150">
        <id>BINNEYPOND</id>
        <desc>Binney Ponds</desc>
        <type>Pond</type>
        <sym>Fishing Area</sym>
        </wpt>
        <wpt lat="42.730400" lon="-71.918250">
        <id>LUNCH</id>
        <desc>Lunch Rocks</desc>
        <ele>517.300000</ele>
        <type>Summit</type>
        <sym>Picnic Area</sym>
        </wpt>
        <wpt lat="42.696767" lon="-71.892517">
        <id>WATATIC</id>
        <desc>Mount Watatic</desc>
        <ele>568.500000</ele>
        <type>Summit</type>
        <sym>Summit</sym>
        </wpt>
        <wpt lat="42.711233" lon="-71.898167">
        <id>STATELINE</id>
        <desc>NH-MA State Line</desc>
        <ele>485.600000</ele>
        <type>Intersection</type>
        <sym>Dot</sym>
        </wpt>
        <wpt lat="42.696850" lon="-71.904567">
        <id>PARKING</id>
        <desc>Parking Lot</desc>
        <ele>417.300000</ele>
        <type>Parking</type>
        <sym>Car</sym>
        </wpt>
        <wpt lat="42.733017" lon="-71.919000">
        <id>PRATTMOUNT</id>
        <desc>Pratt Mountain</desc>
        <type>Summit</type>
        <sym>Summit</sym>
        </wpt>
        <wpt lat="42.720267" lon="-71.912383">
        <id>STREAM</id>
        <desc>Stream Crossing</desc>
        <type>Stream</type>
        <sym>Dot</sym>
        </wpt>
        <rte>
        <rtept lat="42.696850" lon="-71.904567">
        <id>PARKING</id>
        <desc>Parking Lot</desc>
        <ele>417.300000</ele>
        <type>Parking</type>
        <sym>Car</sym>
        </rtept>
        <rtept lat="42.696767" lon="-71.892517">
        <id>WATATIC</id>
        <desc>Mount Watatic</desc>
        <ele>568.500000</ele>
        <type>Summit</type>
        <sym>Summit</sym>
        </rtept>
        <rtept lat="42.711233" lon="-71.898167">
        <id>STATELINE</id>
        <desc>NH-MA State Line</desc>
        <ele>485.600000</ele>
        <type>Intersection</type>
        <sym>Dot</sym>
        </rtept>
        <rtept lat="42.720267" lon="-71.912383">
        <id>STREAM</id>
        <desc>Stream Crossing</desc>
        <type>Stream</type>
        <sym>Dot</sym>
        </rtept>
        <rtept lat="42.723617" lon="-71.912150">
        <id>BINNEYPOND</id>
        <desc>Binney Ponds</desc>
        <type>Pond</type>
        <sym>Fishing Area</sym>
        </rtept>
        <rtept lat="42.730400" lon="-71.918250">
        <id>LUNCH</id>
        <desc>Lunch Rocks</desc>
        <ele>517.300000</ele>
        <type>Summit</type>
        <sym>Picnic Area</sym>
        </rtept>
        <rtept lat="42.733017" lon="-71.919000">
        <id>PRATTMOUNT</id>
        <desc>Pratt Mountain</desc>
        <type>Summit</type>
        <sym>Summit</sym>
        </rtept>
        </rte>
        </gpx>




        --
        Dan Foster
        TopoGrafix - GPS Software, Waypoints, and Maps
        http://www.topografix.com - mailto:egroups@...
      • davewissenbach@yahoo.com
        ... a ... this ... I ve posted sample data at my incomplete web site. Deep link to the URL www.cableone.net/cdwissenbach/FileFormat.gpx The document is not
        Message 3 of 24 , Nov 7, 2001
        View Source
        • 0 Attachment
          --- In gpsxml@y..., Dan Foster <egroups@t...> wrote:
          > Hello,
          >
          > Here's a sample GPX file containing some waypoints and a route from
          a
          > hike I took recently. Comments? Anyone else have sample data at
          this
          > point? Eventually, we should have a set of GPX files to validate
          > against.
          >
          > Hike.gpx
          >

          I've posted sample data at my incomplete web site. Deep link to the
          URL

          www.cableone.net/cdwissenbach/FileFormat.gpx

          The document is not valid against the DTD because I did not define
          the tags contained in the <private> element. The ANY keyword does not
          absolve one from the responsibility of defining the other elements.
          My reading of the XML specification indicates that we should be able
          to have both an external DTD and inline portion. I'll let you know if
          this works. Also, although my application outputs gpx format, the
          application doesn't read its own output! That's next week's job.

          I've not finished the DTD. I'll Add K.J.'s stuff and the attributes
          of gpx, version and src, and correct any other omissions.
        • Kjeld Jensen
          Hi.... A few comments... The owerall format looks great and mine is very similar to it. I am not sure if we have to put a tag at the bottom in order to
          Message 4 of 24 , Nov 8, 2001
          View Source
          • 0 Attachment
            Hi....

            A few comments...

            The owerall format looks great and mine is very similar to it. I am not
            sure if we have to put a </xml> tag at the bottom in order to "close" the
            <?xml....> tag?

            I would consider src="EasyGPS" as optional which by your current format
            should then be coded as an element?

            Perhaps we should save the <id> tag for numbering of tracks and routes and
            then use <name> instead for the name?


            <wpt lat="42.723617" lon="-71.912150"> (no id necessary)
            ...
            </wpt>

            <route>
            <rtept id="1" lat="42.723617" lon="-71.912150">
            ...
            </rtept>
            <rtept id="2" lat="42.723617" lon="-71.912150">
            ...
            </rtept>
            <rtept id="3" lat="42.723617" lon="-71.912150">
            ...
            </rtept>
            </route>

            <track>
            <trkpt id="1" lat="42.723617" lon="-71.912150">
            ...
            </trkpt>
            <trkpt id="2" lat="42.723617" lon="-71.912150">
            ...
            </trkpt>
            <trkpt id="3" lat="42.723617" lon="-71.912150">
            ...
            </trkpt>
            </track>

            Maybe the latter example could also solve our problem with interrupted
            tracks? That is we could just reset the id tag to "1"...

            Maybe it would be a good idea to drop the DTD for <gpx version="1.0"> as
            this could speed things up a bit here in the beginning. Then let us create
            one for version 1.1 instead. Or perhaps we should call the first agreed
            format for "0.1" :-)

            What do we do about the time format? Do we use the modified string or do we
            add <msec>? I would prefer 20011021T142003.293Z to keep things simple, but
            it is not important to me.

            I will make a few changes to the Cetus GPS software today in order to adapt
            some of the features in your sample file. Then I will then post the result
            here..

            Kjeld


            ______________________
            Kjeld Jensen
            N 55° 22' E 10° 24'
            Email: gps@...
            http://www.cetus.dk/gps
          • davewissenbach@yahoo.com
            ... routes and ... Why is ID a required attribute of a trkpt and a rtpte? (Remember that we have previously agreed that required attributes of an element would
            Message 5 of 24 , Nov 8, 2001
            View Source
            • 0 Attachment
              --- In gpsxml@y..., Kjeld Jensen <gps@c...> wrote:
              > Hi....
              >
              > A few comments...
              >
              >
              > Perhaps we should save the <id> tag for numbering of tracks and
              routes and
              > then use <name> instead for the name?
              >
              >

              Why is ID a required attribute of a trkpt and a rtpte? (Remember that
              we have previously agreed that required attributes of an element
              would be coded with the XML attribute syntax, and that optional data
              would be coded as elements.) Have we changed this rule while I was
              offline? If so, I'd suggest that we use id in the XML sense of the
              word, as a special kind of attribute that must be unique from all
              other attributes. The Garmin GPS requires that all waypoints be
              unique, and the ID is used for display purposes. Usability would be
              improved if the source application could control the ID which is
              displayed on the Garmin unit, which therefore implies that all
              waypoints have a Unique ID. And because routes are composed of
              waypoints, routepoints must also have unique ID's. But I'd make these
              optional.

              So my view of the GPS/GPX world is in conflict with the numbered ID
              sequence which you propose below. Or is what I'm referring to above
              as an ID now to be called a name?

              What is the purpose of sequentially numbering points within a track
              or route? If the purpose is just to detect loss of lock on the GPS,
              or a break in a track, can we code this in some other way?

              > <wpt lat="42.723617" lon="-71.912150"> (no id necessary)
              > ...
              > </wpt>
              >
              > <route>
              > <rtept id="1" lat="42.723617" lon="-71.912150">
              > ...
              > </rtept>
              > <rtept id="2" lat="42.723617" lon="-71.912150">
              > ...
              > </rtept>
              > <rtept id="3" lat="42.723617" lon="-71.912150">
              > ...
              > </rtept>
              > </route>
              >
              > <track>
              > <trkpt id="1" lat="42.723617" lon="-71.912150">
              > ...
              > </trkpt>
              > <trkpt id="2" lat="42.723617" lon="-71.912150">
              > ...
              > </trkpt>
              > <trkpt id="3" lat="42.723617" lon="-71.912150">
              > ...
              > </trkpt>
              > </track>
              >
              >
              > I will make a few changes to the Cetus GPS software today in order
              to adapt
              > some of the features in your sample file. Then I will then post the
              result
              > here..
              >
              > Kjeld
              >
              >
              > ______________________
              > Kjeld Jensen
              > N 55° 22' E 10° 24'
              > Email: gps@c...
              > http://www.cetus.dk/gps
            • Kevin Read
              I agree somewhat.... rather than id for trackpoints perhaps a sequence is inferred here instead. taking the sample this would equate to ... this would allow
              Message 6 of 24 , Nov 8, 2001
              View Source
              • 0 Attachment
                I agree somewhat....
                 
                rather than "id" for trackpoints perhaps a sequence is inferred here instead.
                 
                taking the sample this would equate to
                 
                > <route>
                > <rtept seq="1" lat="42.723617"
                lon="-71.912150">
                > ...
                > </rtept>
                > <rtept
                seq="2" lat="42.723617" lon="-71.912150">
                > ...
                >
                </rtept>
                > <rtept seq="3" lat="42.723617"
                lon="-71.912150">
                > ...
                > </rtept>
                >
                </route>
                this would allow for an application to confidently sort the route/track using DOM or XSLT for purposes such as track reversal
                 
                Kevin
                 
                 -----Original Message-----
                From: davewissenbach@... [mailto:davewissenbach@...]
                Sent: Friday, 9 November 2001 4:06 PM
                To: gpsxml@yahoogroups.com
                Subject: [gpsxml] Re: GPX Sample File: Hike.gpx

                --- In gpsxml@y..., Kjeld Jensen <gps@c...> wrote:
                > Hi....
                >
                > A few comments...
                >
                >
                > Perhaps we should save the <id> tag for numbering of tracks and
                routes and
                > then use <name> instead for the name?
                >
                >

                Why is ID a required attribute of a trkpt and a rtpte? (Remember that
                we have previously agreed that required attributes of an element
                would be coded with the XML attribute syntax, and that optional data
                would be coded as elements.) Have we changed this rule while I was
                offline? If so, I'd suggest that we use id in the XML sense of the
                word, as a special kind of attribute that must be unique from all
                other attributes. The Garmin GPS requires that all waypoints be
                unique, and the ID is used for display purposes. Usability would be
                improved if the source application could control the ID which is
                displayed on the Garmin unit, which therefore implies that all
                waypoints have a Unique ID. And because routes are composed of
                waypoints, routepoints must also have unique ID's. But I'd make these
                optional.

                So my view of the GPS/GPX world is in conflict with the numbered ID
                sequence which you propose below. Or is what I'm referring to above
                as an ID now to be called a name?

                What is the purpose of sequentially numbering points within a track
                or route? If the purpose is just to detect loss of lock on the GPS,
                or a break in a track, can we code this in some other way?

                > <wpt lat="42.723617" lon="-71.912150"> (no id necessary)
                > ...
                > </wpt>
                >
                > <route>
                > <rtept id="1" lat="42.723617" lon="-71.912150">
                > ...
                > </rtept>
                > <rtept id="2" lat="42.723617" lon="-71.912150">
                > ...
                > </rtept>
                > <rtept id="3" lat="42.723617" lon="-71.912150">
                > ...
                > </rtept>
                > </route>
                >
                > <track>
                > <trkpt id="1" lat="42.723617" lon="-71.912150">
                > ...
                > </trkpt>
                > <trkpt id="2" lat="42.723617" lon="-71.912150">
                > ...
                > </trkpt>
                > <trkpt id="3" lat="42.723617" lon="-71.912150">
                > ...
                > </trkpt>
                > </track>
                >
                >
                > I will make a few changes to the Cetus GPS software today in order
                to adapt
                > some of the features in your sample file. Then I will then post the
                result
                > here..
                >
                > Kjeld
                >
                >
                > ______________________
                > Kjeld Jensen
                > N 55° 22' E 10° 24'
                > Email: gps@c...
                > http://www.cetus.dk/gps



                To unsubscribe from this group, send an email to:
                gpsxml-unsubscribe@yahoogroups.com



                Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.
              • Kjeld Jensen
                Hi, ... No :-) If so, I d suggest that we use id in the XML sense of the ... The id was changed to name yes, and I thought id could be a sequential numbering
                Message 7 of 24 , Nov 9, 2001
                View Source
                • 0 Attachment
                  Hi,

                  >Why is ID a required attribute of a trkpt and a rtpte? (Remember that
                  >we have previously agreed that required attributes of an element
                  >would be coded with the XML attribute syntax, and that optional data
                  >would be coded as elements.) Have we changed this rule while I was
                  >offline?

                  No :-)

                  If so, I'd suggest that we use id in the XML sense of the
                  >word, as a special kind of attribute that must be unique from all
                  >other attributes. The Garmin GPS requires that all waypoints be
                  >unique, and the ID is used for display purposes. Usability would be
                  >improved if the source application could control the ID which is
                  >displayed on the Garmin unit, which therefore implies that all
                  >waypoints have a Unique ID. And because routes are composed of
                  >waypoints, routepoints must also have unique ID's. But I'd make these
                  >optional.

                  >So my view of the GPS/GPX world is in conflict with the numbered ID
                  >sequence which you propose below. Or is what I'm referring to above
                  >as an ID now to be called a name?

                  >What is the purpose of sequentially numbering points within a track
                  >or route? If the purpose is just to detect loss of lock on the GPS,
                  >or a break in a track, can we code this in some other way?

                  The id was changed to name yes, and I thought id could be a sequential
                  numbering of all track/route records.

                  But to me the name is not necessarily unique. I see the requirement for a
                  unique id (for the purpose you describe) the same way as limitations on
                  string lengths and so on. Limitations that may be convenient for one
                  application should not cause limitations in the gpx standard. If you for
                  instance export let say 100 waypoints from a database it is quite difficult
                  to make sure that each name is unique at least in some applications.

                  If we define that trackpoints must follow in a sequence, then I guess the
                  id="" is not that necessary. Actually I like the element name seq="" better
                  because it tells more about what it is, so if we are to include a
                  sequential number for tracks and routes, then I think it should be called
                  seq rather than id.

                  Kjeld

                  ______________________
                  Kjeld Jensen
                  N 55° 22' E 10° 24'
                  Email: gps@...
                  http://www.cetus.dk/gps
                • Kjeld Jensen
                  ... I agree on the seq name and you have a good point regarding the sorting option. Kjeld ______________________ Kjeld Jensen N 55° 22 E 10° 24 Email:
                  Message 8 of 24 , Nov 9, 2001
                  View Source
                  • 0 Attachment
                    >> <rtept seq="2" lat="42.723617" lon="-71.912150">
                    >> ...
                    >> </rtept>

                    >this would allow for an application to confidently sort the
                    >route/track using DOM or XSLT for purposes such as track reversal

                    I agree on the seq name and you have a good point regarding the sorting option.

                    Kjeld

                    ______________________
                    Kjeld Jensen
                    N 55° 22' E 10° 24'
                    Email: gps@...
                    http://www.cetus.dk/gps
                  • Adrian Seccombe
                    ... that ... data ... these ... sequential ... **************** Having read to the end ( A great body of work) the only concern I have ((after the Time Zone
                    Message 9 of 24 , Nov 25, 2001
                    View Source
                    • 0 Attachment
                      --- In gpsxml@y..., Kjeld Jensen <gps@c...> wrote:
                      > Hi,
                      >
                      > >Why is ID a required attribute of a trkpt and a rtpte? (Remember
                      that
                      > >we have previously agreed that required attributes of an element
                      > >would be coded with the XML attribute syntax, and that optional
                      data
                      > >would be coded as elements.) Have we changed this rule while I was
                      > >offline?
                      >
                      > No :-)
                      >
                      > If so, I'd suggest that we use id in the XML sense of the
                      > >word, as a special kind of attribute that must be unique from all
                      > >other attributes. The Garmin GPS requires that all waypoints be
                      > >unique, and the ID is used for display purposes. Usability would be
                      > >improved if the source application could control the ID which is
                      > >displayed on the Garmin unit, which therefore implies that all
                      > >waypoints have a Unique ID. And because routes are composed of
                      > >waypoints, routepoints must also have unique ID's. But I'd make
                      these
                      > >optional.
                      >
                      > >So my view of the GPS/GPX world is in conflict with the numbered ID
                      > >sequence which you propose below. Or is what I'm referring to above
                      > >as an ID now to be called a name?
                      >
                      > >What is the purpose of sequentially numbering points within a track
                      > >or route? If the purpose is just to detect loss of lock on the GPS,
                      > >or a break in a track, can we code this in some other way?
                      >
                      > The id was changed to name yes, and I thought id could be a
                      sequential
                      > numbering of all track/route records.
                      > ****************
                      ****************
                      Having read to the end ( A great body of work) the only concern I
                      have ((after the Time Zone question)) is the question of uniqueness
                      of NAME (See below)
                      I believe this one will bite if Name is not required to be Unique
                      within each .GPX file, or am I missing a trick?
                      Am I the only paranoid thinking this?
                      ***************
                      ***************
                      > But to me the name is not necessarily unique. I see the requirement
                      for a
                      > unique id (for the purpose you describe) the same way as
                      limitations on
                      > string lengths and so on. Limitations that may be convenient for one
                      > application should not cause limitations in the gpx standard. If
                      you for
                      > instance export let say 100 waypoints from a database it is quite
                      difficult
                      > to make sure that each name is unique at least in some applications.
                      >
                      > If we define that trackpoints must follow in a sequence, then I
                      guess the
                      > id="" is not that necessary. Actually I like the element name
                      seq="" better
                      > because it tells more about what it is, so if we are to include a
                      > sequential number for tracks and routes, then I think it should be
                      called
                      > seq rather than id.
                      >
                      > Kjeld
                      >
                      > ______________________
                      > Kjeld Jensen
                      > N 55° 22' E 10° 24'
                      > Email: gps@c...
                      > http://www.cetus.dk/gps

                      Adrian
                      Email: adrius42@...
                    Your message has been successfully submitted and would be delivered to recipients shortly.