[OSM-talk] relating trace points to gpx files

[Andy Robinson]

Wading through the quagmire of discussion here there were a few points I
noted and wish to comment upon.

1. Why yellow dots? - As I recall the gpx dots started out grey/white I
think. Then with the previous incarnation of the applet it became difficult
to discern the grey/white dots from the line segments being drawn so Steve
increased the size of the spot and changed the colour to pink. That was a
problem for a particular user who pointed out a colour blindness issue so it
was changed to yellow. It could in theory be any other colour as far as I
can see and there was no great debate over the yellow. As far as I can see
yellow works fine?

2. When a gpx file data is imported to the db I assume that each point gets
referenced with the user id and/or the name of the original log file. If the
log file is made public then there is no obvious reason not to make the
origin of the point available (ie the log file from which it comes). In JOSM
it's really helpful with your own logs to be able to colour the different
tracks separately. However this level of ability in the applet is probably
not required if the heavy users all use non applet editing clients. For logs
that are not made public then the current anonymity must surely remain.
Unless of course the project decides to change the acceptance basis for logs
in that future uploads are all made public in the first place. There is some
argument for this to help stop the upload of inappropriate track logs but
that seems the only argument for me and thus I prefer the current status
quo.

3. History and user information for map data is another avenue entirely and
one I wish to see supported at some stage. I have yet to hear any cry of
concern about making the history data for map edits available. The project
at some point needs to show its true wiki colours in this respect. When I
create a feature on the map I do so using more information that just the
tracklog data underneath. I have pictures, my memory and quite possibly an
intimate knowledge of the area being mapped. Thus as the initial creator of
that node, segment, way or area I impart additional knowledge. By providing
a means to access or query that knowledge (via a secure interaction through
the server to the user at the other end) it may be possible to gain a lot
more info about a location and not necessarily just about the creation of
the map. It opens up the possibility for a much richer user experience. To
give an example: I do a lot of genealogical research. The Genes Reunited
website allows users to upload their family tree. The trees are private
unless you open them up to another specific user. Other users can query all
the trees looking for names but do not know who a name in the big database
originates from. To get in touch with the other researcher an email is sent
by the Genes Reunited server to the other user who will either respond (via
the server again) to the original requester or will ignore the request
entirely. These requests for information from other users of the system are
called your "contacts". I believe that OSM could potentially work in a very
similar way. No direct contact between users would be possible via the data
unless a user grants it outside of the system. Once a contact is made it
either deals with the initial enquiry and then no further contact is made or
the two contact become lifelong buddies.... well you get the idea. On Genes
Reunited I have been astounded how many fellow researches (not just distant
relatives) I have been able to communicate with and find over the last few
years. OSM could go along way to doing something similar.

For me these issues have nothing to do with privacy. The privacy aspect I
leave to a technical solution. Rather I believe we can bring a much richer
experience to the existing OSM community and help draw in more users because
of the level of community interaction that is on offer.

Cheers,

Andy

Andy Robinson
Andy_J_Robinson at blueyonder.co.uk 

>-----Original Message-----
>From: talk-bounces at openstreetmap.org [mailto:talk-
>bounces at openstreetmap.org] On Behalf Of Lars Aronsson
>Sent: 12 July 2006 05:05
>To: talk at openstreetmap.org
>Subject: Re: [OSM-talk] relating trace points to gpx files
>
>Nick Black wrote:
>
>> > This mantra is being repeated so often, that it is almost becoming
>> > a truth.
>>
>> What do you mean?  The reason tracklogs aren't public is because of
>> legal concerns , mainly relating to future contraints of the Freedom
>> of Information Act.  It isnt a Mantra - its a legal opinion.
>
>No, it's not the result of an informed decision.  It's a feature
>not yet implemented (which I can accept) and an excuse invented
>after the fact, often mentioned as an explanation.
>
>In the edit applet, trackpoints are yellow.  This doesn't mean
>that all other colors have been tried and found to be inferior to
>yellow.
>
>If I draw a line segment and next day find that the segment has
>moved, there currently is no function to tell me who moved it.
>The one who moved it also has no means to tell me why they did so.
>This is just because these functions haven't been implemented yet.
>We have absolutely no consensus that they should never be
>implemented.
>
>And we have no consensus that trackpoints must never be possible
>to trace back to a file.  We have some concerns (not a consensus,
>but still valid and interesting concerns) that it might be a
>privacy problem if we can locate a person to a point in place and
>time.  But providing the map editor with information that this
>point belongs to tracklog file #134 is *not* a privacy problem.
>
>I don't require that somebody should implement these functions now
>or anytime.  This is a volunteer project, and I'm happy every time
>anybody makes improvements.  But we should keep a clear record of
>what has been decided and what has not.
>
>
>--
>  Lars Aronsson (lars at aronsson.se)
>  Aronsson Datateknik - http://aronsson.se
>
>_______________________________________________
>talk mailing list
>talk at openstreetmap.org
>http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk