[Talk-ca] GeoBase discussion

[Sam Vekemans]

Hi all,I got a little more feedback from the
talk at openstreetmap.orgdiscussion list :)
and from sending out my message to as many users in canada as i can find...
in hopes to get more feedback.

Harry Wood (OSM) made a comment about the discussion vs. the statements, and
so, im seeing some great points being discussed :)

So here's some more mud i'll through at the wall... see what sticks :0).
  I have made some arguments... which of course can be defeated with some
good facts.  :)  lets play.


1. .. We use the power of Wiki and community, and go ahead and import the
most recent data available, and import all the different types. (the Canadian
Geodetic Network<http://wiki.openstreetmap.org/wiki/GeoBase_Import#Geobase:_Canadian_Geodetic_Network>
can
probably be imported after the water features. .. as for sure there wont be
any confusion there.)
So for even in the areas that will result in 'ghost lines', we can post on a
chart all the areas that need work. .. so say your working in
the Toronto or Calgary area. .. those working on it, and can see that they
need help, can post it on the chart. . so anyone who is available to help.
can jump in there, contact the users to announce that you are helping. .. so
it's a coordinated effort.  (we've done these charts before, for measuring
what roads are done, and level of completeness.  This is a much larger
scale, but still manageable)

Personally, i think that going at it this way, the positives (tonnes more
data) far out way the negatives (fixing up and moving other peoples work).
... another positive is that once all the GeoBase data has been imported...
as were going along fixing it, where we are there in person, we can also add
things like the land  use, and where bus stops, and trails, and shops that
GeoBase wont have.

2. .. The discussion about the idea that the GeoBase dataset is also
maintained, brings me to this conclusion.
Well... you know how, for the time being we are using 'relatively crappy'
yahoo imagery, and it's only been certain areas that we have had better
imagery for? ... yahoo imagery is also (in a sence) a maintained database of
aerial imagery..
... then once the new imagery became available... we see that the place we
traced is different than the image.. so we need to go back there and fix it.
 ... this will be the same thing with GeoBase updates.
(and Linz Updates  BTW)

Well.... what if we (after the 1st big import of everything, and are
up-to-date) ... and notifying Geo Base about the import progress.  (perhaps
as a community we could help push Geo Base to get a bit more updates
available? ... or as much as possible. ... if the results can
be physically shown to each province, perhaps that along can be motivation
to be as up-to-date as possible?? ..  Setting a time-line, cut-off date,
were any more data available doesn't directly get imported? So even the
future updates don't get imported.
... so creating a WMS layer for the new data, and having that available, so
we can use it to trace with.
(my assumption, if the technical process of identifying current OSM data,
and adding the GeoBase ID to it) .. so every piece of GeoBase data, has it's
own id attached to it.. ... is NOT technically possible.)

I think it is safe to say, that the desire is there for OpenStreetMap to
become even better than GeoBase. .. and YES.. it will.  For the moments
after the last piece of current available data is uploaded. .. and before
the next update of Geobase. .. it will :)

So the looking at the future sinerio. .. say it's all imported, and we fixed
up all the ghost lines... and are all upto date. .. then GeoBase announces
that an update for BC roads is available. ... so what now? ..
We see that some of this update, has already been done by OSM users.
I think that in this case... the ONLY options are to 1 - import it with a
'smart - detecting and adding' script... or to 2- have it as a WMS layer
where we can trace over it. ... just as we would flip on the yahoo imagery
layer to get a better view.  3- skip it, and try to get GeoBC to start
posting their data directly to the largest OpenSource Database on the planet
(directly from the survay people who are out there in the field), or try to
get them to have it available as a WMS layer, so anyone can use to trace
with.

I like the idea of tracing, and only importing datasets were we know that
the current data either doesn't exist, or is available in a state which
is manageable.
In this case, although yes it is frustrating for those who contributed data,
we do have the people power available to fix it, in an orderly fashion.

So anyway... on another note:
For those of you who have now been (hopefully) using the (old and outdated)
Ibycus topo,  you see that there are tonnes of poi's that are included in
there.   Does anyone know what data set this information came from. .. ie.
Campgrounds, schools, churches, libraries.  etc. all of which are available
on the Ibycus Topo.  - I havent yet added the source of those features, but
would like too.

A tonne of information,
Hopefully points can be shot down :)  ...
Once we get a firm agreement on the process boundaries, i think it's easier
to define all the steps to take. (we do have lots so far) .. now to organize
it all to manageable chunks.

Cheers,
Sam Vekemans
Across Canada Trails
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/talk-ca/attachments/20081205/24f58918/attachment.html>