dive location database?

[Dirk Hohndel]

On Wed, Nov 25, 2015 at 12:56:35PM -0800, K. "pestophagous" Heller wrote:
> On Wed, Nov 25, 2015 at 11:50 AM, Dirk Hohndel <dirk at hohndel.org> wrote:
> > On Wed, Nov 25, 2015 at 10:17:40AM +0100, Guillaume Gardet wrote:
> >> Hi,
> >>
> >> I am working on a project of dive sites (and dive centers) database, which will be released under an open license.
> >> It will be available on a GIT repo, probably github.
> >> A website will be also available with an API to access the data (XML and JSON outputs).
> >>
> >> It would be very interesting if Subsurface could make use of it.
> >> I can give access to the testing website if some subsurface devs want to play with it.
> >
> > We have a couple of other sources of dive site data bases. The more the
> > merrier. But the gating factor is that someone needs to implement the UI
> > to use this. It's on my todo list. So are many other things.
> >
> > It's our usual problem. No shortage of good ideas.
> > A distinct shortage of people willing and able to work on the UI.
> 
> i would be happy to work on this.  as a newcomer, i have been trying
> to find all the "things that need doing" without bugging anyone with a
> lot of questions.

Ha! Wrong approach.
Let me rephrase this. Great approach, but maybe not the best possible way
to do this :-)

I have noticed the patches and appreciate them a lot.

But asking questions, creating discussions on the mailing list and
eliciting feedback is actually the best way to move the project forward.
There are weeks where it feels to me that I'm the only one doing that and
that is at times a bit overwhelming. So other chiming in, driving parts
forward is extremely helpful.

For example, Robert has completely taken over maintenance of the planner
and the deco side of things (and to some degrees the handling of photos).
That to me has been among the best things that happened this year.

I was thrilled to see Guido come in and rework the Uemis downloader.
Desperately needed work, long overdue (but he's been rather quiet,
lately... given his background I hoped that he would start pushing things
further on the QA / testing side of Subsurface... guess I need to poke him
a bit).

Simon came in with the AppImage. OK, sorry Robert, THAT was the best thing
that happened in the last few months. Or maybe it was Sebastian joining
the team and taking over the QML UI of Subsurface-mobile?

OK, I guess I shouldn't complain, there are a lot of people doing really
amazing work here...

> i have mainly operated using the trac tickets as a todo list. (patches
> for 733, 820, and 912 so far.) in trac, people had already taken the
> time to describe what needed doing, so i was able to find work to do
> without waiting for anyone to explain things.
> 
> the idea of saving time entering locations sounds very appealing.
> what is the envisioned workflow?
> is reverse geocoding the main interest? (the diver has GPS
> coordinates, and subsurface retrieves the "friendly" site name from
> the public/shared database?)

That we already have and it's working reasonably well. This could always
be improved and there was some talk about trying to figure out ways to
better interoperate with other dive logs and the way they store things (in
their taxonomy).

> or would the workflow be the opposite: enter some human-readable text
> first, and then subsurface attempts to grab corresponding GPS coords
> from the database?
> or maybe people want both kinds of lookup?

Yes they do. The rough idea of using the dive site data bases was
something like this:

a) user enters information where they are diving (this could be "for this
trip" or something along those lines, or it could be when working on a
specific dive site). The information could be textual, but I think that's
way too complicated. So maybe it's easier if we ask the user to use the
globe to show a rectagular area that encloses where the dives are. So the
user can pan and scroll until the right area is shown and then click "this
is where I'm diving"

b) now the user tries to set the location of a specific dive site; they
enter a name or partial name. We use the dive site data base lookup to
populate both a drop down list and the Marble display (the globe) with
dive sites that match the partial name. If nothing is entered, all dive
sites in the area are shown

c) the user picks one of the sites (either on the globe or in the drop
down

d) success

> besides Guillaume's dive site database, what other dive site databases
> should be considered? (all i found mentioned in mail list archives was
> general stuff like OpenStreetMap)

divelogs.de has an API that allows us to search for dive sites around a
gps fix. divesites.com does as well. The maintainers of both websites have
agreed to help us with the integration (so if we need changes to the API
to make this easier they will work with us). Rainer (divelogs.de) is even
on this mailing list.

> is there a trac ticket for this?  even if i am not the right person
> for the task, writing a "rough spec" in a trac ticket might attract
> the attention of others like me who have the quiet, lurker mindset :)

Good point. I'll add one with the information above.

/D