Hi Kieran

You fixed issue 40 using the database, but as you can see from issue
41 and 43 I'm not very keen on the database. If osmdroid is made as a
library to replace Google Maps that will mean getting rid of the
database and the service.

MySpeed is not open source but I'm willing to share commonly useful
code by contributing to osmdroid and osn.

To detect if you've gone off the route it simply finds the shortest
distance between your position and the route. This could be quite a
bit of processing so I only do it every 30 seconds, although I think
you could probably easily do it more often. I originally tried only
checking part of the route but it seemed to be easier just to keep the
entire route in memory and check the entire route rather than trying
to figure out which part of the route to check. Some points on the
route could be miles apart, so you need to check a pretty big area,
and then how do you define "pretty big"?

I had a look at osn before you moved to launchpad and didn't see any
significant fixes to the osmdroid part of the code. Since then I fixed
a lot of little things in osmdroid to make it more stable.

My top priority now is to make osmdroid a library but I'm a bit
reluctant to make myself the main developer and start rewriting the
whole thing. This shouldn't be too much work but it's a fundamental
change. Plus I need to find time.

Will be in touch!

Neil



On Tue, May 11, 2010 at 14:58, Kieran Fleming <[email protected]> wrote:
> Hi Neil,
> Making osmdroid a library would be fantastic and we'll definitely use
> it. At the moment our codebase has diverged from osmdroid so Steve is
> trying to merge the changes in. We've also made some fixes to osmdroid
> that you may want - we've definitely fixed bug 40 from your list plus a
> lot of little changes. Our code is at:
> https://code.launchpad.net/opensatnav
> Please don't hesitate to ask if you need help with it.
>
> I had a quick look at MySpeed and it looks like a great idea. Is there a
> link for the source somewhere? In particular I'd like to know how you
> detected if the user has gone off the route as I'm not very happy about
> our code in this area.
>
> Thanks for the email, and I hope we keep in touch.
> Cheers,
> Kieran
>
> On Tue, 2010-05-11 at 10:05 +0200, Neil Boyd wrote:
>> Hello
>>
>> I see that you're keeping track of the osmdroid project. I'm one of
>> the contributors to that project. I hope we can keep osmdroid useable
>> as an external library without you feeling the need to create your own
>> branch. Hopefully this issue will help with that:
>> http://code.google.com/p/osmdroid/issues/detail?id=43
>>
>> I also have routing on my own app - perhaps I can share some code with
>> you if you see anything you like:
>> http://L6n.org/myspeed/
>>
>> Regards
>>
>> Neil
>>
>>
>> On Mon, May 10, 2010 at 16:54, Kieran Fleming <[email protected]> 
>> wrote:
>> > Hi
>> > These are the minutes from Steve and my meeting today. The nicely
>> > formatted view is here:
>> > http://openetherpad.org/x4phODf7yt
>> > (I don't know how long this will stay up for though)
>> >
>> > Would 2pm UK time on a Sunday be a better choice for another meeting?
>> > Steve and I were talking about having them every 2 weeks at least while
>> > the GSoC is running.
>> >
>> > Cheers,
>> > Kieran
>> >
>> > Much more reliable trace recording tools, including voice and  picture
>> > notes. This will take about a week, as much of the work is  already done
>> > by osmtracker and opensatnav.
>> >      * http://wiki.openstreetmap.org/w/images/4/4e/JOSM-Jan-2009.jpg
>> >      * Consider talking to JOSM or Merkaartor people about anything we
>> >        need to support this
>> >
>> > Add in support for alternative online routers other than yournavigation.
>> > This will take less than a week, and will allow for the  next point:
>> >      * Use OpenRouteService ATM, there are no fully open source routers
>> >        available. Cloudmade is probably not a good choice due to
>> >        restrictive license terms. OSM community is going to provide
>> >        something soon enough.
>> >
>> > Turn-by-turn driving directions
>> > This is the larger task, and  will bring it much closer to the standard
>> > set by AndNav and other satnav  products. This will include text to
>> > speech, using the standard android  libraries if using Android 2.0+, or
>> > the TTS download from the  marketplace. I estimate that this will take 3
>> > weeks for the basic visual  view, and algorithms to detect moving off
>> > route and moving past a step  in the journey, and another 2 weeks to get
>> > the TTS implemented,  including intelligent TTS when there are several
>> > instructions close  together in a route.
>> >      * Look into AndNav code but probably only use overlay rendering
>> >        code
>> > POI search around a place (choose a destination, then search  for POIs
>> > near it such as pubs or atms, using the Nominatum geocoder).  This will
>> > take about a week.
>> >      * Long click on map, goes to new menu
>> >      * POIs near here
>> >      * Route to here
>> >      * Visual search results - tapping a result brings up further
>> >        information about the POI, allows the user to ring the location
>> >        if there's a phone number in OSM and other nice use cases
>> >
>> > Finally, there are two features that will be actually  innovative:
>> >      * A widget view, so that someone can leave the application closed,
>> >        but still see driving directions in a 2x2 android widget, and
>> >        thus  change music or other tasks without loosing view of the
>> >        driving  directions. This will take another week.
>> >      * Main app can maybe send Intents to the widget so it can update
>> >        its view
>> >      * Driving directions to the status bar when the application is
>> >        minimised: driving directions will be posted as notifications,
>> >        so even  if someone is in a call, they will still be able to see
>> >        where to go.  This will take about a week (could be done much
>> >        more quickly, but will  take a week to get perfect)
>> >      * Android might kill the app when it gets a call, requires
>> >        research (maybe OSN needs to always run as a service)
>> >
>> >
>> > 1.0 Roadmap (try for release in about 6 weeks)
>> >      * Kieran will add autocomplete to text search
>> >      * Kieran will get translations working in Launchpad
>> >      * Kieran will have a look at stats service - it seems that it
>> >        doesn't close when the program does and wastes battery
>> >      * bug 574176 - out of memory - Steve will have a look at what new
>> >        osmdroid does to try to alleviate this problem
>> >      * database is locked bug -
>> >        https://bugs.staging.launchpad.net/opensatnav/+bug/574187 -
>> >        Kieran will try Guillaume's suggestion as this hasn't been
>> >        tried?
>> >
>> >
>> > Future
>> > Release new 1.x version as each GSoC feature is complete
>> > Meetings - Sundays at around 2pm UK time if it's a good time for
>> > everyone else
>> >
>> >
>> >
>> >
>> > _______________________________________________
>> > Osn-dev mailing list
>> > [email protected]
>> > https://lists.kentgeek.org/mailman/listinfo/osn-dev
>> >
>> _______________________________________________
>> Osn-dev mailing list
>> [email protected]
>> https://lists.kentgeek.org/mailman/listinfo/osn-dev
>
>
> _______________________________________________
> Osn-dev mailing list
> [email protected]
> https://lists.kentgeek.org/mailman/listinfo/osn-dev
>
_______________________________________________
Osn-dev mailing list
[email protected]
https://lists.kentgeek.org/mailman/listinfo/osn-dev

Reply via email to