[Loadstone] Question for Shawn
shawn at odyssey.cm.nu
Wed Nov 22 01:40:29 GMT 2006
The heading announcement is just a display change, nothing about the
calculation logic was changed. The problems you're having, and will continue
to have, are with your receiver having its static navigation feature
switched on. All the program hacks in the world aren't going to fix this.
For things to work correctly loadstone needs the latitude, longitude, speed,
and heading from the gps receiver. With the receiver's static navigation on
half the needed information is dropping out at low speeds.
The checkpoint announcements need a speed value from the gps receiver in
order to work. Without this the checking isn't done. This is because there's
no point doing point checks when you're standing still. The checks could be
done based on latitude/longitude but when you're stationary the gps position
will drift and that drifting can give random false point announcements.
Reducing the max approach distance below 10 meters wouldn't effect this
since that's about the normal accuracy for gps anyway. The max approach
distance setting should not come in to effect in normal opperation anyway.
The phone does the point checks roughly every 5 seconds so reducing the
approach time wouldn't help either. If you have a lot of checkpoints close
together then you probably will get multiple announcements especially if
you're approaching the group at high speed. In that case you may want to
consider unchecking some points or having a different checkpoint list file
for the high speed route.
The chained point mode will be similar to the checkpoint mode but the points
will be announced in the order they're checked. There will have to be some
controls for inserting a point into the list and starting a route and
reversing the route direction etc. All that's still being worked out.
The static threshold setting seems correct. The problem is your static
navigation again. You probably want to have static threshold set to 0.
On Tue, 21 Nov 2006, Aedan O'Meara wrote:
> Hi Shawn,
> Would I be correct in thinking that the addition of the distance and
> heading to the ordinary point announcing in v0.65, is making it possible
> that points that used to be announced in v0.63 are not now announced
> because the walking speed is not in excess of 4kph?
> I am still experiencing a proper display of points on the joystick
> particularly in the city whilst walking without any announcing of my
> approach to these points. These points are progressing correctly in
> distance on the joystick. There would be as many as 8 satellites being
> seen by the gps. I know you explained this because Loadstone is using
> the last known heading, but if the heading announcement was turned off
> is ift possible to have these points pre announced?
> Secondly, Would the removal of the time aspect of measuring the approach
> distance and the ability to reduce the max approach distance below 10m
> not arrive at the same effect as building the sequential point route
> system we previously requested?
> I am coming to the conclusion that the excessive announcing of too many
> adjacent points is definitely speed related because unfortunately at my
> bus's journey end it goes downhill and picks up quite considerable speed
> and the points it starts calling out would all be within the new
> calculated radius.
> This would not matter either if it only called out the points at 12
> but it calls them out all around.
> At present the speed setting cannot be dropped below 10 secs.
> So, what I am asking is would it be possible to give us experimentally a
> version of Loadstone with an option to drop the max speed to 0?
> Thirdly, I have a feeling that the new setting of static threshold is
> not returning to full 0 after it has been put up to any other figure. Is
> it possible that you have written a default setting here that is above
> 0? The reason I ask is I am convinced that judging over a period of time
> The v0.63 definitely gave better results when walking.
> I have checked the Loadstone.ini file and it is set to 0 there ok .
> Finally, if we proceed to the sequential point route idea could this not
> be simply activated by adding another field to the database in which one
> stores the sequential number of the point to be announced, so that when
> that point is approached it is not announced unless it is in correct
> sequence as stored in a route file?
> The rest of the engine can be left alone.
> Regards to all,
> Loadstone mailing list
> Loadstone at loadstone-gps.com
More information about the Loadstone