Forum: General Topics

Hi Stephen!

I've noticed that for GPX uploaded routes (i.e. without time info) the Graphs page on the web site either is not available at all or it does not display the applet (it hangs and IE must be restarted). I can image it's due to the fact that no elapsed time and speed data are available ... Nevertheless the Graphs page would be very useful for the altitude vs. point plot and for the altitude vs. distance one (at least as far as bike, hiking and running activities are concerned).

Would you please have the time to look into that issue, if not too much complicated (as far as I'm concerned, you can add a dummy constant speed in such cases, if it helps) ?

Thanks a lot in advance!

Cheers,
yano

Hi Yano,

Right, I've had a look at this... The problem with the Graphs page hanging looks to be due in part to the GPX import and Manual Track Creation functions. When a track is imported that doesn't have any timing data, each trackpoint is automatically separated by a short gap (of around 1 second) to avoid 'divisions-by-zero' that might otherwise occur if two consecutive points have identical times. The problem is that if the trackpoints are several hundred metres apart this can result in a very high estimated speed (often > 1000mph!). The Graphing applet has a bug where the y-axis sub-divisions have a maximum gap-limit, so for high speeds the graph tries to plot a silly number of subdivisions, which slows it down considerably (and make it appear to hang).

I still need to fix the graph applet, but I thought it would be more useful to be able to manually override the track speed to a better estimated value. I've therefore created a new page (linked from the Track/Info page) to overwrite the track timing/speed data with estimated values based on the track gradient (ie fast cycling downhill, slow cycling uphill). Its pretty crude, but at least allows the track to be populated with reasonable real-world figures, which should then plot fine in the graphs and map pages.

Hope it helps!

Cheers,
Stephen

(Note I haven't fully tested the code yet, so I'd advise testing it on duplicated or unimportant routes initially!)