SuuntoPlus Improvement Suggestions

[Woodster]

Hi All,

Here’s a few suggestions for improvements of the SuuntoPlus development environment - I’ve not heard what the future roadmap looks but here’s what I have gathered based on my few months of experience so far :

floating point duration and distance

The current solution is providing duration and meters rounded to the nearest second and meter respectively despite the datatype being a float.

For precision pacing it would be better to have floating point precision

distance only seems to update every 2s

this creates a jitter effect when pulling duration/distance & using this to calculate speed from

onInterval() event not triggered and the start of an interval set

but is triggered at the end of an interval set. This causes issues with SuuntoPlus app logic thinking you’re “running” again.

See posting here:
https://forum.suunto.com/topic/15908/oninterval-event-not-triggered-on-starting-a-set-of-intervals

be able to know workout phase instead of inferring it from events and making assumptions such as the warmup happens first

There is currently ( as far as I can see ) no way to determine if you’re in a workout and what phase ( warm-up/run/rest/etc ) you’re in.

know what the current workout target pace/power/heart rate is

So we can make some cool workout screens tailored to the sport you’re doing

make the watch simulator closer to the watch hardware you’re trying to test for example if the app wouldn’t load due to an issue compiling :

ERR DUKTAPE : JSalloc:4904 oversize
ERR DUKTAPE : Compiling js failed: Error: 1

disclaimer: this is based on my own experience - feel free to correct anything that’s already solved or can easily work around

[elbee]

@Woodster said:

For precision pacing it would be better to have floating point precision

Even the best gnss isn’t accurate enough to justify floating point precision.

[Woodster]

@elbee said:

E Offline
elbee
8 minutes ago

@Woodster said:

For precision pacing it would be better to have floating point precision
Even the best gnss isn’t accurate enough to justify floating point precision.

My angle is that a 1m accuracy on the ground is fine but this can be stored with all the natural variation included in the data e.g if you’re travelling at 1m/s you might get:

1.134
2.092
2.998
4.001
etc

And keep it unaltered like an analogue value ( Garmin does this - I develop on there too ).

Truncating the natural variation can lead to temporal artifacts.

The same goes for time - it should be possible to capture the milliseconds in your code so you can compute accurate averages etc.

I might try and compute distance myself using LAT/LON raw data and see if I can improve on it.

[Łukasz Szmigiel]

@Woodster said:

make the watch simulator closer to the watch hardware

That would require running the actual firmware.