You are reading a single comment by @myownself and its replies.
Click here to read the full conversation.
-
Having used the beta of widadjust and the above-mentioned web page to check the time, I've probably come as close as I can to determining my watch's drift and fixing it. I'm still experimenting with a few details.
I'm surprised that others haven't commented on drift. It would be useful to know other estimates of that.
-
I actually implemented something like that already for this webpage. All measurements are now saved in local storage and you can compare any two of them to get clock drift in PPM.
It's somewhat inaccurate so I'd recommend taking 3-4 measurements in a row to see that they are consistent and removing all except one which seems like best measurement. Then wait at least few hours before repeating that.
Numerist
malaire
@myownself
@malaire this was more a curiosity for me because I haven't noticed click drift large enough to personally care about, but I was thinking you could do something with this to make measuring the drift easier for use in your widget... maybe press a button, it sets the bangle time to computer time and puts a timestamp in a cookie or local storage, then the user can come back at a later point and press another button to calculate the drift? I don't know how long you need to reliably determine the drift, perhaps it could be done without storing anything at all?
Edit: if you made it that easy, you could probably convince people to contribute their clock drifts and you could work out what the range is across bangles.