Hi Hubert,
It seems no different than orginal after estiminated.
The last instance:
No of Lap Trigger (1km) Lap dist GPS inaccuracy
---------- --------------- ----------- -----------------
0 1000m 1006m +6
1 994m 999m +5
2 1001m 1004m +3
3 992m 997m +5
4 1003m 1005m +2
5 995m 1001m +6
Total comulatived error is 6-1+4-3+5+1=12m;
About yours:
No of Lap Trigger Lap dist (T+GPS inaccuracy)
(T-GPS inaccuracy) Diff
---------- ------------- --------- -----
0 1000m 1006m + 6
1 994m 999m -1
2 995m 998m -2
3 997m 1002m +2
4 995m 997m -3
5 998m 1004m +4
Total comulatived error is +6-1-2+2-3+4=12m;
Needs to check more... 
Regards, Geoffrey 
Hi Geoffrey,
thank you for thinking about my suggestion for the cummulative error in auto lap function !
I do
not agree with you, that there is no difference between the two methods !
First of all: Please re-check your calculation of the cumulative error in my suggestion:
Total comulatived error is +6-1-2+2-3+4=12m;
Needs to check more...
I learned at school, that +6-1-2+2-3+4=6m (not 12m !)
And this is:
- Less than your cumulative error,
- always the same as the last GPS inaccuracy
- and that means, there is no cumulation !!
Maybe you think, this is a result of this special example
and if there is an error of 6m or 12m really doesn't matter.
But it's more than that:
Let me give two more examples.
First is an example, where GPS error is nearly constant and toggles from 6m to 8m:
With your algorithm:
No of Lap Trigger (1km) GPS Error Lap dist Cumm Error
---------- ------------- --------- -------- -----------
0 1000m + 6m 1006m + 6
1 994m + 8m 1002m + 8
2 998m + 6m 1004m + 12
3 996m + 8m 1004m + 16
4 996m + 6m 1002m + 18
5 998m + 8m 1006m + 24
With my algorithm:
No of Lap Trigger (1km) GPS Error Lap dist Cumm Error
---------- ------------- --------- -------- -----------
0 1000m + 6m 1006m + 6
1 994m + 8m 1002m + 8
2 992m + 6m 998m + 6
3 994m + 8m 1002m + 8
4 992m + 6m 998m + 6
5 994m + 8m 1002m + 8
As you can see, yours is cumulating, mine is not !
And the total error with my method is always exactly the last GPS error. No more.
Need one more example ?
This is an example, where GPS error toggles from nearly okay (1m) to bad (13m):
With your algorithm:
No of Lap Trigger (1km) GPS Error Lap dist Cumm Error
---------- ------------- --------- -------- -----------
0 1000m + 1m 1001m + 1
1 999m + 13m 1012m + 13
2 988m + 1m 989m + 2
3 1011m + 13m 1024m + 26
4 976m + 1m 977m + 3
5 1023m + 13m 1036m + 39
With my algorithm:
No of Lap Trigger (1km) GPS Error Lap dist Cumm Error
---------- ------------- --------- -------- -----------
0 1000m + 1m 1001m + 1
1 999m + 13m 1012m + 13
2 987m + 1m 988m + 1
3 999m + 13m 1012m + 13
4 987m + 1m 988m + 1
5 999m + 13m 1012m + 13
Here also yours is cumulating, mine is not.
And even more: Your method tends to build up ! Errors are getting bigger, effective
laps are much too short or much too long.
As a result, I'm sure my suggestions can solve the current problem with the
algorithm used in firmware X and I would like to ask you to check it again.
(If you decide to use my suggestion, you don't need to mention the famous
Hubert algorithm in your advertising...) 
Best regards, Hubert
Author
Topic: Firmware X release!! (Read 33772 times)