Garmin Edge1040 with Beta 20.08: no Lap in GC

[mute mutnig]

I use V3.6 Buid Aug. 5 2023 on Manjaro Linux, updated my Garmin Edge 1040 to Beta 20.08.

When I import a fit from Edge to GC no Laps were shown.

If I import the same ride via Strava the Laps were shown.

Debug output:

GoldenCheetah_v3.6_x64.AppImage --debug
QApplication: invalid style override 'gtk' passed, ignoring it.
Available styles: Windows, Fusion
Python loaded [3.7.5 (default, Dec  9 2021, 17:04:37)  [GCC 8.4.0]]
Debug: Sandboxing disabled by user.
Debug: "FitRideFile: unknown global message number 327; ignoring it"
Debug: "FitRideFile: unknown global message number 113; ignoring it"
Debug: "FitRideFile: unknown global message number 288; ignoring it"
Debug: "FitRideFile: unknown global message number 141; ignoring it"
Debug: "FitRideFile: unknown global message number 104; ignoring it"
Debug: "FitRideFile: unknown global message number 147; ignoring it"
Debug: "FitRideFile: unknown global message number 79; ignoring it"
Debug: "FitRideFile: unknown global message number 216; ignoring it"
Debug: "FitRideFile: unknown global message number 326; ignoring it"
Debug: "FitRideFile: unknown record field 137; ignoring it"
Debug: "FitRideFile: unknown record field 107; ignoring it"
Debug: "FitRideFile: unknown record field 144; ignoring it"
Debug: "FitRideFile: unknown record field 138; ignoring it"
Debug: "FitRideFile: unknown global message number 327; ignoring it"
Debug: "FitRideFile: unknown global message number 113; ignoring it"
Debug: "FitRideFile: unknown global message number 288; ignoring it"
Debug: "FitRideFile: unknown global message number 141; ignoring it"
Debug: "FitRideFile: unknown global message number 104; ignoring it"
Debug: "FitRideFile: unknown global message number 147; ignoring it"
Debug: "FitRideFile: unknown global message number 79; ignoring it"
Debug: "FitRideFile: unknown global message number 216; ignoring it"
Debug: "FitRideFile: unknown global message number 326; ignoring it"
Debug: "FitRideFile: unknown record field 137; ignoring it"
Debug: "FitRideFile: unknown record field 107; ignoring it"
Debug: "FitRideFile: unknown record field 144; ignoring it"
Debug: "FitRideFile: unknown record field 138; ignoring it"

Snapshot build does not work:

/tmp/.mount_GoldenDM1Psc/GoldenCheetah: symbol lookup error: /usr/lib/libtiff.so.6: undefined symbol: jpeg12_write_raw_data, version LIBJPEG_8.0

How can I solve the Lap issue?

thx

Helmut

[marcen]

First one i can't help. Maybe it's help to have a FIT file for tests.

I have the same behavior on Arch Linux with this error: /tmp/.mount_GoldenDM1Psc/GoldenCheetah: symbol lookup error: /usr/lib/libtiff.so.6: undefined symbol: jpeg12_write_raw_data, version LIBJPEG_8.0

What helps:

1. ./GoldenCheetah_v3.6-DEV_x64.AppImage --appimage-extract

2. search under ../squashfs-root/lib/ for libjpeg* and delete

3. start GC ../squashfs-root/AppRun

[mute mutnig]

thx for the hint - I will try it

attached the fit - export from Garmin Connect.

[Ale Martinez]

El lunes, 15 de enero de 2024 a la(s) 3:43:54 a.m. UTC-3, mute mutnig escribió:

attached the fit - export from Garmin Connect.

I imported that FIT file to latest version and effectively there are no laps detected by GC, so something relevant has changed, the good news is the logic Mark's added for laps details as XData seems to be working, so likely it is an easy fix to create GC laps:

[mute mutnig]

Thank you both!

marcen's work around let me use the latest version, and after import I see the LAP tap.

[Mark Liversedge]

Can you share the interval information you see on Garmin Connect please- i.e. start and finish time for each interval

[mute mutnig]

link to Garmin: https://connect.garmin.com/modern/activity/13502531764

[Mark Liversedge]

I don't understand that lap data since the lap starts overlap and go back in time.

Did you edit the interval data in Garmin Connect?

Mark

[mute mutnig]

No I did not modify anything, but I will upload the original fit from my Edge later when I am at home.

[mute mutnig]

attached the fit directly copied von Edge.

[mute mutnig]

as still GoldenCheeta does not show the laps I asked in Garmin Beta forum

https://forums.garmin.com/beta-program/edge-1040-series/public-beta-bug-reports/i/edge-1040-series---20-xx-bug-reports/20-15-change-of-lap-information-in-fit-file-causes-troubles-if-imported-to-other-apps

Answer:

aweatherall 5 hours ago

I would assume that these changes are permanent .

There are two types of changes. Data field names. Garmin provided the old and new data field names  for around 5 years in the FIT file to allow 3rd parties to switch over before they deprecated the older fields.

The other change is the sequence that the data is written out into the FIT file. This is new in the beta builds. Provided that applications are using time information correctly and not making assumptions about the record order in the FIT file they should be OK. In other words they should be using things like start time for laps, rather than assuming that the record timestamp is the start time of the lap. The record timestamp for a lap is just when it gets written to the FIT file. This used to be when it occurred in the activity so the timestamp would equal start time. Things like laps and splits are now stored in memory and get written out when the activity is saved. This requires applications to use the lap start time. This is what they should always have been doing, but they were able to get away with using timestamp instead until this change.

Can you please check if this could be the reason for not seeing laps?

Thanks!

Helmut

[Mark Liversedge]

On Sunday 18 February 2024 at 20:04:47 UTC mut...@gmail.com wrote:

as still GoldenCheeta does not show the laps I asked in Garmin Beta forum

https://forums.garmin.com/beta-program/edge-1040-series/public-beta-bug-reports/i/edge-1040-series---20-xx-bug-reports/20-15-change-of-lap-information-in-fit-file-causes-troubles-if-imported-to-other-apps

Answer:

aweatherall 5 hours ago

I would assume that these changes are permanent .

There are two types of changes. Data field names. Garmin provided the old and new data field names  for around 5 years in the FIT file to allow 3rd parties to switch over before they deprecated the older fields.

The other change is the sequence that the data is written out into the FIT file. This is new in the beta builds. Provided that applications are using time information correctly and not making assumptions about the record order in the FIT file they should be OK. In other words they should be using things like start time for laps, rather than assuming that the record timestamp is the start time of the lap. The record timestamp for a lap is just when it gets written to the FIT file. This used to be when it occurred in the activity so the timestamp would equal start time. Things like laps and splits are now stored in memory and get written out when the activity is saved. This requires applications to use the lap start time. This is what they should always have been doing, but they were able to get away with using timestamp instead until this change.

This is highly plausible- the way timestamps and start times are handled in the FIT parser is not at all clean. There is definitely room for improvement in this space. I did see some issues regarding this when working on the generic parsing. Will take another look.

Mark 

[mute mutnig]

Dear Mark, thank you for taking care!

Helmut

[Paul Greenberg]

FYI: I looks like the same thing is going on with my Garmin 965 which was just updated to software version 18:22 last night.

[Ale Martinez]

El miércoles, 6 de marzo de 2024 a la(s) 12:56:00 p.m. UTC-3, Paul Greenberg escribió:

FYI: I looks like the same thing is going on with my Garmin 965 which was just updated to software version 18:22 last night.

Can you attach a sample file for testing? 

[Paul Greenberg]

Ale:

Happy to oblige!  Please see attached file.

Best Regards,

Paul

[Paul Greenberg]

Also, here is a file from the same device from a couple of days ago before the upgrade which GC could read the laps for comparison.  Both workouts are runs.

[Paul Greenberg]

Fit File Repair tool was able to export into the older format which shows the laps in GC. Also attached is a screen shot from Fit File Repair which gives a hint on the changes to the .fit file format.

-Paul

[Ale Martinez]

Thank you, that's very useful, I will open a GitHub issue for tracking this.

[Ale Martinez]

El miércoles, 6 de marzo de 2024 a la(s) 5:35:43 p.m. UTC-3, Ale Martinez escribió:

Thank you, that's very useful, I will open a GitHub issue for tracking this.

https://github.com/GoldenCheetah/GoldenCheetah/issues/4451, a lap swim file from a 965 with the new firmware would be useful too.

[Paul Greenberg]

Let me reach out to my swim friend with a 965 to get a file. I only bike and run!

[Ale Martinez]

El jueves, 7 de marzo de 2024 a la(s) 2:09:56 p.m. UTC-3, Paul Greenberg escribió:

Let me reach out to my swim friend with a 965 to get a file. I only bike and run!

Great! Open water swimming is similar to bike/run, but lap/pool swimming has a different structure.

BTW, besides to convert the FIT file as you did, the other workaround until we fix this is to use https://github.com/GoldenCheetah/GoldenCheetah/wiki/UG_Side-Bar_Activities-view#intervals---find-intervals to recreate the intervals. 

[Paul Greenberg]

Good to know about the lap/pool swimming vs open water swim differences. My friend is an open water swimmer so let me reach out to a swim group to get a pool .fit file.

[Paul Greenberg]

Here are open swim and lap swim in both the classic and modern formats. -Paul

[Ale Martinez]

Thank you, Paul, there are no problems with Pool/Lap swimming files, this was somewhat I expected because we create laps from length records in that case.

Files labeled Open also corresponds to pool swims, they don't have GPS data but length records, but this should not be a problem, OW swims format is like Bike/Runs/etc. 
Again, thank you very much!

[Paul Greenberg]

Ale:

I am happy to help in my small way since my programming skills stopped with FORTRAN on punch cards!  

Paul

[Ale Martinez]

El viernes, 8 de marzo de 2024 a la(s) 1:08:20 p.m. UTC-3, Paul Greenberg escribió:

Ale:

I am happy to help in my small way since my programming skills stopped with FORTRAN on punch cards!  

Been there, done that ;)

Starting from scratch the new structure would be far easier to parse, but GC parser is tailored for the old format including a bunch of peculiarities from non Garmin sources.

The main offender seems to be start date and time, in the old format GC got it from the 1st record since the natura sources (activity/session) came at the end.

Lap decoding also checks the present of sample records and uses timestamp as the lap stop.

Changing these 3 items laps seems to be ok, but we need more testing with new and old files.

I have not looked at the new decodeGeneric for now, may be some adjustment be required there too.

[mute mutnig]

2 Files from my mtb ride today: 

Coros Pace 2 Firmware V3.0309.0 (Coros...fit)
Garmin Edge 1040 Solar FW20.18 (.zip)

tried to press the lap buttons every time on both devices - lap should be on nearly the same location, but not the same ride time because I started the Coros watch little earlier.

The Laps in the Coros fit are recognized by GC, Edge not.

If you need some older Edge files - just tell me what you need.

(only cycling indoor & Road & MTB some I recorded in parallel Coros & Edge)

Helmut

[Ale Martinez]

El domingo, 10 de marzo de 2024 a la(s) 1:30:43 p.m. UTC-3, mute mutnig escribió:

2 Files from my mtb ride today: 

Coros Pace 2 Firmware V3.0309.0 (Coros...fit)
Garmin Edge 1040 Solar FW20.18 (.zip)

tried to press the lap buttons every time on both devices - lap should be on nearly the same location, but not the same ride time because I started the Coros watch little earlier.

The Laps in the Coros fit are recognized by GC, Edge not.

Thank you for the files, importing the Garmin one using my modified GC I get this json file, lap seems similar to the Coros one, you can import it for checking: 

[mute mutnig]

it is perfekt!

compared the laps to Garmin Connect - just some minor deviations in averages.

thx

Helmut

[Ale Martinez]

There is a GitHub Pull Request with these changes for testing, Windows users can download an installer from https://ci.appveyor.com/project/Joern-R/goldencheetah-knhd8/builds/49371041/artifacts, but beware PR builds don't include cloud keys so they have no Strava/CloudDB/etc. access.

[Ale Martinez]

El domingo, 10 de marzo de 2024 a la(s) 9:57:26 p.m. UTC-3, Ale Martinez escribió:

There is a GitHub Pull Request with these changes for testing, Windows users can download an installer from https://ci.appveyor.com/project/Joern-R/goldencheetah-knhd8/builds/49371041/artifacts, but beware PR builds don't include cloud keys so they have no Strava/CloudDB/etc. access.

The above Pull Request is merged now, a Windows installer can be downloaded from https://ci.appveyor.com/project/Joern-R/goldencheetah-knhd8/builds/49378281/artifacts, this is similar to snapshot builds, basically v3.6 plus some fixes and small new features, no big changes yet. Feedback is welcome.

[Paul Greenberg]

Ale:

I am getting this error message when I try to download:  {"message":"Artifacts download limit (1024 MB/day) exceeded."}

Am I doing something wrong?

-Paul

[Ale Martinez]

El lunes, 11 de marzo de 2024 a la(s) 5:22:59 p.m. UTC-3, Paul Greenberg escribió:

Ale:

I am getting this error message when I try to download:  {"message":"Artifacts download limit (1024 MB/day) exceeded."}

Am I doing something wrong?

No, it is a free service which has a daily download quota that was already used, it resets every day, try again later.

[Paul Greenberg]

👍

--
_______________________________________________
Golden-Cheetah-Users mailing list
golden-che...@googlegroups.com
http://groups.google.com/group/golden-cheetah-users?hl=en
---
You received this message because you are subscribed to a topic in the Google Groups "golden-cheetah-users" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/golden-cheetah-users/dbIRjYtCBoU/unsubscribe.
To unsubscribe from this group and all its topics, send an email to golden-cheetah-u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/golden-cheetah-users/3bdab8af-5f88-4e8b-96e8-8588e8544d50n%40googlegroups.com.

--

Paul S. Greenberg
30 Owenoke Park
Westport, CT 06880
(203) 253-7436

[Ale Martinez]

I updated snapshot builds in the main site including this fix to make testing easier, AppVeyor artifact download is not intended for multiple user trying to download simultaneously.

[mute mutnig]

Good Morning!

thank you for the snapshot - my try to compile on my own was not successful...

I tried with one random activity with Edge on Beta FW.20.13 - with the new GC snapshot the laps were recognized: Perfekt!

I compared the laps to garmin connect - there are some small differences in the some average values, e.g. average power

the txt file shows the Garmin connect values - fit is attached.

I am not claiming something from you - I just want to provide some help to allow you to make this amazing tool even better!

Thank you very much for your work!

Helmut

[Ale Martinez]

El miércoles, 13 de marzo de 2024 a la(s) 2:48:23 a.m. UTC-3, mute mutnig escribió:

Good Morning!

thank you for the snapshot - my try to compile on my own was not successful...

I tried with one random activity with Edge on Beta FW.20.13 - with the new GC snapshot the laps were recognized: Perfekt!

Thank you for the feedback, lap start and stop time seems to be aligned up to the second as before.

 

I compared the laps to garmin connect - there are some small differences in the some average values, e.g. average power

the txt file shows the Garmin connect values - fit is attached.

This change only affects lap start/stop times, all other lap metrics are computed by GC as before and there are reasons explained in the wiki and discussed in other threads about the reason the may not align exactly with Garmin ones, I have no intention to discuss this here again.

If you go to Data > Raw Data > LAPS you will find Garmin metrics imported from the FIT file, this is a new feature Mark's added after v3.6 release.

[mute mutnig]

Dear Ale, 

I do not see this "LAP" at Data/Raw Data if I import my Edge file with the build from yesterday night.

But I see LAP when I import the fit from my coros watch.

Both shows the laps in the Intervals window.

Helmut

[Ale Martinez]

El miércoles, 13 de marzo de 2024 a la(s) 10:41:43 a.m. UTC-3, mute mutnig escribió:

Dear Ale, 

I do not see this "LAP" at Data/Raw Data if I import my Edge file with the build from yesterday night.

It works for the other files provided in this thread I used for testing, including your previous one, I will have to look at this one... eventually 

[Paul Greenberg]

The new snapshot build worked perfectly for me today with my run recorded on my Garmin 965.

Ale, thanks again for updating GC!

-Paul

[Ale Martinez]

El miércoles, 13 de marzo de 2024 a la(s) 10:41:43 a.m. UTC-3, mute mutnig escribió:

Dear Ale, 

I do not see this "LAP" at Data/Raw Data if I import my Edge file with the build from yesterday night.

It is weird because there no SESSION tab either when importing this file, I can reproduce this behavior using the previous snapshot build without the fix discussed in this thread, so it doesn't seem to be related to the changes I made. 

[mute mutnig]

Dear Ale, 
since some time I am on v20.19 with my edge1040 solar, today I did a workout from a Garmin trainingsplan on my Neo 2T indoor trainer - this time I get in GC  the lap and the session tap (all my other rides don't show these taps). GC on Linux, snapshot from 12th March.

fit file from Edge attached.

[Ale Martinez]

El martes, 2 de abril de 2024 a la(s) 10:30:33 a.m. UTC-3, mute mutnig escribió:

Dear Ale, 

since some time I am on v20.19 with my edge1040 solar, today I did a workout from a Garmin trainingsplan on my Neo 2T indoor trainer - this time I get in GC  the lap and the session tap (all my other rides don't show these taps). 

Again, I have no idea which the problem may be.

These tabs are a new feature (https://github.com/GoldenCheetah/GoldenCheetah/issues/4416) I have no experience with.