ALERT: TrainingPeaks Sync Deprecates from December 7th

4,277 views
Skip to first unread message

Mark Liversedge

unread,
Nov 27, 2016, 6:30:43 AM11/27/16
to golden-cheetah-users
Hi,


As part of this change TrainingPeaks are also revoking read access to users data from GoldenCheetah (and many other tools and services).

They are happy for us to continue to support device download and file format conversion and upload to their service, but we will not play along. They have deprecated their own Device Agent preferring to push responsibility for device and file format support to us. They claim that this is because the FIT file format is now a de-facto standard and they no longer need to support new devices and file formats. Good luck with that.

I will no longer recommend any users pay for their closed service and will deprecate support on the 7th.

Mark

Mark Liversedge

unread,
Nov 27, 2016, 7:17:35 AM11/27/16
to golden-cheetah-users
The code has now been removed from the codebase, here is the commit message.

commit 6ba6d079b97d79fcbf0debbc3fffe81098f184f3 (HEAD -> master)
Author: Mark Liversedge <@gmail.com>
Date:   Sun Nov 27 12:10:57 2016 +0000

    Deprecate TrainingPeaks
    
    .. from December 7th 2016 read access (for downloading
       from TrainingPeaks) has been revoked, even for paid
       up members.
    
    .. write access (upload) is not revoked.
    
    .. This is essentially anti-competitive behaviour and
       returns TrainingPeaks and their products, once again
       to that of a 'closed' product.
    
    .. We will no longer support their service since we
       cannot and will not promote such behaviours now
       or ever.
    
    .. We are now considering the best way to purge all
       other references to their trademarks and remove
       any indication that their products are endorsed.

Mark Liversedge

unread,
Nov 29, 2016, 1:44:33 PM11/29/16
to golden-cheetah-users
I will be removing all references to TrainingPeaks and related trademarks in GoldenCheetah v3.4.1

I realise that for many of our users this will be disappointing. I would suggest, politely, that perhaps you should consider paying for the TrainingPeaks products to get access to them if they are of value to you. I realise many will think this is cutting off my nose to spite my face; that we will lose users. I would counter that this is an open source project that is not for profit and is not intended as a competitor to their tools.

I also realise that many of you may suggest that I am doing this because of my well known dislike of Andrew Coggan. I would counter that I added support for TrainingPeaks and TSS et al long after making public my view of his online behaviour and self-aggrandizement. Further, I was quite happy to port GC to the new TrainingPeaks API, until I realised they did not allow users to access their own data using 3rd party tools.

It is by closing their system to all 3rd parties, limiting download of data to their own tools, that they closed the door to us. I thought the arguments regarding open vs closed systems had long been resolved. Open always wins, Oracle, Google, IBM, Microsoft or more pertinently Saris, Garmin, SRM, Tacx have all recognised the value in being open and benefitted from it. 

Here is why we no longer support TrainingPeaks:

1. We do not endorse the "science" behind the terms. 

Despite the furious claims of the originator, there is zero evidence of the validity of TSS, NP, IF, VI, CTL, ATL, TSB and FTP. There are a handful of references to these terms, but no study has ever been published that verifies any of the numerous claims made.

The marketing of these terms has made them popular with amateurs the world over and are considered the lingua-franca of cycling with power. This is because they are promoted in tools such as GC and related. They are not at all widely used by professionals and sports scientists.

2. We did not "steal someone else's ideas"

I added these to GoldenCheetah in 2010 as many of our users were keen to use them, I did so with permission from Gear Fisher (CEO Peaksware). This permission has now been rescinded and I am more than happy to honour this.

3. We promote real science

There is no need to use them anyway. There is plenty of science in the literature that has been validated and proven to be robust. The vast output of research stemming from the scientific method is our main focus. Our challenge is to put this science in the hands of amateurs and practitioners in a way they can use them.

4. We value integrity, honesty and openness

GoldenCheetah has always been an open platform; the source code is freely available, our tools go to great lengths to be transparent and not contain "secret sauce". We are honest about the shortcomings of much of the content of GC. By using these terms and algorithms we have long tainted our integrity. But no longer.

I strongly advise that all users take measures to extract their data from TrainingPeaks before Dec 7th since they have revoked access to ALL 3rd party applications, not just GC. There is no batch export function in their products. Vendor lock-in is a real risk, you have been warned.

--
Mark

Mark Liversedge

unread,
Nov 29, 2016, 4:11:11 PM11/29/16
to golden-cheetah-users
It is possible to batch export data from TrainingPeaks.


You cannot use a date range longer that a year, so do it for each year of data. This appears to work for free users too, not just premium.

My apologies, I did not think to look in settings for this option.

Mark
Reply all
Reply to author
Forward
This conversation is locked
You cannot reply and perform actions on locked conversations.
0 new messages