Request testing and feedback on proposed new features

[Ale Martinez]

Hi, if you browse our GitHub repository you will notice we have several pending Pull Requests (https://github.com/GoldenCheetah/GoldenCheetah/pulls), these are proposed new features.

In case you are interested in any of them I encourage you to give them a try (there is a Windows installer available for download and you can build from source for Linux/macOS) and provide feedback.

If you prefer to wait others do the testing, the features be merged and included in a new release please consider:

a) this may not happen, we will not merge features for which we think there is no interest, unless we are personally interested, which is not always the case.

b) the moment to provide useful feedback and influence in development is when features are developed and testing, 2/3 years later is likely too late.

Thanks in advance and best regards.

Ale.

[Joachim Kohlhammer]

Ale, thanks for this call to action!
Your points are spot on. Every line of code can create new bugs and will require additional maintenance, every feature will require support.

I would therefore like to take up Ale's call and ask everybody for help on my pull requests:

• https://github.com/GoldenCheetah/GoldenCheetah/pull/4543
  • Eases the creation of power zones by proposing the estimated values
  • This is still under development (mainly the UI), and allows strong influence by you!
• https://github.com/GoldenCheetah/GoldenCheetah/pull/4536
  • Improve filtering for the trends view in compare mode
• https://github.com/GoldenCheetah/GoldenCheetah/pull/4534
  • Allows user defined relative seasons

To me, help means feedback, positive as well as negative:

• Do you get the purpose of the feature?
• Do you consider it useful?
• Would you expect things differently?
• Did you actively test the feature? Did it fit your needs?
• Did you find unwanted behaviour or just plain bugs?

Also don't hesitate to look into the other open PRs, there is very interesting stuff!

Otherwise my developments will remain exclusively in my local version - which I would find a great pity pity, because as a user I want to make Goldencheetah an even better product through my contributions.

Regards

 Joachim

[pepe]

 Installed today  ci.5372 and seems awesome, facilitating much CP settings.

 AeT seams overestimated though.

[pepe]

The AeT numbers look pretty good for cycling, for running a little overestimated

[Joachim Kohlhammer]

AeTP is not part of the estimates for any model. Therefore this is calculated within the new code as 85% of the CP (as given by the model). This is also the reason why manually  changing AeTP doesn't signal a deviation from the estimate.

Do you think it would be a benefit to make the factor configurable by sports? Or should the AeTP be calculated differently? Also: should manual changes to the AeTP signal a difference to the model?

[pepe]

Hi Joachim,

AeTP  and AeTHr are usually determined by:

Metabolic test

Lactate ramp test

Run /Cycling   ‘conversational’  test
Run Decoupling test

Calculating  AeTP as a fix percentage of CP  negates the individualised approach of these tests, given the inter-individual variation and same individual variation with training   example: In well Aerobic trained athletes, CP / AeTP is lower than CP / AeTP in less Aerobic trained athletes.

So its better AeT values be introduced manually, unless, there is a Standardised Decoupling test that can be processed automatically for  AeTP and  AeTHr .

Pre calculating it as 80-85% of the CP can go with the explanation that a real test is needed.

Also: should manual changes to the AeTP signal a difference to the model?

Don't think so. 

AeTP and CP represent different metabolic energy systems thresholds

[Joachim Kohlhammer]

Hi Pepe,

thanks for the hints, I will make my thoughts on them. Maybe Lennarts comments on GitHub (he asked for a  preview of the estimate before adoption) can help out. I started working on the following dialog that shows up when adopting an estimate:

AeTP is still calculated as 85% of CP but not automatically adopted (the accept checkbox is not pre-checked) and a tooltip was added to the AeTP info-icon.

For new, prefilled ranges I have to make some more thoughts

Regards

 Joachim

[pepe]

Nice, thank you.

[Rui_B]

You guys are awesome — I'm still struggling to do clever things with Python & R, so I can't be much help here... BUT I'd like to  let you know, that I really appreciate your time and effort to make Golden Cheetah even more awesome. Thank you.

Cheers

[Joachim Kohlhammer]

The semi-automatic zones got updated, again - feedback is welcome. Here is what changed:

• Added a dialog to inspect and accept only some values while adoption
• Added info messages
  • when the model does not provide FTP or PMAX
  • that AeTP is only a very rough estimate
• Added a tolerance in comparison before proposing new values
• Using the following order for defaults when adding a new manual range
  • selected row
  • last row
  • predefined defaults
• Zones-Tab: To prevent crashes, a message is shown instead of the real
• interface if a metric refresh is ongoing
• Changed Pace- and HR-Tabs to use inline editing

[pepe]

Started testing ci.5389, and my first impression is that everything looks awesome.

Thenk you.