I guessed that 
That’s fine, but no I’ve not heard from anyone except you, Paul Beach (audacity-nyquist list) and a PM from waxcylinder.
Here’s the link to the track: http://audacity.easyspacepro.com/examples/Track-No02-beeped.ogg
and a link to Koz’s famous piano solo: http://audacity.easyspacepro.com/examples/piano.flac
That’s easy - there’s no problem with generated tones because the amplitude is constant.
If I change the amplification scale to dB, then this can be very easily achieved by simply generating the tones at amplitude 1.0, then, if they require a fade from a target level of, say -12dB to say -3dB, they simply set the start amplification to -12dB and the end amplification to -3dB.
Hands up who would expect a fade effect to dynamically compress the audio…
but that is only one symptom of the problem.
If we look at the last example, the “sensible” thing for setting a target level would be to base it on the peak value for the last 5% of the waveform, but what about Koz’s piano? the last 5% of Koz’s piano is almost silent. What if the audio is completely silent at the end? No matter how much you amplify silence it is still silence.
How do I calculate the required amplification for a set target level (any target level) with this sample:
According to David Bailes, they would just use text input, but this is why he particularly wanted the range of values to be specified on the left of the control rather than on the right (screen readers do not always read text on the right of the control).
It would mean an additional control (could be a “choice” drop down menu).
I think having -50dB as silence would be pretty safe as long the user was clearly informed. If, in the unlikely event, a user specifically wanted a fade of -80 dB to 0dB, they could run a fade of -40 dB to 0dB twice (though I think it would be extremely rare that anyone would ever need to do that).
I’ll go for the dB scale (which is what I had originally ) as there does appear to be consensus on this, and it ties in nicely with the Amplify effect.
The comment from Paul Beach on this was “dB is familiar to the ears; however % is easier to see. Try viewing the waveform as dB. It is not very helpful.”
A dB (amplify) version should be up by the end of the week.
BTW, I posted an updated version of my “Noise Gate” here: Noise Gate - #3 by steve
bar any typo’s that need correcting the noise gate is complete.