Hey all,
I've decided to pull my patch for issue 1986 off the current countdown. After incorporating Keith's suggestions and moving the calculations for a beam's concaveness into Beam_scoring_problem, the patch has changed significantly enough from yesterday to merit a new countdown.
With these additional changes there are no differences from the previous regtest comparisons save beam-broken-classic.ly, where strict-prolongation is more accurate (before, errors in the concaveness calculation led to slight slope discrepancies for strict-prolongation when the slope was flat).
Lastly, there is one regtest change that's been bothering me, and I want to get feedback on it before i push the patch. It is in chord-tremolo.ly (see attached).
The shorter stem in the 1/4 bar doesn't bother me (I actually prefer it, but then again, I know nothing about classical typesetting and don't have access to any of the canonical texts), but it is not consistent with respect to the 2/4 and 3/4 versions. This comes from the fact that the 1/4 version doesn't link up with the stem and thus has a shorter x_span_ than its 2/3 and 3/4 sistren. The fix is easy (something like x_span_ = max (x_span_, distance_between_normal_stems);), but I first want to make sure that this is the right move to make. Specifically, if Ross, Reed, or Gould have anything to say, please let me know!
Cheers,
MS
I've decided to pull my patch for issue 1986 off the current countdown. After incorporating Keith's suggestions and moving the calculations for a beam's concaveness into Beam_scoring_problem, the patch has changed significantly enough from yesterday to merit a new countdown.
With these additional changes there are no differences from the previous regtest comparisons save beam-broken-classic.ly, where strict-prolongation is more accurate (before, errors in the concaveness calculation led to slight slope discrepancies for strict-prolongation when the slope was flat).
Lastly, there is one regtest change that's been bothering me, and I want to get feedback on it before i push the patch. It is in chord-tremolo.ly (see attached).
The shorter stem in the 1/4 bar doesn't bother me (I actually prefer it, but then again, I know nothing about classical typesetting and don't have access to any of the canonical texts), but it is not consistent with respect to the 2/4 and 3/4 versions. This comes from the fact that the 1/4 version doesn't link up with the stem and thus has a shorter x_span_ than its 2/3 and 3/4 sistren. The fix is easy (something like x_span_ = max (x_span_, distance_between_normal_stems);), but I first want to make sure that this is the right move to make. Specifically, if Ross, Reed, or Gould have anything to say, please let me know!
Cheers,
MS