Topic/fix pdef clear using Psync #4792
Conversation
Where the old source pattern is shorter than the quant value, Pdef / EventPatternProxy would start the new pattern too early. As Psync both cuts and extends, we can use this feature here.
telephon
changed the title
|
Was the Psync problem that a zero-duration pattern would roundUp to 0? I'd like to understand it before reviewing. |
yes. In My strategy was to acknowledge that in such cases we implicitly assume a |
| var <>quant, <>maxdur, <>tolerance; | ||
| *new { arg pattern, quant, maxdur, tolerance = 0.001; | ||
| ^super.new(pattern).quant_(quant).maxdur_(maxdur).tolerance_(tolerance) | ||
| var <>quant, <>maxdur, <>tolerance, <>mindur; |
There was a problem hiding this comment.
Ok, I see the reason for mindur.
My only suggestion here is to document the new argument.
The ordering is unfortunate but alternatives are worse.
| } | ||
| storeArgs { ^[pattern,quant,maxdur,tolerance] } | ||
| storeArgs { ^[pattern,quant,maxdur,tolerance,mindur] } |
There was a problem hiding this comment.
Let's add spaces after commas (since this is already touching that line).
Sorry, I had tried to include this in the other review but it got lost.
|
Should this go into 3.11 ? |
That would be great, if it can! Thanks again for this. |
Based on the track record of the patterns library i think we should wait until after 3.11.0 is released to move this change into the 3.11 branch. then it can go through a beta cycle. |
|
but as release manager it's ultimately @joshpar 's decision |
|
@telephon quick reminder to add labels to your PRs! |
|
I think 3.11.1 is best. 3.11 RC is about hours away. I think a 3.11.1 release in the very near future is part of the plan in my mind. |
Purpose and Motivation
Discussion see #4779
Types of changes
To-do list