Hello Will and the others,
now I've worked a bit with the new 1.5.8 and I noticed some surprising GUI
behaviour at two places, which I weren't aware off previously. Not sure if
this is actually something changed in 1.5.8, or if it was present in
earlier 1.5.x versions, since I used for a long time 1.4.x versions
Anyway, here are my observations:
(1) When editing envelopes in "free" mode...
- when you delete a node
- then the selection is set to the next but one node to the right
This means, you can not "chew away" a section of an elaborate curve
just by hitting del several times.
Now the question here is: what should be the proper behaviour?
I see two options:
- selecting the node /before/ the deleted one would mean that
consecutive deletes would reduce the curve to the left, and
the selection itself would wander to the left
- selecting the node immediately after the deleted one would
mean that the selection stays at the same point in time and
consecutive deletes would reduce the curve to the right.
(2) Selection after saving a part
- when working on a part other then the first one, with several detail
windows opened; when you save that part by overwriting the original slot
- then the first part gets selected, causing all detail windows to close.
This is quite annoying in practice, since not only you have to re open
several windows (kit, part, sub-part detail) but also you loose the
current positions of those windows. E.g you have to re-arrange two envelope
subwindows when you are comparing envelopers.
For me this is quite a bummer, since I often look at values in the XML
with a Git diff to the original setting, since it can be quite difficult
not to move some node in a controlled way. So basically it breaks the
convenient edit-check-tweak cycle.
Sometimes (say one out of 5) I even get both the part #1 and the current
part selected. But I could not find a clear rule how to reproduce it.
But even in that case, clearly part #1 has the actual selection after
saving, which can be seen by hitting the edit button; it brings up the
state in memory for part #1, and not the part you where working on.
I have re-loaded my instance and re-started jack. So it's at least reproducible
locally. Well, I haven't rebooted my system yet (paranoia mode). Can anyone else
also confirm those observations?
Yoshimi source code is available from either:
Our list archive is at: https://www.freelists.org/archive/yoshimi
To post, email to yoshimi@xxxxxxxxxxxxx