Xing Tu:
Okay, so that's why ctrl-c has to be done twice to exit. It'd
certainly be useful if you guys can figure this one out.
You can now dump the undo stack and if you started cin from a terminal window, it will display
something like the following there up to 32 undo-s with program addresses that it keeps also.
UndoStack::dump
0 0x68e3af0 record patch 119 0004 *
k 1 0x47b0080 record patch 37413 0004
2 0x68e0bb0 record patch 2802 0004
3 0x469f1b0 record patch 1529 0004
4 0x46c1700 split | cut 1529 0011
5 0x46c1eb0 split | cut 1193 0011
6 0x45a8f10 split | cut 1193 0011
k 7 0x46c2060 split | cut 37076 0011
8 0x469e660 split | cut 2580 0011
9 0x68cf280 split | cut 1071 0011
10 0x46bc070 record patch 1071 0004
11 0x409b810 record patch 956 0004
12 0x45a8eb0 record patch 956 0004
13 0x45bf510 record patch 841 0004
k 14 0x7ffe3465ffc0 paste assets 36905 0001
15 0x7fff6c00b5e0 paste assets 4 0001
k 16 0x68eea70 add track 36205 ffffffff
17 0x409a610 add track 4 ffffffff
k 18 0x40e7af0 add track 35677 ffffffff
k 19 0x4599000 add track 34373 ffffffff
k 20 0x7ffe4c4073a0 load 30481 ffffffff
21 0x7ffe4c01bc40 load 841 ffffffff
k 22 0x46cd670 move edit 26271 ffffffff
23 0x45a2360 move edit 2937 ffffffff
k 24 0x4698f10 load 26106 ffffffff
k 25 0x65161e0 load 3768 ffffffff
can you explain why both groups were separated after the second "undo"?
It is not really that both groups were separated after the second "undo" - what it really is, is just
like an Oracle database "commit". As the program runs, a commit writes down the current state
of the user's work; when you operate the first undo, you are just reverting to the state of the
space when it was written down. With the 2nd undo, you are now reverting to what was written
down before that. I have explained this as well as I could, but my field is computers and it is
so automatic to me, that I have no better way to convey this (so my apologies).
It's counter-intuitve that when trying to extend or shorten the length of
the video stream, the corresponding audio stream remains unaffected
(and vice-versa) despite being grouped.
I will look at Sam's qqjni.mp4 steamable to get a better understanding of exactly what you are referring
to so I can relay this to gg.
Actually, one more thing would be if "drag source only" would also apply to groups.
By this are you referring to section 41.3 Trim Feature with Drag Handle, Button 3? Either way, it is
now a little easier for me to keep track of what needs to be added to Cin, if it is added to MantisBT
bugtracker.