• Anywhere that text entry is permitted on the sample screen (and as far as I can tell, ONLY the sample screen) numpad * gets interpreted as "8". I seem to recall this being temporarily fixed and then promptly unfixed ages ago, for whatever reason I never brought it up again.

• Current builds have some strange things going on when you try to enter a note while the song is playing. For one, the cursor is fixed on the current row. But other weird stuff starts to happen when you play samples on the sample screen, like whatever is in the audio buffer gets repeated or something...

(this is what you get, when you get me interested in tracking again :P)

Samples are getting interpreted with the wrong signed/unsigned value. Not a big problem since an Alt-A fixes things, but it's weird that it's only come up recently.

>>4515
er, some samples. Probably less than 1/4 of mine.

Also I tried attaching a plain .wav file and got brought http://sovietrussia.org/test/da.py which seems to be a blank page.

Hitting ctrl-N and making a new song doesn't trigger the "module not saved" confirmation dialog. However making changes, hitting ctrl-N, and then immediately trying to quit or load a new song does trigger the dialog.

>>4516
That's because the board sucks. :/

>>4517
Er? Are you saying Ctrl-N should prompt you to save a blank song? That doesn't seem right...

(also the wav code was rewritten recently, which would explain >>4515.)

>>4519

>Er? Are you saying Ctrl-N should prompt you to save a blank song? That doesn't seem right...

No, I'm saying Ctrl-N (or Escape, File, New) should prompt you to save a non-blank changed-since-saved song, the same way loading a module does. Right now it's like Ctrl-N is ignored entirely regarding saving/loading:

• Load any song.
• Press F2 and change something.
• Press Ctrl-N and make a new song. You get no prompt for save.
• Immediately (without making any changes to the blank new song) try to load another module. You do get a prompt for save. Schism hasn't realized the changes from the first track you loaded have been obliterated with Ctrl-N.

Ohhh.

Old one I forgot about until now: trying loading the left part of a 16-bit stereo sample loads half the left part followed by half the right part.

A new thing seems to be that the left and right parts are reversed on the waveform display. (i.e. left is on the bottom)

(I think that's all the long-standing problems that haven't been fixed already, that I know of.)

The wav thing should be fixed now. It'll be in the repo later (when Mercurial stops being an ass).

>>4522 is fixed – quick test indicates IT clears the state when you erase the patterns. Makes perfect sense, as it also removes the filename then.

The cursor position on the message screen does weird things. It's possible to crash schism.

• load up any module that has a message screen at least 3 lines long
• go to the start of the bottom-most non-blank line
• start hitting backspace

If you don't follow: shift-f9 enter pagedown pagedown backspace*100. If you need an example file, try player abuse test 12.it

>>4551
Not surprising, as the message editor code really sucks. It should probably be rewritten entirely.

With certain songs, trying to reorder pattern numbers on the Order List (F11, Alt-R) causes a segfault every time.

With my tracks it seems to be entirely older ones, I don't know if that has anything to do with it, or if I just used to do something weird that I stopped doing, or what.

>>4577
Aha! You had patterns listed in the orderlist that didn't physically exist. Strange.

Think I've fixed it in current hg.

>>4591

Ah, blank patterns used for padding so that the last notes can fade out, or whatever. Good catch & thanks.