home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
Meeting Pearls 3
/
Meeting_Pearls_III.iso
/
Pearls
/
midi
/
Editors_and_Tools
/
OberSuite-1.03
/
TODO
< prev
next >
Wrap
Text File
|
1993-01-23
|
2KB
|
77 lines
General:
o Use string concatenation to cut down ErrorMsg().
o Make programs return Amiga-convention return codes, particularly
ObType -q.
o Allow files with arbitrary sets of Xp/M12 patches in them
(single mixed with multi, etc.)?????
This probably means a reworking of the PATCHINFO structure.
** Also, this will allow the user to create patch archives
containing patches with the SAME NUMBER! HOW TO HANDLE???
o Test programs under low-memory situations.
If not enough memory, go to alternate algorithm
(1 patch at a time, writing to file each patch.)
(Print a message: "Low on memory, using alternate algorithm")
(Or let the user specify it as a command-line option.)
o Instead of complaining when there isn't enough data, just keep
reading until there is. Something like:
amount = 0;
while ((amount < all the data) && (data buffer isn't empty))
{
specify we want (allData - amount) bytes;
amount += DoTheIO();
}
if (we didn't get it all)
NOW complain and retry.
To see if the data buffer is empty, do a CMD_QUERY.
Put this all inside GetFromMidi(), returning total bytes read.
o Get rid of dependency on "global" ->mode parameter everywhere
possible.
o Add 1 more file type: entire memory dump (single+multi).
ObGet:
o Get rid of the hack in TransmitPatchInfo.
o ObGet s2-5 m91-34 filename
(ObType will show singles/multis in different colors then.)
o Since no checksum, get the data twice and compare.
We don't have to store 2 copies of the data -- just generate
a checksum from the first transfer and use the 2nd.
Make this an option.
ObPut:
o Send a 1-second note-on/note-off after the transfer (user option).
o Send a patch-change command after transmission, so the current
patch is the new one. (user option).
(Problems with both -- which MIDI channel? Maybe require a
number between 1-16 after the option.)
ObType:
o Is pi->rightSize getting reset redundantly (or dangerously) in
FigureOutInstrument()?
ObExtract:
o Option -d to specify extraction directory.
patchinfo.c
o In VerifyPatch, check that the patch type (single/multi)
and the file size are consistent with one another.