one thing i've noticed for a while is that trigger objects defined with 'int' or 'i' cause an error:
error: trigger: i: bad type
pd allows the object to be created but the int outlet behaves the same as a float outlet. is this a feature? or ... is it ... ... a bug?
also, if i have a notein object in a patch, and also have a midi controller plugged in, pd will hang and have to be killed with a sig 9. the dying message reads:
socket receive error: Connection reset by peer (104)
this is using oss and a venerable musicquest midi interface. this setup works fine with other linux midi apps (jazz, etc.). i've tried many combinations of other factors, and if i plug the controller in AFTER the patch is loaded sometimes i can get a couple of notes in before it hangs. has anyone experienced anything like this, and was it solveable?
thanks,
B
On Tue, Dec 19, 2000 at 09:57:43PM -0500, Bill Sack wrote:
one thing i've noticed for a while is that trigger objects defined with 'int' or 'i' cause an error:
error: trigger: i: bad type
pd allows the object to be created but the int outlet behaves the same as a float outlet. is this a feature? or ... is it ... ... a bug?
As I understand it, PD has no integer type. The "int" box might be better called "floor" except that it rounds toward zero (ugh!) and behaves like "float" in terms of eagerness. So the various objects that want to know type don't recognize "int", "integer", or "i" as meaning "round this one off". However, they do seem to recognize "int" as special in some way; "trigger flobnurgl" works, but "trigger int" gives an error.
trigger decides what type to give based on the first character of its argument (don't believe me? look in x_connective.c), and the types allowed are: 'p' pointer. What these are and how you manipulate them I have no idea. 'f' float 'b' bang 'l' list 's' symbol
Hope this helps (sorry, I don't know anything about MIDI for your other question; no money for hardware, not enough energy to try to get ALSA virtual MIDI drviers going)
Andrew Archibald