thx Iohannes for looking into it.
what i do to test it:
in PD: [comport 3 9600] - [print]
in Arduino: in setup:Serial.begin(9600) in loop: Serial.println("z"); delay(3000);
more simple it cannot be, i thought.
using the serial monitor of the Arduino IDE the z's are printed.
Pd gives the RXerrors (10 of them).
when i unplug the cable there's no reaction from [comport].
when i do [close( the following 3 lines are printed:
[comport] ** ERROR ** couldn't reset params to DCB of device COM3
[comport] Couldn't reset old_timeouts for serial device [comport] closed COM3
IOhannes wrote:
note, that virtual serial-ports on the client are forwarded to network on the host, so no actual serial hardware is involved. it shouldn't make a difference for the client though.
i'm not sure what this should tell me. my reasoning is that when my virtual XP doesn't have this problem, it's the combination of [comport] & W10 at fault.
(the Arduino on it's own IDE works fine: driver and hardware are apparantly ok).
IOhannes wrote:
i just tested [comport] on a virtual W10 (running on a Debian host),
so your host is Debian, mine is W10. which tells what?
rolf
PS i installed a fresh copy of Pd-ext 43.4
On 2015-10-07 07:47, rolfm@dds.nl wrote:
IOhannes wrote:
i just tested [comport] on a virtual W10 (running on a Debian host),
so your host is Debian, mine is W10. which tells what?
nothing. i was merely mentioning the fact that my host is Debian for completeness sake.
the point was, that i have a [comport] running fine on a W10 machine. so i doubt your assumption "[comport] buggy on W10".
i'm aware that this doesn't help you much (since it doesn't magically make [comport] work on your W10 installation), but it adds a data point to triaging the actual problem.
PS i installed a fresh copy of Pd-ext 43.4
i used Pd-vanilla 0.46-7 with [comport] being installed either via deken or from pd-iem (probably the latter).
fgmad IOhannes
On Wed, Oct 7, 2015 at 3:54 AM, IOhannes m zmoelnig zmoelnig@iem.at wrote:
On 2015-10-07 07:47, rolfm@dds.nl wrote:
IOhannes wrote:
i just tested [comport] on a virtual W10 (running on a Debian host),
so your host is Debian, mine is W10. which tells what?
nothing. i was merely mentioning the fact that my host is Debian for completeness sake.
the point was, that i have a [comport] running fine on a W10 machine. so i doubt your assumption "[comport] buggy on W10".
i'm aware that this doesn't help you much (since it doesn't magically make [comport] work on your W10 installation), but it adds a data point to triaging the actual problem.
PS i installed a fresh copy of Pd-ext 43.4
i used Pd-vanilla 0.46-7 with [comport] being installed either via deken or from pd-iem (probably the latter).
I get errors on Win10 as well. It seems that the read function always returns immediately and getOverlappedResult needs to be called. I'm trying to fix it but at the moment Pd on Win10 won't load my externals when I build them with Visual Studio 2015 on a Win7 machine, so I'm stuck there until I figure out why.
Martin