beware of dumpOSC (was Re: [OSC_dev] No dumpOSC output in pipe
in non-interactive terminal)
Martin Peach
martin.peach at sympatico.ca
Sun Sep 9 01:32:50 PDT 2007
Pd's [dumpOSC] uses code from dumpOSC and [unpackOSC] is built on
[dumpOSC] with additional support for timetags. [unpackOSC] takes lists
of bytes (as floats) as input. The input bytes are supposed to come from
[udpreceive] or something like that, so the transport is separated from
the OSC protocol. The timetag data is output as a delay in milliseconds
so that pd's delay objects can use it to schedule events.
Martin
Torsten Anders wrote:
> Dear IOhannes,
>
> Thanks for sharing this issue. Just for my information: Pd's
> [unpackOSC] uses code from dumpOSC?
>
> Best
> Torsten
>
> On Sep 8, 2007, at 8:32 PM, IOhannes m zmoelnig wrote:
>> hi
>>
>> Stephen Sinclair wrote:
>>> Looking at the dumpOSC.c source file, there is nothing special
>>
>> apart from the fact that dumpOSC uses some pointer casting magic.
>> this seriously screws up, when you just use the code in a naive way and
>> turn on compiler optimization.
>> (gcc>=3.0 and "-fstrict-alignment" cannot be used in conjunction with
>> pointer reinterpret casts)
>> i just recently discovered this problem in one of Pd's
>> OSC-implementations [1], which is basically a copy'n'paste of dumpOSC.
>>
>> if you want to use dumpOSC code and use compiler optimization at the
>> same time, you should fix this e.g. by explicitely using unions.
>>
>>
>> mmfga.sdr
>> IOhannes
>>
>>
>> [1] see
>> http://sourceforge.net/tracker/index.php?func=detail&aid=1788587&group_id=55736&atid=478070
>>
>> for my troubles...
>> _______________________________________________
>> OSC_dev mailing list
>> OSC_dev at create.ucsb.edu
>> http://www.create.ucsb.edu/mailman/listinfo/osc_dev
>
> --
> Torsten Anders
> Interdisciplinary Centre for Computer Music Research
> University of Plymouth
> http://strasheela.sourceforge.net
> http://www.torsten-anders.de
>
>
> _______________________________________________
> OSC_dev mailing list
> OSC_dev at create.ucsb.edu
> http://www.create.ucsb.edu/mailman/listinfo/osc_dev
>
More information about the OSC_dev
mailing list