[PD-dev] multicast patch to OSCx

[Hans-Christoph Steiner]

It builds on Mac OS X and GNU/Linux (i.e. works on UNIX), but dies on  
Windows/MinGW:

cd src && make
make[4]: Entering directory `/home/pd/auto-build/pd-extended/ 
externals/OSCx/src'
gcc -g -O2 -mms-bitfields -Wall -Wimplicit -Wunused -Wmissing- 
prototypes -O3  -I../libOSC -I../../pd/src -I../../../pd/src -I../src  
-I../../../pd/src -I../libOSC -I../../pd/src -I../../../pd/src -I../ 
src -I../../../pd/src -c -o sendOSC.o sendOSC.c
gcc -g -O2 -mms-bitfields -Wall -Wimplicit -Wunused -Wmissing- 
prototypes -O3  -I../libOSC -I../../pd/src -I../../../pd/src -I../src  
-I../../../pd/src -I../libOSC -I../../pd/src -I../../../pd/src -I../ 
src -I../../../pd/src -c -o htmsocket.o htmsocket.c
htmsocket.c: In function `OpenHTMSocket':
htmsocket.c:236: error: `IP_MULTICAST_TTL' undeclared (first use in  
this function)
htmsocket.c:236: error: (Each undeclared identifier is reported only  
once
htmsocket.c:236: error: for each function it appears in.)



On Jan 2, 2007, at 10:59 AM, Stephen Sinclair wrote:

> That's great!
>
> About the configure.ac stuff, I probably shouldn't have included those
> changes in the submitted patch.
> However, as a separate issue, those changes were necessary for me get
> OSCx to compile against the Pd.app bundle.  Any chance on making
> similar configure.ac changes that won't break Pd-extended?
>
> Steve
>
>
> On 1/1/07, Hans-Christoph Steiner <hans at eds.org> wrote:
>>
>>
>> I committed it without the changes to configure.ac since those  
>> broke the
>> Pd-extended build stuff.  We'll see tomorrow whether it builds on  
>> Windows...
>>
>> .hc
>>
>>
>> On Dec 28, 2006, at 3:11 PM, Stephen Sinclair wrote:
>> Hi,
>>
>> I did this patch quickly because I needed it.
>>
>> The changes are _very_ simple, so I have some confidence that it  
>> should
>> work.
>>
>> Some quick tests under Linux and OS X showed that it was working.   
>> I haven't
>> tried the patch under Windows. (Never compiled Pd under Windows  
>> actually.)
>>
>> Testing the TTL parameter is not so easy... you need to test  
>> several ranges
>> of subnets.  Setting it to 0 showed that a message could be sent to
>> localhost but not to another computer on the subnet, so that at least
>> supports the theory that it works.  But I haven't tried it over the
>> internet.
>>
>> By the way, I made the changes based on this tutorial:
>> http://ntrg.cs.tcd.ie/undergrad/4ba2/multicast/antony/index.html
>>
>> Testing it with Max/MSP's " net.multi.send/receive" objects  
>> produced some
>> garbage... I'm not sure what that was all about. But testing it  
>> with Max
>> 4.6's udpsend/receive objects worked fine.
>>
>> Another change I'd like to make is to have sendOSC support some  
>> arugments
>> for port/host instead of having to send it the "connect" message,  
>> but I
>> thought that kind of change would be out of this patch's scope.
>>
>> Let me know if anyone has comments or criticisms, I'd be happy to  
>> make
>> changes and re-submit the patch.
>>
>> Steve
>>
>>
>>
>> On 12/28/06, Hans-Christoph Steiner <hans at eds.org> wrote:
>> >
>> > This patch adds multicast support to the OSCx libs, it looks  
>> useful,
>> > but I don't really use OSC, so I couldn't say if its worth  
>> taking or
>> > not:
>> >
>> > http://sourceforge.net/tracker/index.php?
>> > func=detail&aid=1612910&group_id=55736&atid=478072
>> >
>> > Shall we trying it in the nightly test builds?  Anyone willing to
>> > vouch for this patch so that it can be included in the next  
>> release?
>> >
>> > .hc
>> >
>> >
>>
>>
>>
>>
>> --------------------------------------------------------------------- 
>> ---
>>
>>                                               http://at.or.at/hans/
>>
>>


------------------------------------------------------------------------

"[W]e have invented the technology to eliminate scarcity, but we are  
deliberately throwing it away to benefit those who profit from  
scarcity."        -John Gilmore