py_interface and extended pids and ports

[Clint Miller]

Extended PIDs and PORTs are simply a change from 18 to 28 bits?  Is that it?

On 7/22/06, Clint Miller <tosvar@REDACTED> wrote:
> Hi there erlang gurus -
>
> I'm trying to reuse a python codebase that *I believe* is a good
> candidate to live it's next life as a set of erlang nodes.  So, I was
> naturally drawn to py_interface.  On running the tests I get
>
> =ERROR REPORT==== 22-Jul-2006::11:02:54 ===
> ** enode1@REDACTED: Connection attempt to node
> py_interface_test@REDACTED aborted since it cannot handle extended
> pids and ports. **
>
> The occurs with R10B-10 and R11B-0 on both linux and OSX.  I've not
> tried any earlier versions.
> Googling erlang.org for this I find this conversation and accompanying
> totally untested patch.
>
> http://www.erlang.org/ml-archive/erlang-questions/200504/msg00012.html
>
> It says that the ports and pids extension occured in R10.  The patch
> doesn't work (being totally untested and all...no surprise there) but
> the tests then do merely time out.   So, I tried the +R flag to erl
> and viola!  Works consistently with +R 9.
>
> Now, my questions are these:
>
> *) what do I lose or risk by using +R 9 throughout my erlang
> development?  Do I give up feeatures/bugfixes/etc introduced in R10
> and beyond?
> *) can someone point me to where I would look in the erlang codebase /
> docs to start implementing extended ports and pids in py_interface so
> that I don't have to live in a R9 world?  (I bet I'm not the only
> person interested in this...)
> *) should I just give up and use the R9 distribution?
> *) is this a fool's errand?
>
> Thanks for your guidance
> --
> Clint
>


-- 
Clint Miller
--
"Never have so many
known so little
about so much"
-- James Burke