Re: RTT Flexiport branch

2012/12/14 Charles Lesire-Cabaniols <[email protected]>

>
>
>
> 2012/12/13 Peter Soetens <[email protected]>
>
>> Hi Charles,
>>
>> On Wed, Oct 24, 2012 at 10:04 AM, Charles Lesire-Cabaniols
>> <[email protected]> wrote:
>> > Hi,
>> >
>> > I have just pushed a new branch to RTT:
>> > https://gitorious.org/orocos-toolchain/rtt/commits/master-flexiport
>>
>> Thanks ! Sorry for not responding sooner, I really had noticed :-)
>>
>> >
>> > It contains some modifications to the Connection chain that allows to
>> > convert types for local connections, i.e., if you connect OutputPort<A>
>> to
>> > InputPort<B> in the same process, and provide a constructor B(A), then
>> the
>> > conversion will be done by the connection itself.
>> >
>> > When I run ctest, it results in:
>> > The following tests FAILED:
>> >       7 - corba-ipc-test (Failed)
>> >      35 - property_test (Failed)
>> >
>> > As these tests also fail for me in the master branch, I guess that it
>> is not
>> > caused by my modifications.
>>
>> They depend on some cmake setup (starting corba-ipc-server and copying
>> some xml/cpf files).
>>
>> >
>> > Peter/Sylvain, I would be reassured if you look at the code to check
>> that I
>> > correctly manipulated ChannelElement and DataSource objects, especially
>> with
>> > respect to allocation...
>>
>> We will
>>
>
> Good. Looking forward to reading from you
>
>
>>
>> >
>> > I join a little test I made that connect ports with a conversion of
>> simple
>> > structs.
>> >
>> > Seems to work ;)
>> >
>> > Next step: extending to out-of-band connections (i.e. managing
>> transports).
>>
>> Do you need any help here ?
>>
>
> I don't think so... but I just have no time for the moment. I will try to
> look again into it by mid-January.
>
>
Hi Peter (and others),

I just came back into the flexiport branch.

In ConnFactory::createConnection, you added the case where the input_port
is not local, but I can't see why and when we could actually lead to this
situation. My intuition is that when a port is not local, we should have
called createStream instead of createConnection.

Why did you consider this case? (the behavior in this case is not fully
implemented, so I guess it is a specific flexiport use case)
Is there a corner case I haven't seen?

A complementary question about connections vs. streams: I think it would
make sense to consider that creating an OutOfBand connection is just like
creating two streams, respectively for the input and output parts, wouldn't
it?

Charles.


>
>
>>
>> Peter
>>
>
>

-- 
Orocos-Dev mailing list
[email protected]
http://lists.mech.kuleuven.be/mailman/listinfo/orocos-dev