Hi everybody,
I already sent a mail a couple of weeks ago about a problem I had with
reacting to NewDataOnPorts events in my home-made reporting component.
I have reproduced the problem with RTT 1.10.3, and I join the test file.
The part of the test file reporting a "DataPort" works fine, while the
tests fails because of the "WriteDataPort" reporting!
Maybe I misuse the Handler (and it miraculous worked with the DataPort) or
there is still a bug with event ports...
I am looking forward to your solution!
Charles.
| Attachment | Size |
|---|---|
| test_reporter_event.cpp | 2.75 KB |
Test file for Write vs. RW Port event signals
On Monday 19 April 2010 18:11:28 Charles Lesire-Cabaniols wrote:
> Hi everybody,
>
> I already sent a mail a couple of weeks ago about a problem I had with
> reacting to NewDataOnPorts events in my home-made reporting component.
> I have reproduced the problem with RTT 1.10.3, and I join the test file.
>
> The part of the test file reporting a "DataPort" works fine, while the
> tests fails because of the "WriteDataPort" reporting!
>
> Maybe I misuse the Handler (and it miraculous worked with the DataPort) or
> there is still a bug with event ports...
>
> I am looking forward to your solution!
I misread your previous mail. This is actually desired behaviour. Only 'read-
capable' ports are notified that new data is available. What use does a write
port have to be informed of new data, if it can not read that data ?
Cheers,
Peter
Test file for Write vs. RW Port event signals
Peter Soetens wrote:
> On Monday 19 April 2010 18:11:28 Charles Lesire-Cabaniols wrote:
>
>> Hi everybody,
>>
>> I already sent a mail a couple of weeks ago about a problem I had with
>> reacting to NewDataOnPorts events in my home-made reporting component.
>> I have reproduced the problem with RTT 1.10.3, and I join the test file.
>>
>> The part of the test file reporting a "DataPort" works fine, while the
>> tests fails because of the "WriteDataPort" reporting!
>>
>> Maybe I misuse the Handler (and it miraculous worked with the DataPort) or
>> there is still a bug with event ports...
>>
>> I am looking forward to your solution!
>>
>
> I misread your previous mail. This is actually desired behaviour. Only 'read-
> capable' ports are notified that new data is available. What use does a write
> port have to be informed of new data, if it can not read that data ?
>
The component may want to be waken up in that case... anyway, I have
just found another way to write it to by-pass my problem
> Cheers,
> Peter
>