incubator-etch-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "scott comer (sccomer)" <scco...@cisco.com>
Subject Re: [jira] Created: (ETCH-54) TransportFactory.newTransport returns stack topped by DeliveryService which makes it hard to implement meta transports
Date Tue, 03 Feb 2009 20:49:53 GMT
i would like to propose fixing this issue in etch release 1.1. it will 
be blocking for the name service implementation.

any objections, questions?

scott out

scott comer (JIRA) wrote:
> TransportFactory.newTransport returns stack topped by DeliveryService which makes it
hard to implement meta transports
> ----------------------------------------------------------------------------------------------------------------------
>
>                  Key: ETCH-54
>                  URL: https://issues.apache.org/jira/browse/ETCH-54
>              Project: Etch
>           Issue Type: Improvement
>           Components: csharp-binding, java-binding
>     Affects Versions: 1.0.1, 1.0.0, 1.0.2
>             Reporter: scott comer
>
>
> To implement a meta transport, such as a transport supporting the name service, it is
necessary to be able to internally construct a transport without having a MailboxManager and
a DeliveryService stacked on top. For example PerfHelper.newServer calls TransportFactory.newTransport
and gets back a DeliveryService which is then used to construct the remote and stub directly.
The change would be for TransportFactory.newTransport to return the transport stack topped
by the filter chain elements and then have PerfHelper add the MailboxManager and the DeliveryService
before constructing the remote and the stub. So, the change is to move the construction of
MailboxManager and DeliveryService to the generated helper code from the likes of TcpTransportFactory.
>
>   

Mime
  • Unnamed multipart/mixed (inline, None, 0 bytes)
View raw message