geronimo-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "kamesh kompella" <>
Subject Re: Axioms for JCA implementation
Date Mon, 25 Aug 2003 22:51:51 GMT
Re: Axioms for JCA implementationHi Richard,
                  Thanks for the info. In that case, I will look into getting RA going for
openjms. This will be the "dummy" RA that can be used for testing the app server's contract's
implementation once the infrastructure is in place.  

I was intrigued by one of your statements:
"Personally, I don't think the CCI offers much value. Most Ras will have very specific APIs.
I think we should avoid supporting it."
Is the variability in API ok as the code will be derived at deployment time with the help
of some tools which will hide the custom API from source code? Also, what happens to two tier
clients ( Should we leave them out altogether since this is as bad as mixing DB calls in a
JSP) ? Do we assume that these clients will talk to the App server? 

  ----- Original Message ----- 
  From: Richard Monson-Haefel 
  Sent: Monday, August 25, 2003 2:23 AM
  Subject: Re: Axioms for JCA implementation

  My 2 cents ...

  On 8/24/03 11:46 PM, in article, "kamesh kompella"
<> wrote:

               There are some tests in place already under the incubator-geronimo\specs\j2ee-connector\src\test
directory. However,  these are against the classes that are specified in the spec. Beyond
these, as I see it, one can write all the tests one wants, (as they will be testing the interfaces)
 but the more important path of figuring out what's to be done to make the implementations
concrete has to be tread. There are several gaps in the connector specification that will
need to be filled in. These are left out for the implementation. Further, we need to figure
out the axioms for the system.

      1.. What can we assume and program against? In other words, what are the givens and
what is it we are trying to accomplish? 

  Support for arbitrary JCA 1.5 connectors (resources).

    2. Are we going to provide hooks to let individual RAs or are we not going to do any implementation
there at all? In fact, what is the purview of this project vis-a-vis the implementation? Are
we going to limit ourselves to simply implementing the contract from the app server side?
 I do realize that providing an implementation for each EIS there is, is hard.

  At a minimum we should provide a JCA RA for OpenJMS. This will allow us to use that impl
for MDBs and as the default JMS provider. Last time I checked, the JMS API wasn't explicitly
aligned with JCA 1.5, but I don't think that will be a problem to work around.

    3. The spec does not require that the CCI need be supported. In the light of this information,
are we planning on evolving an API? 

  Personally, I don't think the CCI offers much value. Most Ras will have very specific APIs.
I think we should avoid supporting it.

    4. In any case, we will need to write dummy RAs that will ensure that our tests run. We
will also need to supply other components that will become a reality eventually when other
groups fill in.

    I recall having have heard certain other ideas that are outside the scope of the spec
per se but seem to be the responsibility of the implementation group. I was wondering if anybody
can fill me in on these.

    In short, since this is a spec that is pretty much laid out, with most of the work being
done by the third parties or first parties which ever you feel like calling it, I see little
reason for not getting the ball rolling. This is, of course, my 2 cents, and I could be splashing
my ignorance for all and sundry to wallow in. If that is the case, please enjoy the humor
I provided and don't forget to mail me the check. :)

  I would advise that we wait until the core architecture of Geronoimo is established. There
is still debate about how the kernel should be implemented, and this will have a direct impact
on the JCA support in Geronimo.  That said, I don't think there is any harm in getting started
on the dummy RAs (one sync and the other async) since this will take some time and will be
extremely useful for testing once development does begin.


      ----- Original Message ----- 
      From: Nick Faiz <>  
      To: '' <mailto:'geronimo-devÑ'>
      Sent: Sunday, August 24, 2003 3:45 PM
      Subject: Newcomers being useful


                 I'm another newcomer. When time permits, I think it would be helpful to give
newcomers a clearer insight into what is needed for the project. For example, I'd be happy
to write unit tests but how do I know that another developer is not doing the same thing at
the same time somewhere else? On the other hand, if someone needs a hand with something like
that, please let me know.


                 I'm looking forward to Alex Rupp's design documents ...


                 My current goal is to keep an up to date build, learn the design, etc.. Hopefully,
when things settle down, it will become clearer for the uninitiated how to participate.



      Nick Faiz.         



      -----Original Message-----
      From: Siva [mailto:siva@^'z!³öÜT(kl^ÓN] 
      Sent: Monday, 25 August 2003 4:40 AM
      To: geronimo-devÑ
      Subject: Contribute [Was Re: Who are working on the JCA integration part]


      IMO,the best way to start contributing is to read the codebase and write unit tests
for it.This way you can start understanding the 

      system and help the system evolve towards a TDD.



        ----- Original Message ----- 

        From: kamesh kompella <>

        To: <mailto:geronimo-devÑ>

        Sent: Monday, August 25, 2003 12:05 AM

        Subject: Re: Who are working on the JCA integration part



          I am interested in contributing. I have added my name and I was perusing through
the cvs. Can somebody let me know the current status and where I can jump in?




View raw message