Skip to main content

[jsr356-users] Re: [jsr356-experts] Re: Package naming and arrangement

  • From: Danny Coward < >
  • To:
  • Cc: Arun Gupta < >
  • Subject: [jsr356-users] Re: [jsr356-experts] Re: Package naming and arrangement
  • Date: Wed, 12 Dec 2012 13:31:04 -0800

Hi Arun,

On 12/12/12 11:59 AM, Arun Gupta wrote:
" type="cite">Danny,

In Java EE 6, javax.servlet.annotation and javax.xml.bind.annotation are the only two packaged that have annotations separated out. All other technologies (JAX-RS, CDI, EJB, JAX-WS, etc) have them in the same packages. It makes sense to flatten out javax.websocket.annotations to javax.websocket package.
Yes we put took out the annotation subpackage in did that in an earlier draft for that reason.

" type="cite">
As Mark mentioned, splitting the APIs in two packages does create cyclic dependency. OTOH, do you expect only APIs in the client package to be bundled for rich client applications ?
Rich client applications will need the javax.websocket.* package, server applications will need javax.websocket.* and javax.websocket.server.*.

" type="cite"> IMHO the names are intuitive and its only 5 classes. I'd rather keep a flat structure with everything in javax.websocket.* package.
OK. Well we are balancing the need not to require server API classes for rich clients, which really pulls us in the direction of needing a separate package to make the separation between the models clean.

We may well have more server specific classes in future releases too !

- Danny
" type="cite">
Thanks,
Arun

On 12/12/12 3:49 AM, Mark Thomas wrote:
On 12/12/2012 00:33, Danny Coward wrote:

<snip/>

This is what the API would look like in this (final?) package structure:-

javax.websocket.server.*
- DefaultServerEndpointConfiguration
- ServerEndpointConfiguration
- ServerContainer
- WebSocketEndpoint
- WebSocketPathParam

javax.websocket.*
-<everything else>

Let me know if anyone sees any issues with this arrangement.
One very minor nit is that it creates a circular dependency between the
two packages. It could be broken by having separate ContainerProviders
for client and server. I'm on the fence as to whether this is worth doing.

Mark




--
Danny Coward
Java EE
Oracle Corporation


[jsr356-users] [jsr356-experts] Re: Package naming and arrangement

Danny Coward 12/12/2012

[jsr356-users] [jsr356-experts] Re: Package naming and arrangement

Mark Thomas 12/12/2012

[jsr356-users] Re: [jsr356-experts] Re: Package naming and arrangement

Arun Gupta 12/12/2012

[jsr356-users] Re: [jsr356-experts] Re: Package naming and arrangement

Danny Coward 12/12/2012

[jsr356-users] Re: [jsr356-experts] Re: Package naming and arrangement

Arun Gupta 12/12/2012

[jsr356-users] Re: [jsr356-experts] Re: Package naming and arrangement

Danny Coward 12/12/2012

[jsr356-users] Re: [jsr356-experts] Re: Package naming and arrangement

Arun Gupta 12/12/2012

[jsr356-users] [jsr356-experts] Re: Package naming and arrangement

Danny Coward 12/12/2012

[jsr356-users] [jsr356-experts] Re: Package naming and arrangement

Mark Thomas 12/12/2012
 
 
Close
loading
Please Confirm
Close