Hi, Sreekant:
Sure. I'm also a subscriber of axis-dev-ml. Are you, too?
Before we move to axis-dev list ...
Please let me notify all of axis-users that jax-rpc handler
is an important arcitecture. And I want to suggest you pay
attention to the difference between 'Axis specific handlers'
and 'jax-rpc handlers' when you just say *'Handler'*.
The following section show us the starting point of this
activity, which is a Q&A with an axis committee.
.........*.........*.........*.........*.........*.........*.........*
| -----Original Message-----
| From: Tom Jordahl [mailto:[EMAIL PROTECTED]
| Sent: Wednesday, January 08, 2003 12:01 AM
| To: '[EMAIL PROTECTED]'
| Subject: RE: gzip support
|
| We will always support the JAX-RPC way of doing things,
| but that doesn't mean that Axis specific handlers aren't
| a good thing.
|
| Pick whichever one you think will get the job done.
| I would prefer JAX-RPC is you think you would ever want
| this handler to move to some other engine.
|
| Please consider contributing the handlers to the project
| if/when you complete them, I am sure many other may want
| to do a similar thing.
|
| --
| Tom Jordahl
| Macromedia Server Development
|
| >-----Original Message-----
| >From: Toshiyuki Kimura [mailto:[EMAIL PROTECTED]
| >Sent: Tuesday, January 07, 2003 3:53 AM
| >To: [EMAIL PROTECTED]
| >Cc: Tom Jordahl
| >Subject: RE: gzip support
| >
| >Hi Tom,
| >
| > I have a question for this matter.
| >
| > AXIS has two implementations of "Handler"; one is "AXIS Handler"
| >as the AXIS original, and the other is "JAX-RPC Handler" as the
| >standard implementation in accordance with JAX-RPC ver 1.0 spec.
| >
| > The AXIS original architecture has extensibility as you say.
| >However, it doesn't follow the JAX-RPC ver 1.0 spec. at the point
| >of "Handler implementation", currently.
| >
| > How do you think what AXIS should choice of the future style
| >for the AXIS handler implementations ? (i.e. Both AXIS original
| >and JAX-RPC , or JAX-RPC only, or ... ???)
| > The AXIS Project has already decided the direction ?
| >
| >Regards,
| >
| > Toshiyuki Kimura <[EMAIL PROTECTED]>
| > R&D Headquarters
| > NTT DATA Corp.
.........*.........*.........*.........*.........*.........*.........*
Best Regards,
Toshi (Toshiyuki Kimura) <[EMAIL PROTECTED]>
R&D Headquarters
NTT DATA Corporation
-----Original Message-----
From: Sreekant Thirunagari [mailto:[EMAIL PROTECTED]
Sent: Monday, March 03, 2003 11:51 PM
To: [EMAIL PROTECTED]
Subject: Re: Handler chain implementation
Hi Toshi,
This is exactly what i wanted to configure, index of the handler
user wants to start in case of 'return false'. You got me. Instead
of storing class name, index makes things lot easier. Ofcourse wsdd
gives more flexibility. I am not sure if we need to support '-1'
option you mentioned. I think we have to move this thread to
developer list so that we can take the opinion of other axis
handler developers. Ultimately we have to fill this gap of non
conference with spec completely (WSDD, handlerchainimpl ... ).
Regards,
Sreekant
