axis-c-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Susantha Kumara <susan...@opensource.lk>
Subject Re: Priorities for Axis C++ 1.3 Release
Date Tue, 29 Jun 2004 13:11:13 GMT
On Tue, 2004-06-29 at 16:23, damitha kumarage wrote:
> On Tue, 2004-06-29 at 15:42, Samisa Abeysinghe wrote:
> > As more and more people are interested in interop it is important to focus on aspects
related to
> > interop as well.

Yes. This is a long term process that we have to carry out through out
next releases.

> > 
> > Some of the areas of improvement would be:
> > 1. xsd:any support - Elements must have inner complex types, avoid unbounded types,
no schema
> > references etc.

I think deciding the DOM model for this is more important.

> > 2. Accepting SOAP messages containing href
> > 3. Encoding used for char* (UTF8, UTF16)

Thanks Samisa I will add these too to the list of tasks.

> > 4. Array declaration should interop with .NET and Axis Java

Please explain this item. What is your alternative suggestion to replace
the existing axis defined arrays ?.

> > 
> > BTW: I think before integrating txpp to Axis C++, txpp itself has to be developed,
and verified to
> > work. I think it will be a good idea to make txpp a seperate project, rathe than
making it part of
> > Axis C++

After the parser abstraction its easy to develop and plug in the parser
to Axis without affecting the Axis code. The txpp developer can work on
it even without any knowladge on Axis code.

> What do you mean by making txpp a seperate project?
> Clearly the development thread can be seperated from the Axis C++
> development process. Yet I think txpp need not be a general pull parser.
> It can be optimized as a soap parser(not really a general xml pull
> parser). Initially it could be bound to Axis C++ requirements.

I totally agree.

>  
> We could avoid DTD support?.

Yes by droping DTD/Schema support we reduce lots of work for the moment.

> We can rename it from txpp(types xml pull parser) to 
> tspp(typed soap pull parser) or tpps(typed pull paresr for soap)
> 

Yes this is a good idea. Some of the SOAP engines use SOAP optimized XML
parsers to acheive performance. So we can use txpp parser for Axis first
and then make it generic later.

> 
> damitha
> > .
> > 
> > Thanks,
> > Samisa...
> > 
> > --- Susantha Kumara <susantha@opensource.lk> wrote:
> > > Hi All,
> > >  
> > > Lets discuss the priorities of tasks for Axis C++ 1.3. As a start I
> > > would like to point out following tasks in order. High priority tasks
> > > first,
> > >  
> > > 1.	The real XML Pull parser integration (txpp)
> > > *         When the existing txpp parser is improved to handle namespaces
> > > we can use it in Axis. No need to implement the type conversion layer
> > > initially because Axis already has the type conversion routines
> > > *         There are some other volunteers to take up some of these work
> > > 2.	Libwww transport layer and chunking support in the existing
> > > transport layer
> > > *         This work will also improve/write transport layers to support
> > > SWA at the same phase
> > > 3.	Improve 2 apache modules to support http chunking (I think
> > > receiving is already done by apache. We need to improve it for sending)
> > > 4.	STL elimination
> > > 5.	RPC/Literal support
> > > 6.	Soap with attachments.
> > >  
> > > IMO tasks 1-4 are set high priority because they will bring Axis C++ to
> > > be the fastest open source SOAP engine (the fact that increases the
> > > customer base).
> > >  
> > > I will add the other tasks when the discussion thread goes on. 
> > >  
> > > What do you think ? Lets discuss and set goals.
> > >  
> > > Thanks,
> > >  
> > > Susantha
> > >  
> > > Lanka Software Foundation
> > >  
> > > 
> > 
> > 
> > 
> > 	
> > 		
> > __________________________________
> > Do you Yahoo!?
> > New and Improved Yahoo! Mail - 100MB free storage!
> > http://promotions.yahoo.com/new_mail
> > 
> 
> 


Mime
View raw message