axis-java-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Demetris <demet...@ece.neu.edu>
Subject Re: SOAP styles
Date Tue, 01 Sep 2009 17:13:06 GMT

He he - I hear ya - indirections indirections ..

Not a bad idea and that's something we looked into it Wolfgang as a last 
resort. Thanks for the pointers and I will keep the
list informed on the outcome.

WJ Krpelan wrote:
> Hi,
> small wonder.
> One possibility is of course given by the saying "every IT-Problem can be solved by another
indirection"
> create new webservices out of the description of the old ones and let the new ones call
the old ones - performance-issues to the side
> kind of usual IT-madness, dont cite me
> Cheers, Wolfgang
>
> --- On Mon, 8/31/09, Demetris <demetris@ece.neu.edu> wrote:
>
>   
>> From: Demetris <demetris@ece.neu.edu>
>> Subject: Re: SOAP styles
>> To: axis-dev@ws.apache.org
>> Date: Monday, August 31, 2009, 6:48 PM
>>
>> Hi Wolfgang,
>>
>>    thanks for the good info - I did see some
>> of these links and I will go over them.
>>
>>    We have an underlying p2p infrastructure
>> that intercepts the SOAP messages from
>> clients to servers without distrupting the operation of the
>> upper layers ( so the soap
>> enginers and the WS implementations remain as is) and
>> routes them across peers.
>> Now that we are adding mobile peers in the mix, including
>> CXF-based ones, Axis2
>> ones etc. we are seeing some incompatability issues - for
>> ex. CXF does not handle
>> rcp/encoded styles. So how do you suggest we avoid the
>> migration and manage to
>> handle these issues?
>>
>> Thanks again
>>
>> WJ Krpelan wrote:
>>     
>>> Hi
>>> http://ws.apache.org/axis/java/reference.html
>>> hopefully answers your last question, dont expect
>>>       
>> bugfixes any more ;-)
>>     
>>> there is a migration guide for axis1 to axis2
>>>       
>> webservices,
>>     
>>> http://ws.apache.org/axis2/1_5/migration.html
>>> dont think there is any special attention to
>>>       
>> preserving soap-styles however. its really not in the spirit
>> nowadays, see
>>     
>>> http://ws-i.org/
>>> you didnt mention why you would want to migrate
>>> good luck,
>>> Wolfgang
>>>
>>> --- On Fri, 8/28/09, Demetris <demetris@ece.neu.edu>
>>>       
>> wrote:
>>     
>>>    
>>>       
>>>> From: Demetris <demetris@ece.neu.edu>
>>>> Subject: Re: SOAP styles
>>>> To: axis-dev@ws.apache.org
>>>> Date: Friday, August 28, 2009, 7:41 PM
>>>>
>>>> Well I will certainly push the notion of upgrading
>>>>         
>> the
>>     
>>>> target servers but there are cases where the
>>>>         
>> customer does
>>     
>>>> not
>>>> want to do that. So we NEED to deal with
>>>>         
>> deprecated styles
>>     
>>>> - so the question will remain if Axis 1.4 can
>>>>         
>> generate
>>     
>>>> one and only or multiple (even if deprecated)
>>>>         
>> styles
>>     
>>>> programmatically?
>>>>
>>>> Cheers
>>>>
>>>> WJ Krpelan wrote:
>>>>      
>>>>         
>>>>> Hi,
>>>>> all SOAP styles except doc/lit are kind of
>>>>>           
>> deprecated
>>     
>>>>>        
>>>>>           
>>>> by now and are no longer fully supported by most
>>>>         
>> frameworks,
>>     
>>>> if at all.
>>>>      
>>>>         
>>>>> You better migrate everything to doc/lit,
>>>>>           
>> resp.
>>     
>>>>>        
>>>>>           
>>>> doc/lit "wrapped" I suppose
>>>>      
>>>>         
>>>>> Cheers, Wolfgang
>>>>>
>>>>>
>>>>> --- On Thu, 8/27/09, Demetris <demetris@ece.neu.edu>
>>>>>        
>>>>>           
>>>> wrote:
>>>>      
>>>>         
>>>>>           
>>>>>           
>>>>>> From: Demetris <demetris@ece.neu.edu>
>>>>>> Subject: SOAP styles
>>>>>> To: axis-dev@ws.apache.org
>>>>>> Date: Thursday, August 27, 2009, 10:10 PM
>>>>>> Hi all,
>>>>>>
>>>>>> we have some legacy systems still using
>>>>>>             
>> Axis 1.4
>>     
>>>>>>          
>>>>>>             
>>>> and we
>>>>      
>>>>         
>>>>>> need clients from them to generate SOAP
>>>>>> rpc/lit or doc/lit instead of rpc/enc -
>>>>>>             
>> does
>>     
>>>>>>          
>>>>>>             
>>>> anyone know if
>>>>      
>>>>         
>>>>>> the latter is the default for Axis 1.4
>>>>>> and how it can be manipulated
>>>>>>             
>> programmatically?
>>     
>>>>>> Thanks
>>>>>>
>>>>>> Ruwan Linton wrote:
>>>>>>            
>>>>>>             
>>   
>>     
>>>>>>> On Thu, Aug 27, 2009 at 11:21 PM,
>>>>>>>               
>> Deepal
>>     
>>>>>>>        
>>>>>>>               
>>    
>>     
>>>> jayasinghe
>>>>      
>>>>         
>>>>>>>          
>>>>>>>               
>>         
>>     
>>>>>> <deepalk@gmail.com
>>>>>> <mailto:deepalk@gmail.com>>
>>>>>> wrote:
>>>>>>            
>>>>>>             
>>   
>>     
>>>>>>>        >
>>>>>>>        > No
>>>>>>>               
>> I can't, I guess
>>     
>>>>>>>        
>>>>>>>               
>>    
>>     
>>>> I
>>>>      
>>>>         
>>>>>>>          
>>>>>>>               
>>         
>>     
>>>>>> have explained why I can't use it as
>>>>>>             
>> well,
>>     
>>>>>>            
>>>>>>             
>>   
>>     
>>>>>>>        >
>>>>>>>               
>> because I cannot
>>     
>>>>>>>          
>>>>>>>               
>>         
>>     
>>>>>> differentiate the undeployment call for
>>>>>>             
>> the hot
>>     
>>>>>>            
>>>>>>             
>>   
>>     
>>>>>>>        >
>>>>>>>               
>> update and real
>>     
>>>>>>>          
>>>>>>>               
>>         
>>     
>>>>>> undeployment. Well, what Amila suggested
>>>>>>             
>> will
>>     
>>>>>>          
>>>>>>             
>>>> work
>>>>      
>>>>         
>>>>>>            
>>>>>>             
>>   
>>     
>>>>>>>        >
>>>>>>>               
>> though :-)
>>     
>>>>>>>        Of
>>>>>>>               
>> course you can if the
>>     
>>>>>>>        
>>>>>>>               
>>    
>>     
>>>> file
>>>>      
>>>>         
>>>>>>>          
>>>>>>>               
>>         
>>     
>>>>>> is there then that is hot-update else it
>>>>>>            
>>>>>>             
>>   
>>     
>>>>>>>        is un
>>>>>>>               
>> deployment.
>>     
>>>>>>>        >
>>>>>>>        >
>>>>>>>        >
>>>>>>>    
>>>>>>>               
>>    >         
>>    
>>     
>>>>         
>>>>         
>>>>>>      >
>>>>>>            
>>>>>>             
>>   
>>     
>>>>>>>    
>>>>>>>               
>>    >         
>>    
>>     
>>>>         
>>>>         
>>>>>>      > I propose
>>>>>>             
>> adding a update method
>>     
>>>>>>          
>>>>>>             
>>>> to
>>>>      
>>>>         
>>>>>> the Deployer interface or
>>>>>>            
>>>>>>             
>>   
>>     
>>>>>>>    
>>>>>>>               
>>    >         
>>    
>>     
>>>>         
>>>>         
>>>>>>      passing
>>>>>>            
>>>>>>             
>>   
>>     
>>>>>>>    
>>>>>>>               
>>    >         
>>    
>>     
>>>>         
>>>>         
>>>>>>      > the state as
>>>>>>             
>> an argument,
>>     
>>>>>>            
>>>>>>             
>>   
>>     
>>>>>>>    
>>>>>>>               
>>    >       
>>    
>>     
>>>>     I
>>>>      
>>>>         
>>>>>>>          
>>>>>>>               
>>         
>>     
>>>>>> would consider undeploy as the update
>>>>>>             
>> method you
>>     
>>>>>>          
>>>>>>             
>>>> can do
>>>>      
>>>>         
>>>>>>            
>>>>>>             
>>   
>>     
>>>>>>>    
>>>>>>>               
>>    whatever you
>>     
>>>>>>>    
>>>>>>>               
>>    >         
>>    
>>     
>>>>         
>>>>         
>>>>>>      want there, and
>>>>>>             
>> you can just ignore
>>     
>>>>>>          
>>>>>>             
>>>> at
>>>>      
>>>>         
>>>>>> when it call deploy
>>>>>>            
>>>>>>             
>>   
>>     
>>>>>>>    
>>>>>>>               
>>    method.
>>     
>>>>>>>    
>>>>>>>               
>>    >         
>>    
>>     
>>>>         
>>>>         
>>>>>>      (I know in
>>>>>>             
>> undeploy method you only
>>     
>>>>>>          
>>>>>>             
>>>> get
>>>>      
>>>>         
>>>>>> the filename, but
>>>>>>            
>>>>>>             
>>   
>>     
>>>>>>>        since
>>>>>>>               
>> your
>>     
>>>>>>>    
>>>>>>>               
>>    >         
>>    
>>     
>>>>         
>>>>         
>>>>>>      deployer is domain
>>>>>>             
>> specific you know
>>     
>>>>>>          
>>>>>>             
>>>> what
>>>>      
>>>>         
>>>>>> to do with the
>>>>>>            
>>>>>>             
>>   
>>     
>>>>>>>        file
>>>>>>>               
>> name)
>>     
>>>>>>>        >
>>>>>>>        >
>>>>>>>        >
>>>>>>>               
>> No, the issue is we
>>     
>>>>>>>        
>>>>>>>               
>>    
>>     
>>>> need
>>>>      
>>>>         
>>>>>>>          
>>>>>>>               
>>         
>>     
>>>>>> to invoke a different code in the case
>>>>>>            
>>>>>>             
>>   
>>     
>>>>>>>        of hot
>>>>>>>        >
>>>>>>>               
>> update.
>>     
>>>>>>>        Yes, as
>>>>>>>               
>> I mentioned
>>     
>>>>>>>        
>>>>>>>               
>>    
>>     
>>>> earlier if
>>>>      
>>>>         
>>>>>>>          
>>>>>>>               
>>         
>>     
>>>>>> the file is there then that is
>>>>>>            
>>>>>>             
>>   
>>     
>>>>>>>    
>>>>>>>               
>>    hot-update, else
>>     
>>>>>>>          
>>>>>>>               
>>         
>>     
>>>>>> un-deployment. So it should not be a big
>>>>>>             
>> issues.
>>     
>>>>>>            
>>>>>>             
>>   
>>     
>>>>>>>        >
>>>>>>>        >
>>>>>>>               
>> Anyway I feel I
>>     
>>>>>>>        
>>>>>>>               
>>    
>>     
>>>> should go
>>>>      
>>>>         
>>>>>>>          
>>>>>>>               
>>         
>>     
>>>>>> for a synapse deployer :-)
>>>>>>            
>>>>>>             
>>   
>>     
>>>>>>>        I
>>>>>>>               
>> though you already have
>>     
>>>>>>>          
>>>>>>>               
>>         
>>     
>>>>>> deployer for synapse.
>>>>>>            
>>>>>>             
>>   
>>     
>>>>>>> I mean a new deployer framework
>>>>>>>        
>>>>>>>               
>>    
>>     
>>>> implementation, not an
>>>>      
>>>>         
>>>>>>>          
>>>>>>>               
>>         
>>     
>>>>>> deployer.. anyway synapse doesn't have a
>>>>>>             
>> deployer
>>     
>>>>>>          
>>>>>>             
>>>> yet.
>>>>      
>>>>         
>>>>>>            
>>>>>>             
>>   
>>     
>>>>>>> Thanks,
>>>>>>> Ruwan
>>>>>>>
>>>>>>> -- Ruwan Linton
>>>>>>> Technical Lead & Product Manager;
>>>>>>>               
>> WSO2
>>     
>>>>>>>        
>>>>>>>               
>>    
>>     
>>>> ESB; http://wso2.org/esb
>>>>      
>>>>         
>>>>>>> WSO2 Inc.; http://wso2.org
>>>>>>> email: ruwan@wso2.com
>>>>>>>          
>>>>>>>               
>>         
>>     
>>>>>> <mailto:ruwan@wso2.com>;
>>>>>> cell: +94 77 341 3097
>>>>>>            
>>>>>>             
>>   
>>     
>>>>>>> blog: http://ruwansblog.blogspot.com
>>>>>>>          
>>>>>>>               
>>         
>>     
>>>>>>            
>>>>>>             
>>   
>>     
>>>>>            
>>>>>           
>>      
>>     
>>>>      
>>>>         
>>>        
>>>    
>>>       
>>     
>
>
>       
>
>   


Mime
View raw message