perfsonar development work

[michael <>]

Hello all,

Antoine, while I'm all for avoiding duplication I'm not sure I'm a fan 
of coupling either.

While current usage is all soap-over-http it doesn't mean there might 
not be a future
use-case where the option of using an other transport protocol for SOAP 
is preferred.

By using the http status codes you tie/(tightly) couple the protocol to 
http.

From a practical point of view it might also not be the way to go as 
developers at the
moment are insulated from having to bother with the http layer

SOAP itself also has 'status' part, which probably triggers similar 
reservations.

Best regards,

Michael

Antoine Delvaux wrote:
> Hi Slawomir, Jeff, All,
>
> Yes, this is good we discuss all that, thanks for initiating  
> discussion Slawek.
>
> Jeff, I like your view about making things as simple as possible for  
> the client, this will make the protocol simpler, I guess this is all  
> what we want to achieve.  I also like your idea to mimic HTTP status  
> messages.
>
> However, I don't understand why we need another layer on top of the  
> existing HTTP status messages, but maybe it's because I'm not that  
> familiar with pS protocol, bear with me then.  Couldn't we use HTTP  
> status messages and add to it more description to the result code.  A  
> full response (including HTTP headers) would then be something like:
>
> _Success example_:
>
> HTTP/1.1 200 Ok
> Content-Type:"text/xml"; Charset="utf-8"
> Content-Length: nnnn
>
> <nmwg:message type="EchoResponse">
> <nmwg:metadata id="resultCodeMetadata">
>    <nmwg:eventType>echo</nmwg:eventType>
> </nmwg:metadata>
> <nmwg:data metadataIdRef="resultCodeMetadata">
>    <nmwgr:datum>This is the echo response from the service.</ 
> nmwgr:datum>
> </nmwg:data>
> </nmwg:message>
>
> _Error example_:
>
> HTTP/1.1 500 Internal Server Error
> Content-Type: "text/xml"; Charset="utf-8"
> Content-Length: nnnn
>
> <nmwg:message type="EchoResponse">
> <nmwg:metadata id="resultCodeMetadata">
>    <nmwg:eventType>not implemented</nmwg:eventType>
> </nmwg:metadata>
> <nmwg:data metadataIdRef="resultCodeMetadata">
>    <nmwgr:datum>EchoResponse command doesn't exist.</nmwgr:datum>
> </nmwg:data>
> </nmwg:message>
>
> On a side note, except for the sake of coherence between requests and  
> answers, I don't see the need to use URL in eventType naming.  IMHO,  
> it seems to me that URN would be more suited in this situation.  Or is  
> there some convention in OGF that it's best to use URL?
>
> Best regards,
>
> Antoine.
>
> Le 8 nov. 2009 à 08:36, Jeff W. Boote a écrit :
>
>