perfsonar development work

["Jeff W. Boote" <>]

 wrote:
> Author: mac
> Date: 2007-02-21 10:09:20 -0500 (Wed, 21 Feb 2007)
> New Revision: 2119
>
> Modified:
>    
> branches/XML-LS-RELEASE-1.1/schema/example-instances/perfSONAR/LS/test/EchoRequest.xml
> Log:
> typo in eventType
>
>
> Modified: 
> branches/XML-LS-RELEASE-1.1/schema/example-instances/perfSONAR/LS/test/EchoRequest.xml
> ===================================================================
> --- 
> branches/XML-LS-RELEASE-1.1/schema/example-instances/perfSONAR/LS/test/EchoRequest.xml
>       2007-02-21 14:35:42 UTC (rev 2118)
> +++ 
> branches/XML-LS-RELEASE-1.1/schema/example-instances/perfSONAR/LS/test/EchoRequest.xml
>       2007-02-21 15:09:20 UTC (rev 2119)
> @@ -8,7 +8,7 @@
>                     xmlns:nmwg="http://ggf.org/ns/nmwg/base/2.0/";>
>  			     
>      <nmwg:metadata id="meta">
> -         <nmwg:eventType>echo.ls1</nmwg:eventType>
> +         <nmwg:eventType>echo.ls</nmwg:eventType>
>          <nmwg:parameters>
>  	 
>               <nmwg:parameter name="testDBConnection" value="yes"/>

Maciej, (and other service developers),

Since it looks like you are changing the eventType value anyway, is there any 
reason not to use the agreed to naming scheme? (See the section under Metadata 
particulars at http://wiki.perfsonar.net/jra1-wiki/index.php/Generalities )

It has been made very clear that a naming scheme is required to ensure there are 
no namespace collisions in the eventType. The naming scheme that is documented 
was selected to allow for the possibility of dynamically loading classes to 
handle those specific types of data. There was no dissent in the Cambridge 
meeting or on the mailing list. As requested at that meeting, I have written up 
the results.

As no one has made any dissenting opinions - I suggest that developers need to 
start adhering to this naming scheme as soon as possible. Note, I am not saying 
that eventTypes that are currently used should be discarded. However, I do think 
it would be reasonable to have services recognize the new eventType values in 
addition to the current ones. That way clients can start using the new naming 
scheme immediately.

In all honesty, I believe we need a perfSONAR specifications document that is 
much more oriented along the lines of this document. (And should hopefully be 
based on an NMWG specifications document.) But, until we have that - I suggest 
we use this part of the wiki to start formalizing our specification.

jeff