Ruwan
I agree with Paul although I like the technical approach you describe
here. Using multiple namespaces would require one to define them with
prefixes and use the correct ones when writing a configuration by hand -
which might incur an additional burden.
asankha
Paul Fremantle wrote:
Ruwan
Can you please post a more detailed example of what the synapse config
language would look like?
I think there is a challenge which is that you want to make it easier
for a machine to read.... but maybe harder for a human to write! I'm
not convinced that is the right balance.
Paul
On 12/13/06, Ruwan Linton <[EMAIL PROTECTED]> wrote:
Hi,
Currently we have a common namespace for every element used within a
synapse
configuration.
This is ok when we are processing the configuration with Java (as the
MediatorFactory that loads the mediators knows each available
mediator), but
if someone wants to process this configuration as XML (for example using
javascript) then he/she has a set of XML elements with the same
namespace
and he/she will not be able to distinguish between mediator elements
and the
non-mediator elements such as endpoints etc. unless we maintain a
list of
mediators to compare with the known element names. This also is
difficult
when we use an XSLT processor to process our configuration.
To overcome this issue I think it is better to change the namespace of
mediators from the general synapse namespace so that one can compare the
namespace of the element and figure out whether it is a mediator
element or
not.
Any comments....
Thanks,
Ruwan.
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
