UCD problem in SSA/SpectrumDM

Douglas Tody dtody at nrao.edu
Sat Nov 28 14:50:34 PST 2009


Hi Petr -

>> We already have protocol versioning support in these interfaces,
>> and for it to work properly it is important to observe these rules
>> as Bob notes below (in addition this represents standard practice
>> outside astronomy as well).  So very minor changes or minor document
>> clarifications without changing the intent might be a level 3 revision
>> (1.04 -> 1.05);
>
> Does to this level 3 belong the addition of some explanatory comments and 
> some "best practise" recommendations  ?  And what about adding whole example 
> - e.g. of ground base high resolution spectra?
> All the mandatory-until-optional paramaters will be unchanged, only more 
> informative stuff added?
>
> What do you think ?

I think we need to keep the changes to a minimum for a micro-update.
Adding extensive new text is probably out of scope.  Otherwise we will
likely require a more extensive review.

I agree that having more user documentation would be very helpful
for something as complex as SSA, however perhaps at some point this
should go into a FAQ or something (a user guide is also a possibility),
instead of the spec itself.  The advantage of something like a FAQ
would be that we could address any number of points, and could add
entries at any time without having to revise the standard itself.

> I am repeating again that for the success of SSA more examples of a really 
> runing service is indispensable - all the implementors should be able to 
> understand how the service works by looking in the Registry, finding the 
> service URL and playing with it. They can compare what is returned and lookup 
> in the (hopefully added)  example what does it mean.

Yes, reference implementations illustrating these capabilities and
the query response are very helpful to understanding a complex spec.
We have made a good start at this, however:

Ideally we should have some reference/test services and a FAQ or
user guide, in additional to the spec itself.  This could help the
implementors quite a lot! (also our ready to use service frameworks,
which are coming online slowly).

> And to be honest - the proxy for JHU spectrum is not in registry (Or I have 
> overlooked it - then sorry).

This is a good point - although we do not have any good way yet to
identify a reference or test/verify service implementation in the
registry.  Nonetheless I guess it could be done with a descriptive
comment.  I agree I should register my reference/test SSA service.
In part, what happened with this service is that it was written before
we could register SSA services properly in the registry.  Once the
micro-update is done to update the spec, it would be a good time to
register the test/demo services.

 	- Doug



More information about the dm mailing list