On 05/11/10 04:31 PM, Alok Aggarwal wrote:
On Tue, 11 May 2010, Sarah Jelinek wrote:
On 05/11/10 03:30 PM, Alok Aggarwal wrote:
Talking with Jean, another thing that came up
with respect to Checkpoints was - do we expect
the Checkpoints to be specified in a manifest?
I'm thinking more in terms of the DC manifest here.
And, will an end user be allowed to insert/remove
arbitrary checkpoints from an installer (DC or some
other)? If yes, will the checkpoints have to be
implemented much like some of the other standard
checkpoints such as TI and Transfer or can they
be more free form shell scripts?
Darren/Sarah, do you know?
Yes, the manifests will have to have the checkpoints defined and they
are registered from that for the AI and DC applications. Checkpoints
must implement the checkpoint interface.
ace, but the internal workings are up to the checkpoint developer.
Although, I wouldn't think we need scripts to do any checkpoint work.
That would seem to mean that the scripts would
have to be in python and not shell scripts. That's
most certainly a requirement then, correct?
In the long run I think python is the way to go. For the prototyping
effort for DC to get hooked into the engine I think leaving them as
shell scripts wrapped in a python object and called with the execute()
checkpoint method would be ok.
sarah
Alok
_______________________________________________
caiman-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/caiman-discuss