On 05/11/10 11:53, [email protected] wrote:
Hi Ethan,
Thank you for the clarifications here. One question, below:
On Fri, 7 May 2010, Ethan Quach wrote:
[snip...]
On 05/04/10 12:28, Sarah Jelinek wrote:
[snip...]
-verify(), why do we need this? It seems to me when a user sets an
xpath value
we could do verification then, or at load of the new instance
document. I am
not sure we need a separate verify. Shouldn't it just be automatic?
Because the separate set() calls are used to set individual elements,
the state after a set possibly won't validate.
Since the user is writing a script to generate the XML, I don't see
why they can't provide us full chunks of XML at a time to the set()
operation. I'm unaware of where one could provide a chunk of XML which
would require a disjoint (or non-continuous) change elsewhere; but
perhaps I'm not being creative this morning.
Dave commented on something similar to this as well, but as
an augmentation for the load subcommand to accept incremental
block loads. I think that should fulfill what you're asking for here.
I do think the continuous validation would be very helpful for
debugging when an invalid bit of XML gets inserted to prevent having
to debug manually. I feel this way especially since we're providing
this approach under the assumption that sys. admin.s' knowledge of XML
is severely lacking.
As already mentioned, we can't validate upon every set call because the
state of the manifest in between set calls during the execution of the
script possibly won't validate. One thing we can probably do is have
the aimanifest command keep a transaction log for itself and have it log
the return status of every call and the validity state of the manifest after
every call. Having this log file should help with debugging during the
dev phase of the scripts.
-ethan
Thank you,
Clay
_______________________________________________
caiman-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/caiman-discuss
