Hi,
HSQLDB is actually reasonably fast, but it has other problems, namely that
it stores whole DB tables in memory so if your crawl is large enough it
will run out.
The reason for Documentum connector slowness is almost always poor
Documentum performance, and has nothing to do with MCF itself.
Thanks for the information, Karl.
I have a couple more questions:
It looks like the format is being applied in cases where ordering might
not really matter, e.g. listing jobs as in the example I gave. (Or
perhaps order does matter here?)
Is there any more doc on the non-sugared format than
The output format did change, and the reason was because the "syntactic
sugar" format would not preserve ordering, so that if you output and
re-input, you'd lose information.
The more complex form is being used only where there is a possibility of
ordering confusion. It was always accepted as
