Andrus Adamchik wrote:
>
> On Sep 15, 2005, at 3:08 PM, Gili wrote:
>
>> Coming from a Hibernate background ...
>
> But this is exactly where your usability complaints are coming from.
> You are porting an existing application written with a different
> underlying design philosophy. Nobody ever claimed that *porting from
> Hibernate* is easy.
Yes and no. This is true for why I was complaining about how unobvious
it was to flush the context cache but it isn't true for other aspects of
the API I'm not so crazy about.
For the most part I think I can be appeased by adding a more
task-oriented documentation similar to the XStream "two minute
tutorial": http://xstream.codehaus.org/tutorial.html
If you give me Wiki access I can add a huge bunch of those... but you
guys still need to be open to API changes :) As I was mentioning to Mike
a few minutes ago, I consider query caching to be a very problematic
feature (I've totally disabled it for now). As a new user I fired up the
modeler, created some queries and when I saw the query caching option I
enabled it because "heck, who wouldn't want caching?" only to find out
much later on that huge implications of that setting. The documentation
should have mentioned in big bold letters the implications of this
setting :)
Can you please add a new API for selectively flushing query cached
results? i.e. query.invalidateResults() would flush "local" or "shared"
cache results.
Gili
This archive was generated by hypermail 2.0.0 : Thu Sep 15 2005 - 15:39:59 EDT