On Feb 11, 2005, at 9:12 PM, Mike Kienenberger wrote:
> Andrus Adamchik <andru..bjectstyle.org> wrote:
>> Out of all databases that we support, HSQLDB is probably the easiest
>> one
>> to get running and maintain. Download it, unpack, and run the startup
>> script. From there just update your ~/.cayenne/connection.properties
>> with
>> this data:
>
> The startup script doesn't exist :)
Sorry. Here is the one I use on Mac OS X:
#!/bin/sh
HSQLDB_HOME=~/opt/hsqldb
HSQLDB_JAR=${HSQLDB_HOME}/lib/hsqldb.jar
cd ${HSQLDB_HOME}/data && java -cp ${HSQLDB_JAR} org.hsqldb.Server &
> Eventually, I figured out I could simply stick
> test-hsqldb.jdbc.url = jdbc:hsqldb:mem:aname
> in my connections file, and include hsqldb.jar in the classpath.
So you are running with embedded DB instead of client-server setup.
That's what I like about HSQL - you can do nice things like that. I
guess the fact that this is not a "standard" procedure is that HSQL got
on my radar after I figured all the hard ways to do the tests...
Here is a random idea - we should kill our current "developer guide"
and move it to Confluence (of course after we complete its testing and
I add backups) - all such nice little tricks have to be easily
documentable.
> Also, are the regression tests supposed to work? Mine all fail with a
> similar error to below using jdbc:hsqldb:mem:aname: as well as with
> OpenBase.
No, IIRC I could only make them work on Oracle and we haven't looked at
them in a while... They were supposed to be used specifically for
regression testing of entity sorting. As we are not making changes to
this code, I neglected the tests.
Andrus
This archive was generated by hypermail 2.0.0 : Sat Feb 12 2005 - 00:09:34 EST