On 9/1/05, Elias Torres <[EMAIL PROTECTED]> wrote: > Thanks Dave, > > Finish the Roller 2.0 EA demo and I'll be working on a Roller that is > deployable out of the box using Derby. Imagine roller being deployed > w/o having to install a database? I think this will be great.
FWIW - Roller already has this functionality with HSQL. Dave distributes a "demo" version that includes Tomcat and runs w/o installing anything - just unzip and start Tomcat. ;-) Matt > > Elias > > PS. The only remaining issue is the long index names for DB2 and the > necessary dialects that will have to include for DB2 and Derby because > hibernate2.1 is just too old. > > On 9/1/05, Dave Johnson <[EMAIL PROTECTED]> wrote: > > Thanks Elias. > > > > I think this stuff should be addressed and I will attempt to do so in > > the createdb script. IMO Derby is a key database for Apache, IBM and > > Sun and we need to support it. I hope to get to it tomorrow, but > > probably after I make the Roller 2.0 EA demo available. > > > > - Dave > > > > > > On Sep 1, 2005, at 4:36 PM, Elias Torres wrote: > > > > > Inspired by the message from Brent [1] and with the help of IBM > > > colleagues on the Derby end, we've seem to find some other things in > > > the script that could be improved so we can support Derby 10.1 as > > > well. > > > > > > 1. website table create an "extra" index on the primary key. This > > > gives a warning on both derby and db2, not sure what do the other > > > databases do in this case. Is it ok if we remove the extra index? > > > [[[ > > > id varchar(48) not null primary key, > > > ]]] > > > > > > [[[ > > > create index website_id_index on website(id); > > > ]]] > > > > > > 2. I believe that the rag_group table adds a constraint the handle > > > column to be unique and also creates an index. Again, both derby and > > > db2 think it's redundant. > > > > > > [[[ > > > alter table rag_group add constraint rag_group_handle_uq unique ( > > > [EMAIL PROTECTED]@ ); > > > create index rag_group_handle on rag_group([EMAIL PROTECTED]@); > > > ]]] > > > > > > 3. There's an inconsistency in a foreign key relationship and the > > > table definition of the webpage table. website.id is VARCHAR(48) and > > > webpage.websiteid is VARCHAR(255). Is this an oversight or can we > > > change websiteid in webpage to also be VARCHAR(48), this way Derby > > > won't give us an error. > > > > > > from webpage table: > > > [[[ > > > websiteid varchar(255) not null, > > > ]]] > > > > > > [[[ > > > alter table webpage add constraint weblogpage_websiteid_fk > > > foreign key ( websiteid ) references website( id ) > > > ]]] > > > > > > from website table: > > > [[[ > > > id varchar(48) not null primary key, > > > ]]] > > > > > > Again, thanks for all your support and time with all these requests, I > > > know you are very busy doing the next release, especially Dave. > > > > > > Elias > > > > > > [1] http://tinyurl.com/doqut > > > > > > > >
