|
From: Rod J. <rod...@in...> - 2004-06-15 06:26:14
|
Sean > > I'd vote for Toplink support first. I know it's not open source but a lot > > of us have to use it and I am pretty sure there are several Toplink users > > on the list, each of us rolling our own Spring integration. I don't think it's either/or. Ideally we'd like the development for either of these to be done outside the core Spring team--with our support of course. Especially as Juergen and I (mainly Juergen!) are currently putting a lot of effort into improving our JDO integration, which is already part of Spring. TopLink support would be good to have, and the topic has come up a few times, mainly in forums. I think someone once volunteered to contribute some code, but it didn't happen. Good TopLink integration is also non-trivial, because of the need to register objects selectively with the unit of work (via cloning) to minimize the cost of dirty checking. This makes the issues a bit different to those with Hibernate or JDO, for example. I know TopLink pretty well myself, but I don't have the time to actually do the integration, as it isn't currently a priority for a client. (It came close to being a priority for me last year.) I'd certainly be happy to advise anyone who wanted to work on it in Spring's data access architecture, and the issues I see in TopLink integration: getting it right is an interesting challenge. I tried to interest the TopLink team in Spring integration a few weeks ago--after an encouraging suggestion from someone at Oracle--but so far haven't had a response. We certainly don't have any objection to integrating with a product just because it's not open source. If it's a proven product and people want to use it, we're in favour of integrating with it... I'd be interested to hear about what solutions Spring/TopLink users are using now. R |