Re: [jira] Commented: Search Components (plugins)

classic Classic list List threaded Threaded
2 messages Options
Reply | Threaded
Open this post in threaded view
|

Re: [jira] Commented: Search Components (plugins)

Yonik Seeley-2
On Feb 1, 2008 7:41 AM, Michael Dodsworth (JIRA) <[hidden email]> wrote:
> As an aside, method signature changes are usually trivial to fix;

If there is a general policy of not making backward incompatible
changes in point releases, it really puts developers in a bind.  For
Solr, that really only strictly extends to classes that users are more
likely to extend or implement.

Regarding QParser specifically, it hasn't been in an official release
yet, so we can still make changes if we want.

> I would love to contribute a move towards using Spring.

And I'd love to see a move to something like Spring happen :-)
It's not on any roadmap, but it's been discussed a number of times in the past.

-Yonik
Reply | Threaded
Open this post in threaded view
|

Re: [jira] Commented: Search Components (plugins)

Michael Dodsworth
Yonik Seeley wrote
On Feb 1, 2008 7:41 AM, Michael Dodsworth (JIRA) <jira@apache.org> wrote:
> As an aside, method signature changes are usually trivial to fix;

If there is a general policy of not making backward incompatible
changes in point releases, it really puts developers in a bind.  For
Solr, that really only strictly extends to classes that users are more
likely to extend or implement.

Regarding QParser specifically, it hasn't been in an official release
yet, so we can still make changes if we want.

> I would love to contribute a move towards using Spring.

And I'd love to see a move to something like Spring happen :-)
It's not on any roadmap, but it's been discussed a number of times in the past.

-Yonik
Great! Figuring out a plan of attack is going to be tricky. I envisaged using spring for most (if not all) of the configuration - removing all the boiler-plate loading/configuration code completely...but that's not going to work for Solr users.

I think slowly removing some of static references and allowing Spring to deal with the singleton creation and injection would be the first step to take. I'll try to get something in a patch as-soon-as-possible. I'm in the midst of porting our dynamic multicore changes over to the latest drop of Solr, which is taking up most of my time at the moment. I took a similar route to Solr-350...so hopefully this stuff can also be contributed.

I'll post details on my progress when I have something to report.
Thanks,