> I'd prefer not rushing that in, as it is a big contribution and needs  
> a bit more review.  (personally I think Spring MVC is overkill and  
> less flexible, more verbose, than alternatives, but I'm strongly  
> opinionated and will defer to others that desire this UI in Solr's  
> codebase).

It probably is, but in terms of the code you have to write for the model and
controller components, I think it's pretty good. I'm not a fan of heavyweight
solutions myself, but as Solr itself needs to be deployed to a servlet
container, using the same technologies for a UI seemed like a good choice.

And in the end this will be but one of a large range of example Solr user
interfaces.

> Feedback: I noticed an error when just pressing the "Search" button on  
> your demo without putting in a query.  You'll probably want to default  
> to a *:* query.  And it'd also be cool to see facet values displayed  
> _before_ doing a "search" so a user can browse without having to do a  
> full-text search.

Ok, I hadn't considered that. The main purpose of the UI was to provide search
functionality, and facets and everything else was designed more or less on top
of that.

I wouldn't want to make any changes now, but once this is part of Solr this can
be added quite easily.

> Procedure-wise, most patches, including large ones like the  
> DataImportHandler, are submitted via JIRA and assigned to the ASF,  
> allowing folks one spot to download and comment on them.  I consider  
> that the first step to your push to get this in.  Wiring your  
> contribution in to the example area would ensure an easy way for folks  
> to try it out.

Ok. It'd be good if somebody could take a close look at it before though,
because there're some jars which Solr uses as well, and dependencies on a Solr
war for integration testing. Maybe the build script can be modified to not
duplicate any jars in the repository and integrate the whole thing with the
top-level build file.

As for the timeframe, I'm obviously keen on getting it in as quickly as possible
:), but I appreciate that you need to take a closer look at it. Let me know if
there's anything I can help you with.

Lars

Reply via email to