#ckan

/

      • kalxas has quit
      • slacko11820 joined the channel
      • tomkralidis joined the channel
      • tomkralidis has quit
      • TheCoffeMaker has quit
      • yanik joined the channel
      • yanik has quit
      • yanik joined the channel
      • realitygaps joined the channel
      • realitygaps has quit
      • realitygaps joined the channel
      • rossjones
        tktech not that I know off, will ask at the tech call today though if you'd like?
      • kalxas joined the channel
      • kalxas has quit
      • kalxas joined the channel
      • shellac joined the channel
      • yanik has quit
      • kalxas has quit
      • tomkralidis joined the channel
      • slacko11820 has quit
      • wardi
        rossjones: might be just you and i today
      • rossjones
        will be quick then :)
      • wardi I've done the 2 issues, and no new PRs.
      • Only other thing is tktech's question about Solr5 docs/.
      • wardi
        i don't know of any
      • i was going to submit our info on the SFC form to set up a meeting
      • that's all i've got. meeting done?
      • rossjones
        You can go back to bed now ;)
      • wardi
        sounds good
      • rossjones
        We should try and avoid you having to get up early if the meeting isn't likely to happen
      • wardi
        nah, this is normal schedule for me anyway.
      • rossjones
        I'll join the call in case anyone turns up.
      • kalxas joined the channel
      • kalxas has quit
      • kalxas joined the channel
      • yanik joined the channel
      • joetsoi isn't working on ckan any more, is he?
      • tktech
        rossjones, did the meeting already finish?
      • rossjones
        It did, 3 hours ago ;)
      • 12UTC on Tuesdays
      • tktech
        Welp
      • rossjones
        wardi mentioned above he didn't know of any Solr5 docs ... so contributions very welcome ;)
      • It's later on Thursdays.
      • It's 14:00UTC.
      • tktech
        The documentation needs updates for 4.3 and 5, both need changes to the default schema.xml and changes to the solrconfig.xml (due to knew Boolean types)
      • rossjones
        Ah, that's going to be more problematic I think.
      • Am sure I run on 4.3 somewhere with the standard schema though
      • tktech
        Depends on what point version you're running :)
      • Looks like we actually discussed one of the small changes in September
      • Do the circle.ci tests run against multiple solr versions?
      • That might be a good change, I think
      • rossjones
        I can't assign you that bug unless I add you to the contributors team. May I do that?
      • tktech
        Sure, I don't mind.
      • This issue is fixed in Solr in 5.2.2, but 5.2.1 is still the most common in the ubuntu repos.
      • rossjones
        Do you think it'll need a second schema alongside the other one, and documentation around which to use?
      • tktech
        Nope, the changes should work for >=4
      • For example, the default solrconfig.xml in 5 references new boolean types
      • But the default CKAN schema.xml is missing the type definition.
      • Simply adding it to schema.xml for all versions is fine
      • rossjones
        Oh that's good. And testing against multiple solrs in circle would be great. I think there's a hope we can move to circle eventually.
      • tktech
        Looks like it's currently testing against 4.3.1
      • Ugh, Ubuntu 12.04LTS has solr 1.4.1, 14.04LTS has 3.6.2.
      • rossjones
        :) Let me know when the invite arrives via github and I'll assign you that ticket ;)
      • I hate solr
      • tktech
        rossjones, I've accepted it
      • Is there some specific case for keeping solr around?
      • rossjones
        Welcome to the madhouse.
      • tktech
        Postgres has more than sufficient plain-text search capabilities for what CKAN actually exposes, and as a caching layer it leaves a lot to be desired.
      • rossjones
        I think we use Solr because it works, and nobody wants to replace it, or abstract it out to support more than one backend
      • I think it is the lack of facets that is the problem with pg
      • One day, I hope we can get to a place where we don;t need to cache in Solr, but we're probably still some way.
      • tktech
        Out of the box faceting support is a good point
      • I think the background workers will probably come first.
      • Redis being a hard requirement means you have another option for caching
      • rossjones
        This is true, when I ever get around to finishing that PR
      • yanik has quit
      • yanik joined the channel
      • yanik has quit
      • yanik joined the channel
      • kalxas has quit
      • yanik joined the channel
      • yanik has quit
      • kalxas joined the channel
      • magically_ joined the channel
      • magically_ is now known as magically
      • LondonAppDev__ joined the channel
      • rossjones has quit
      • rossjones joined the channel
      • yanik joined the channel
      • tomkralidis has quit
      • wardi
        rossjones: multi-select facets development is in my future, it seems solr already has a really nice way of handling that (exclude the current facet when returning numbers for each facet)
      • oh, actually not on my list yet, tktech might be the one working on it
      • yanik joined the channel
      • LondonAppDev__ has quit
      • TheCoffeMaker joined the channel
      • yanik has quit
      • yanik joined the channel
      • yanik has quit