[10:28:30 CDT(-0500)] <athena> EricDalquist: do we have any documented SQL to determine how many users have customized their layout?
[10:28:39 CDT(-0500)] <EricDalquist> nope
[10:28:47 CDT(-0500)] <EricDalquist> and its more complex than that
[10:28:57 CDT(-0500)] <EricDalquist> you also need to account for any user that has portlet entity preferences as well
[10:29:13 CDT(-0500)] <EricDalquist> if you are considering exported layout #s
[10:29:32 CDT(-0500)] <EricDalquist> in theory select distinct user_id from up_layout_struct might work
[10:29:46 CDT(-0500)] <EricDalquist> but I'm guessing that in most installs users have layout data even if they haven't customized
[10:32:05 CDT(-0500)] <athena> yeah
[10:32:20 CDT(-0500)] <athena> i'm worried that it would - especially since this is a 3.0 install
[10:32:40 CDT(-0500)] <athena> think they're just trying to get a general idea of how much customization has been done, so ignoring portlet prefs is probably fine
[10:33:08 CDT(-0500)] <EricDalquist> you could probably do something like:
[10:33:28 CDT(-0500)] <EricDalquist> select user_id, count
as node_count from up_user_layout group by user_id
[10:33:39 CDT(-0500)] <EricDalquist> that would give you the # of layout nodes per user
[10:34:06 CDT(-0500)] <athena> makes sense - layout nodes in that case are essentially overrides, right?
[10:34:14 CDT(-0500)] <EricDalquist> and you might be able to simply say any user with X or fewer nodes haven't customized
[10:34:15 CDT(-0500)] <EricDalquist> right
[10:34:31 CDT(-0500)] <EricDalquist> like users without customization might just have 5-6 nodes
[10:34:51 CDT(-0500)] <EricDalquist> actually I think they would have the same node count as the defaultTemplateUser
[10:35:10 CDT(-0500)] <EricDalquist> but yeah, each node would be an override of some sort
[10:35:17 CDT(-0500)] <EricDalquist> with a few extra nodes for container nodes
[10:35:19 CDT(-0500)] <athena> thanks
[10:35:23 CDT(-0500)] <athena> i'll play around w/ some options
[11:45:08 CDT(-0500)] <EricDalquist> grr
[11:45:18 CDT(-0500)] <EricDalquist> this layout code is so inefficient
[11:45:29 CDT(-0500)] <EricDalquist> how can it be using more connections than there are threads in the jvm to export user layouts?