<headius>
seems consistent, may be a valid platform difference
<headius>
since it is just these two failing I will skip them on Darwin for now but if you can look into that it would be helpful
<headius>
I tested on Darwin x86_64 if it matters
<headius>
I guess we need to get a Darwin CI set up eh? ðĪŠ
<headius>
and we need an automated release process... I forgot to update things along the way again ðĒ
<headius>
thank goodness for maven release staging
<headius>
there's really not much value to having these be separate libraries when we have to update all dependencies to appease maven... just a lot of separate error-prone releases every time something changes
<basshelal[m]>
I was just thinking about that yesterday like "Damn, maybe some of the library locating tests will fail on Darwin" ð
<basshelal[m]>
I'm not too surprised given our recent troubles
<basshelal[m]>
Added an issue for the failing tests so that I can deal with it later when I have time
<basshelal[m]>
Been quite busy with my day job recently ðïļ
<headius>
basshelal: ahh good call
<headius>
enebo: so you mentioned last 9.3 issue but I see two
<headius>
I assume you mean the one related to JI concrete class extension... the refinement issue could be punted and looked at after 9.3 release
<headius>
I describe the problem there... we copy the refinements at "using" time and MRI continues to reference them from original location, so we aren't seeing new refinements adding to an existing refinement module
<headius>
so there's a clear fix it is just cumbersome and refinement method lookup and caching needs some overhaul anyway