evaluate strikes again :/ Stuart ran into this, so I went hunting in evaluate to find it not abiding by the allon setting correctly. Specifically, if allon=1, return every atom (removing the conditional ops); if you look in the attached patch, you'll see the flaw straight away, w/ allon it was still filtering out !blah? ( uri1 uri2 ) operators.
Created attachment 35531 [details, diff] 2.0.50-r9-evaluate.patch This corrects all known issues w/ .50 evaluate.
The fix for bug #16159 is included in this patch.
Could the "Copyright header Error. Possibly date related" repoman warnings be fixed with a -r10, too - please!? (Yes, I'm too lazy to patch)
Could ths one be fixed please? There are quite a lot ebuilds affected. We can't wait weeks until .51 is ready.
blah