Jun 15 17:17:13 mst: can you tell frew what you told me about FilterColumn @ GPW Jun 15 17:18:11 I said two methods Jun 15 17:18:15 filter_column(... Jun 15 17:18:20 and filter_column_row(... Jun 15 17:18:28 frew: the idea being if the row object is only in one signature Jun 15 17:18:36 hashrefinflator can assume the non-row ones are safe Jun 15 17:29:31 oh clever Jun 15 17:29:44 hmmmm Jun 15 17:29:55 * frew wonders why we didn't do this when we discussed it before Jun 15 17:30:07 frew: mst did suggest this Jun 15 17:30:14 I think I was hung up on that if we did that we'd need a FC and an FCLite Jun 15 17:30:27 I don't remember the suggestion of two sep methods at al Jun 15 17:30:33 but I don't doubt that he said it Jun 15 17:31:44 when he said it to me in .de I still didn't get it Jun 15 17:31:51 so obv. the problem is in mst Jun 15 17:32:05 so I am getting it right? Jun 15 17:32:08 two sep methods? Jun 15 17:32:09 yeah Jun 15 17:32:12 that's totally simple Jun 15 17:32:20 except we use the already existing method for objs Jun 15 17:32:37 and make a new one for ($data, $rsrc) Jun 15 17:32:42 exactly Jun 15 17:41:09 mst: I already said - I might be overtinking this, two sep. coderefs are fine Jun 15 17:41:24 mst: two non-overlapping signatures Jun 15 17:41:47 mst: the original name is obj-first filters Jun 15 17:41:47 yeah Jun 15 17:41:57 a new name for $val, $rsrc filters Jun 15 17:41:58 happy? Jun 15 17:42:17 frew: ^^ please note the bastard is happy Jun 15 17:42:23 $source, $val though Jun 15 17:42:32 so you can use method names on the source, as originally discussed Jun 15 17:42:56 wtf... (wrt originally discussed) Jun 15 17:43:02 fine $source, $val Jun 15 17:43:10 I dunno if we'll ever use that Jun 15 17:43:17 but applying a role to the source smells like fun Jun 15 17:43:22 mst: meh, it's proactive I like it Jun 15 17:43:49 mst: the wtf is you assuming you said what you were thinking :) Jun 15 17:45:27 ribasushi: I'm ignoring this till I have to code on it Jun 15 17:45:39 frew++ # smart man