--- Log opened Sun Sep 26 00:00:18 2021
00:38 < darosior> Ok then
03:48 < darosior> sipa: could you share with me the script you generated with gramtropy here https://github.com/sipa/miniscript/pull/66#issuecomment-924454783 ? Could be a nice addition to the fuzzing corpus
06:47 < darosior> So for and_or() both Y and Z must be of the same type? We'd need a way to check for the type without properties
06:49 <@sipa> why?
06:51 < darosior> Because GetType() == GetType() may fail because properties of Y and Z may diverge
06:52 < darosior> For instance take Y is older(...) and Z is sha256(...)
06:53 < darosior> Older may have "g" or "h" but Sha256 never
06:54 < darosior> I'm assuming we'll merge the timelock mix branch
07:06 <@sipa> ok but why do we need to be able to test for that without using type properties?
07:06 <@sipa> the type system enforces these things
07:09 < darosior> When parsing an and_or() i wanted to test that Y->GetType() == Z->GetType() to enforce it. How could the type system help here?
07:10 <@sipa> why not do it after?
07:10 <@sipa> construct the node, and if it is not IsValid(), abort?
07:16 < darosior> I overlooked that we'd only give the type property to the parent node if the subs are of the valid type.. It's much smarter than my coarse solution: https://github.com/sipa/miniscript/pull/72/commits/f3519f5a2dde07e7020400e0ba10d5d22fbe15fa
07:21 < darosior> Thanks.
09:00 -!- robertspigler [~robertspi@2001:470:69fc:105::2d53] has quit [Quit: You have been kicked for being idle]
22:48 -!- Netsplit *.net <-> *.split quits: hugohn, jnewbery
22:48 -!- Netsplit over, joins: hugohn, jnewbery
22:54 -!- Netsplit *.net <-> *.split quits: sanket1729, sanket_cell
22:56 -!- Netsplit over, joins: sanket_cell
22:56 -!- Netsplit over, joins: sanket1729
--- Log closed Mon Sep 27 00:00:19 2021