12:40:01scymtymjackdaniel: i didn't get to the visual appender for log4cl yet, but i have accumulated a bunch of cleanup changes on the way. would you be willing to review a pull request?
13:36:34jackdanielscymtym: all looks good to me, I have some objections regarding bumping system definition requirement to ASDF >=3.1 (though BT already does that for us and we depend on it)
13:36:57jackdanielI'd still add explicit package accessors
13:37:58scymtymjackdaniel: thanks, reading comments now
13:41:05scymtymjackdaniel: the final comment is funny because i specifically avoided UIOP:SYMBOL-CALL because i thought you wouldn't like it :)
13:42:30scymtymas you mention, because of bordeaux threads, the compatibility point is kind of moot
13:43:26scymtymi see the benefits of using (asdf:defsystem …) for SLIME, though
13:43:38jackdanielI'd keep asd definitions compatible with ASDF 2.26, but I won't fight it if we jump into ASDF train (of constant change). I'd still add package accessors
13:44:21jackdanielso if you have stronger than barely-against opinion, then go for it. in that case though I'd add error (like BT does) if older asdf is used
13:44:31jackdanielso there will be a clear message instead of some random failures
13:45:59scymtymi will keep/add ASDF: package to DEFSYSTEM forms and fix the typo in any case - those are easy
13:52:43scymtymok, i will reduce the ASDF-related changes to a minimum (remove log4cl.system package, asdf:defsystem everywhere, string designators for systems) and leave the rest for later. that way, there should be no fallout from breaking clisp compatibility