Search
19:36:00
drmeister
I kept both classes for the time being so I could switch back and forth to debug it.
19:36:18
drmeister
You are now removing those vestigal, useless template classes. We keep the `Simple` ones.
19:36:38
drmeister
Nothing refers to them - they are never expressed - they are dead code.
19:37:35
Bike
looks like there's a similar issue in core/wrappers.h
19:40:37
drmeister
These are generating compile time errors right?
19:40:50
Bike
i have deleted a whole bunch of identifiers that these refer to.
19:40:54
drmeister
Can you comment out the offending functions with #if 0/#endif ?
19:40:59
drmeister
I'll clean them out later.
19:41:29
Bike
can i? i mean, this is core::BuiltinClosure_O, is it not in use?
19:41:37
Bike
there's an #if 1 above it, so i guess i just toggle that
19:41:37
drmeister
Crap - I can't load slime because it wants compile-file.
19:41:53
drmeister
BuiltinClosure_O is doing stuff?
19:42:36
drmeister
Oh wait. You can't remove LambaListHandler_O yet. Methods still use them.
19:42:41
drmeister
Crap - sorry - I forgot.
19:42:52
drmeister
I only got part way through what I wanted to do there.
19:42:59
Bike
toggling that did not work because it tries to include wrappers_methoids.h instead, except that file doesn't exist
19:43:06
Bike
it's fine, not like we're in a rush here
19:43:10
drmeister
We got the whole discriminating function dohicky thing involved there.
19:43:34
drmeister
Ok, if it compiles the C++ then you didn't cut too deep.
19:44:14
Bike
the stuff i pushed to the vm-sans-interpreter branch builds fine. i'm just trying to cut more now, and that's not working
19:44:47
drmeister
I know - I'm using it already - but since we don't have compile-file for bytecode I'm rethinking my options.
19:45:02
drmeister
I'll go back to M-x run-lisp
19:47:28
yitzi
duh....I get it via `(fdefinition '(setf logical-pathname-translations))`
19:48:59
drmeister
https://getyarn.io/yarn-clip/f7421a44-6336-43d8-ac14-a90249a51f4e
20:40:34
Bike
drmeister: when you say methods, you mean single dispatch?
20:53:13
yitzi
dang that bytecode compiler is fast.
21:01:34
yitzi
About to push to vm-boehm to reduce the stage count to 3
21:03:08
yitzi
drmeister: the translation tables are now loaded in stage 0.
1:11:40
drmeister
Bike: Yeah - single dispatch.
1:12:13
drmeister
I just got back from dinner with a visiting speaker.
1:12:20
drmeister
I'm working on vm-boehm now.
4:02:32
Bike
the lambda lists for those are handled differently?
4:29:38
Bike
i mean, i know there's the implicit "this" parameter, but other than that it should be the same, right?