aboutsummaryrefslogtreecommitdiff
path: root/dev/doc/debugging.md
diff options
context:
space:
mode:
authorEmilio Jesus Gallego Arias2019-03-26 22:09:30 +0100
committerEmilio Jesus Gallego Arias2019-03-26 22:15:30 +0100
commita74d87aca7dd5d09f305be5772b1ee14c926d904 (patch)
tree07cfc71c364b333861a8dd41e65ddd58f57fa444 /dev/doc/debugging.md
parent2ac275c0f3e65a402951de86a61c77dd0e0782f8 (diff)
[kernel] Don't re-declare scheme side-effects that are already there.
This is an experimental PR as I am not sure I can follow the reasoning in e1ba72037191b1d4be9de8a0a8fc1faa24eeb12c Note that in safe_typing we avoid re-declaring effects twice. This removes a huge number of redeclaration of schemes side-effects that in fact are already generated such as `eq_ind`. Anyways we should deprecate the declaration of Schemes on-the-fly, I don't see the point honestly; just make sure your theory has the right ones. Well, going to a more eager declaration scheme could be costly in terms of size, so OMMV. TODO: only declare side-effects if the scheme is generated on-the-fly, add a parameter to the declaration function.
Diffstat (limited to 'dev/doc/debugging.md')
0 files changed, 0 insertions, 0 deletions