Rewrite and implement }karmalink & }karmaunlink #9
Conversation
dsprenkels
force-pushed
the
karmalink
branch
2 times, most recently
from
4ab4baf
to
c44e3cc
Compare
Cargo.toml
Outdated
| [dependencies.chrono] | ||
| version = "0.2" | ||
| [profile.release] | ||
| overflow-checks = true |
There was a problem hiding this comment.
The overhead is negligible, but why?
There was a problem hiding this comment.
Because the overhead is negligible and I hate overflow bugs. :P
Feel free to veto me on this.
There was a problem hiding this comment.
Most logic and storage interaction is in karma.rs including the logic for computing the group from an initial karma, but not the logic for linking karma. That logic depends on each other for correctness though. I think it makes sense to put all logic and storage interaction in karma.rs and only the human interface in handler.rs.
Co-authored-by: Lars Jellema <[email protected]>
Co-authored-by: Lars Jellema <[email protected]>
I will take a look at this when applying the fixes. I'm unsure whether I'll fix this though. In any case I'd like to prevent |
|
Btw. Perhaps I will take a look at that after merging this change. :P |
Changes in this commit:
using the
}karmalinkand}karmaunlinkcommands.This patch rewrites the whole plugin almost completely. The reason that this
was necessary was because some of the previous dependencies changed so much
over the years, that their usage would have to be rewritten from scratch
anyway. In particular in the case of
pegandrustc_serialize, we have muchbetter crate support nowadays, so I replaced them with their modern
substitutes (
nomandserde).Fixes #5
Fixes #8