Init.Client: don't use both Core_kernel and Core #17284
Conversation
|
!ci-build-me |
|
!ci-bypass-changelog |
| Format.eprintf "A state hash should not be given for %s@." | ||
| ledger_kind ; | ||
| Core_kernel.exit 1 ) | ||
| Core.exit 1 ) |
There was a problem hiding this comment.
I would simply use Stdlib.
There was a problem hiding this comment.
That's what I did on iteration 1, but for consistency I didn't refer to Stdlib, but only Core
| (Yojson.Safe.pretty_print ?std:None) | ||
| (Error_json.error_to_yojson err) ; | ||
| Core_kernel.exit 1 ) ) | ||
| Core.exit 1 ) ) |
src/app/cli/src/init/client.ml
Outdated
| Core.print_endline signature_kind_string ) | ||
| print_endline signature_kind_string ) |
There was a problem hiding this comment.
The reason why CI broke is that this is inside a Command.basic and not a Command.async, so I'm pretty sure the async scheduler is not active here. The existing Core.print_endline just prints to stdout, so it works. This print_endline is the async one and tries to add the write to the async queue, so nothing ends up being printed - I guess just silently does nothing when the async scheduler isn't active?
There was a problem hiding this comment.
(The failing test was the one that makes sure that mina advanced print-signature-kind still works).
There was a problem hiding this comment.
I can build with this though
There was a problem hiding this comment.
But anyway it's bad practice to call sync printline in async code. Let me just make that async
glyh
changed the title
|
Rebased this atop compatible: now it's taking forever to load nix on this branch |
glyh
force-pushed
the
corvo/clean-up-client
branch
from
161fdad to
7ff5c42
Compare
glyh
added
the
enhancement
|
!ci-build-me |
EDIT: didn't expect this to break CI