Update the non-toolchain examples to bzlmod #684
Conversation
z8v
force-pushed
the
update-other-examples-bzlmod
branch
6 times, most recently
from
a8ce95a to
164b2a0
Compare
z8v
force-pushed
the
update-other-examples-bzlmod
branch
from
164b2a0 to
eb822fa
Compare
layus
left a comment
layus
left a comment
There was a problem hiding this comment.
Such an impressive rework. Not much code, but so many subtle edits. Thanks !
I am a bit concerned about the use of deprecated constructs, and the reason why we diverge from common conventions. Not a big worry, but an explanation in some readme would help, or next to the changes.
A bigger issue is that we impose all our users or rules_nixpkgs_cc to configure their own module extension. Is that by design ? Or a shortcut taken in these examples ?
| # Toolchains | ||
| # | ||
|
|
||
| cc_configure = use_extension("//:extension.bzl", "cc_configure") |
There was a problem hiding this comment.
Do we absolutely have to define our own module extension to use cc toolchains ?
I feel like that extensions pertains to rules_nixpkgs_cc itself, with tags for the config. But maybe it is not implemented yet ?
| import %workspace%/../../.bazelrc.remote-cache | ||
|
|
||
| build --host_platform=@rules_nixpkgs_core//platforms:host | ||
| build --crosstool_top=@nixpkgs_config_cc//:toolchain |
There was a problem hiding this comment.
Having crosstool_top used seems weird. Should it not be selected by default by the toolchain selection engine ?
Or is this intended as a demo on how to force a specific toolchain ?
I feel like crosstool_top is on the verge of deprecation.
There was a problem hiding this comment.
Is this intended ? Do we want this version to diverge from ../../.bazeliskrc ?
| @@ -1,2 +1,3 @@ | |||
| bazel-* | |||
| result | |||
| MODULE.bazel.lock | |||
There was a problem hiding this comment.
Again, why is this repo different ? Is this related to python 7 and any run altering that lockfile ?
Uh oh!
There was an error while loading. Please reload this page.