Open jpluscplusm opened 2 months ago
but it
Looks like this sentence was left unfinished.
As I mentioned offline.
If we do it for vet
, we need to do it for all commands (otherwise we end up saying "and take a look at vet
to see how to do X"). cue help inputs
is the canonical reference. I'm minded to think we need better examples on cuelang.org that are richer than the cue help
text. Perhaps a single page with examples, using different commands to drive the point home, of the various permutations of inputs. Rather than trying to create a table test of permutations for each command in each help text or a documentation page specific to each command. And then link to this helpful page from cue help inputs
and the related content of any page that mentions cue vet
, cue export
etc.
but it
Looks like this sentence was left unfinished.
I'm not sure what I was going to say, there. I'm going to update this issue to track being the cue vet
reference page that I believe we need, regardless of any changes we make to cue help vet/inputs/etc
.
As I mentioned offline.
If we do it for
vet
, we need to do it for all commands (otherwise we end up saying "and take a look atvet
to see how to do X").cue help inputs
is the canonical reference. I'm minded to think we need better examples on cuelang.org that are richer than thecue help
text. Perhaps a single page with examples, using different commands to drive the point home, of the various permutations of inputs. Rather than trying to create a table test of permutations for each command in each help text or a documentation page specific to each command. And then link to this helpful page fromcue help inputs
and the related content of any page that mentionscue vet
,cue export
etc.
I've tracked this in https://github.com/cue-lang/docs-and-content/issues/144.
In https://cuelang.org/cl/1193603 I attempted to quickly address https://cuelang.slack.com/archives/CLT4FD7KP/p1711532252834579, which was caused by a user not being aware that
cue vet
can resolve constraints from packages, not just individual files.However, if we accept https://cuelang.org/cl/1193603 then we'll also need to expand the help text for other commands.
cue help inputs
already exists, but itWhilst we work out the best approach for the built-in
cue help
texts in this space, on cuelang.org we should publish a primer oncue vet
that, amongst other things, pushes readers towards package-based constraint resolution, not file-based.Given the non-task-specific nature of this primer, I suggest it's best placed inside the reference hierarchy, and should be written with that location in mind.
I propose we:
cli
component of thecue
command's URLs, by moving the contents of the https://cuelang.org/docs/reference/cli/ hierarchy to https://cuelang.org/docs/reference/command/-help-
) to better reflect the actual underlying command whose output they demonstratecue vet
primer at https://cuelang.org/docs/reference/command/cue-vet/This scheme leaves us with a good setup in which to place reference-level primers/etc on other
cue
subcommands in the future - and other top-level commands likecuepls
, as well, if so desired.