So looking at this it doesn't look like there's really a super satisifying path forward. Though it seems @mkitti has dealt with this on the conda side and honestly we should just follow that IMO. I...
Yes, I can take a stab at it this weekend.
> On Oct 21, 2024, at 1:29 PM, Timothy ***@***.***> wrote:
>
>
> Thanks for reporting and investigating this @awadell1 <https://github.com/awade...
This issue has two parts.
# Issue 1
The following NEWS entry is incorrect:
> If no path is given after (i.e. `--project=@script`) then (like `--project=@.`) the directory and its parents are...
I am slightly doubtful the same fix will apply to both issues because in Issue 2 above an absolute path somehow gets introduced. In other words, if Issue 2 did just depend on Issue 1 getting the co...
Profile from while it's doing it. Looks like completion is finding a massive list of `MethodCompletion` and the time is spent doing repr for each of those..
```
dOverhead ╎ [+additional indent] Co...
At a glance, I suspect that Issue 2 might be solved by fixing Issue 1 (if the base path is determined incorrectly, paths relative to it will be incorrect too).
I agree with @nsajko. It's always possible to add more generic functionality later, but it's impossible to go backwards. The most generic `zero` gets is `::Number`. If we have a function in base,...
I'm not particularly excited about publicizing `require_one_based_indexing` — the reason for its existence is entirely about adding safety checks to _old_ code. But I see that the pair reference e...
"Subtly different", but with no explanation anywhere? At least, I can't find anything in `--help` or the manual that `--trace-compile value` is accepted, when `--trace-compile=value` is what is lis...
@tecosaur I agree that Issue 1 above is a dupe of #55069. However, #55069 doesn't mention anything about the `--project=@script/sub_dir` functionality (what I call Issue 2 above). So this issue isn...