I approved DIP1036e
Meta
jared771 at gmail.com
Sun Jan 21 19:20:21 UTC 2024
On Sunday, 21 January 2024 at 17:37:41 UTC, Richard (Rikki)
Andrew Cattermole wrote:
> On 22/01/2024 1:35 AM, Dom DiSc wrote:
>> Sadly, because it's easier bikeshed, especially when the
>> reviewer
>> doesn't actually understand the code, most code reviews
>> turn into
>> nit-picks focusing on sub-optimal but ultimately
>> insignificant
>> deviations from the reviewers stylistic preferences, all
>> in the name
>> of "legibility". The single best way to derail a
>> conversation on
>> code reviews is to bring up the topic of ... line length.
>>
>> Oh please. Everybody submitting a PR can easily run dfmt on
>> it. This done there should be no more "too long lines",
>> "multiple expressions in one line", "not correct indented
>> code", "curly braces not on their own line", etc. pp. If
>> someone submit PRs without running dfmt, (s)he should be
>> advised to do so. End of discussion.
>
> Yes, that has been added to the PhobosV3 proposal by Adam.
>
> I also suggested that we automate it as part of the CI. Because
> who has time to care about white space at the end of a line? No
> thanks.
Yes, yes, and yes. At work we have a linter that IS integrated
into the CI pipeline, and a formatter that is NOT, for some
unfathomable reason. Every time one of my PRs fails the CI
pipeline because I forgot to run the formatting tool before
pushing, I want to scream. Just automate it and a whole class of
PR nitpicks/arguments disappears, and everyone is better off.
More information about the Digitalmars-d
mailing list