@Trott I'd like to eventually get rid of "jslint" in the codebase though. Would you be less opposed if make jslint displayed an error telling the user to use make eslint instead?

Having make eslint and make jslint be aliases of each other could be confusing. Someone might erroneously think that the latter runs JSLint, while the former runs ESLint.

@seishun How about we leave the underlying tool out of the naming entirely? Maybe something like this?:

Given that:

• make test tests (nearly) everything.
• make test-addons and similar variants only test addons, etc.
• make lint lints everything.

Perhaps instead of make jslint and make cpplint, we should reformat like the test-* stuff?:

• make lint-js will do what make jslint currently does.
• make lint-cpp will do what make cpplint currently does.

We can have a period of time where jslint is an undocumented alias for lint-js. After a period of time, jslint emits a warning pointing at the new name but is still an effective alias. At some point after that, it emits the warning and does nothing else. And maybe it stays that way forever or maybe we remove the alias at some point after that.

@Trott I'm open to that, although I'm not totally comfortable with inventing a generic term for linting JS. Let's see what other @nodejs/collaborators think.

We can have a period of time where jslint is an undocumented alias for lint-js. After a period of time, jslint emits a warning pointing at the new name but is still an effective alias.

What harm is there in introducing a warning right away?

I think keeping 'jslint' is best, especially if we decide to switch to another tool in the future (so we don't have to change the name everywhere again at that point).

• make lint-js will do what make jslint currently does.
• make lint-cpp will do what make cpplint currently does.

I'd rather we do this, it's not tool specific and it fits the rest of the makefile. I've also been confused by the use of jslint in our codebase.

What harm is there in introducing a warning right away?

Mostly people complaining. :-D There will undoubtedly be those who feel that the current task name should work indefinitely. If you're up for having that conversation sooner rather than later, then go for it.

My inclination is usually to make the smallest changes reasonable, one at a time. But that isn't always the best thing to do, so...
¯\(ツ)/¯

"jslint" is the name of a different tool, which can cause confusion.

Has it actually ever though? Has anyone tried to run the jslint tool on Node files because their make jslint failed?

@benjamingr That's what the PR description seems to describe.

I'm in favor of what @Trott is suggesting here. Make js and css postfixes. It fits the style, and it removes all ambiguity.

@benjamingr when I set up my repo to open my first Node.js PRs I was very confused about the name.

I like make lint-js a lot as it does not comply to any library and I think that is a good way to make the live of newcomers a tad easier.

I was also confused by the name at one point, although I wouldn't say my confusion actually caused any issues (since I never needed to know what the file was doing anyway).

I am inclined to what @mscdex has said.

I think keeping 'jslint' is best, especially if we decide to switch to another tool in the future (so we don't have to change the name everywhere again at that point).

@mscdex @thefourtheye Given the reason, would lint-js and lint-cpp be acceptable? That would be analogous to test-addons etc.

@Trott Those names themselves are fine, but I just don't see a need to change anything.

I am rather for keeping it. Let alone to avoid the churn and bikeshedding. Also don't feel there is a need.

I'm also in favor of keeping the existing names.

As suggested by @Trott, replaced eslint with lint-js and cpplint with lint-cpp. The old names still work, but print warnings.

This makes it more consistent with other task names such as test-addons and prevents confusion, so why not?

One thing I'm not sure about is vcbuild.bat output when running lint tasks. I wonder if "running eslint"/"running cpplint" would be preferable to "running lint-js"/"running lint-cpp".

One thing I'm not sure about is vcbuild.bat output when running lint tasks. I wonder if "running eslint"/"running cpplint" would be preferable to "running lint-js"/"running lint-cpp".

So for lint-cpp the invocation trough call :run-python tools/lint-cpp.py %cppfilelist% will also output the actual CLI invocation of python, so the output is:

running lint-cpp 'src\*.c src\*.cc src\*.h test\addons\*.cc test\addons\*.h test\cctest\*.cc test\cctest\*.h test\gc\binding.cc tools\icu\*.cc tools\icu\*.h'
"C:\bin\dev\python27\python.exe" tools/lint-cpp.py  src\async-wrap.cc src\backtrace_posix.cc src\backtrace_win32 ...

Which IMHO is nice and useful
(tl;dr keep echo prompt as is, fix lint-js per my comments)

Which IMHO is nice and useful

Sorry, I messed up. The filename is actually cpplint.py, and it shouldn't be renamed because it's actually the name of the tool.

I am a bit surprised that there are a couple of persons saying that there is no need for changing something when there were four people mentioning that they were confused about the name.
Especially when the old name is kept as a fallback as is it is right now. The confusion can lead to some unnecessary overhead for new people.

Does anyone of you @mscdex @cjihrig @thefourtheye @eljefedelrodeodeljefe have a strong opinion about this? I do not see any negative side as I do not see the churn as a real negative point here.

No strong opinion, but ideally I would want the understanding that churn and bikeshedding is distraction from doing valuable things. Not only the authors' but also those of multiple reviewers. And hence should be discouraged.

See, I learned to be conservative about the codebase since it's a very serious business running it in production.

Just think of this: what is the consequence in closing this now. Nothing really. Go ahead as you deem it pls.

CI: https://ci.nodejs.org/job/node-test-commit/12361/

So what is the conclusion here?

AFAICT we have three +1 and two -0. Unless someone wants to explicitly -1, this could land.

I think there was enough time to express a clear -1 and this has not happened, so it should be fine to land this.

@seishun it needs a rebase though.

Running CI one last time since there was a rebase: https://ci.nodejs.org/job/node-test-pull-request/10217/

Landed in b2eb987

Should this be backported to v6.x-staging? If yes please follow the guide and raise a backport PR, if not let me know or add the dont-land-on label.

If it is backported I think it would make sense to remove the warning when calling make jslint, as that could be considered a behavior change in CI systems.