Require let or const instead of var (no-var) #633

albinekb · 2016-09-19T14:15:08Z

Referring to this rule.

Standard should read node version from the engines (more information) property in package.json, if the version supports const, no-var should be added to the list of rules.

Any thoughts?
Is there any reason to use var in node versions that support const and let?

The text was updated successfully, but these errors were encountered:

dcousens · 2016-09-19T14:20:05Z

Would probably break wayyy to much. Although, I agree with the rule fundamentally.
Might be worth linking in #523

albinekb · 2016-09-19T14:40:16Z

Maybe....

Will standard never release a breaking version, where developers need to fix their code? 😏

I mean, there must be room to add rules that break some codebases to make standard stricter (and thus better)?

dcousens · 2016-09-19T14:46:45Z

#382 might be what you're after, but in the end, I write code for environments where let isn't defined, so this would stop me from using standard since my local node version is >6.0.0.

albinekb · 2016-09-19T14:58:07Z

Cool!
Yes but if your code supports environments where let isn't defined, you don't have engines in your package.json set to node 6? 💭

See https://docs.npmjs.com/files/package.json#engines

dcousens · 2016-09-20T00:23:49Z

Isn't engines only enforced if the user specifies engines-strict? Which is default off...

feross · 2016-09-20T05:20:17Z

@dcousens I believe that npm will still print a warning in that case.

Anyway, I don't think that magical behavior like changing rules based on engines field is the way to go.

Once let works in more environments (like 90% of browsers on CanIUse I think we can consider enabling it. There's definitely an argument to be made that var serves no purpose once let is pervasive. Also, this rule is automatically fixable.

dcousens · 2016-09-20T06:02:47Z

Agreed @feross, I'm just saying I don't think that time is now. Hopefully soon though :)

feross · 2016-09-26T20:42:53Z

Cool -- I'll close this issue for now since it's not actionable.

ForsakenHarmony · 2017-10-13T21:37:52Z

currently at 83%

redbmk · 2018-04-04T21:18:21Z

Up to 91.4% now

LinusU · 2018-04-04T21:35:39Z

@feross is it time? 😎

feross · 2018-05-15T11:22:24Z

Yep, I think it's time 👍

harrysarson · 2019-03-09T19:20:11Z

May I inquire about the status of this one?

LinusU · 2019-03-11T12:27:49Z

This is something that we want to get in, if you want to work on this I would start with compiling a list of all the repos that would break if we change this. If it's not too many, we can open a PR against eslint-config-standard and start sending PRs to upstream projects (e.g. as in this one)

You can get the list by cloning this repo, adding the rule to eslintrc.json, and running the tests.

feross · 2019-07-29T00:10:58Z

Anyone want to help start the process of opening PRs against repos that would break? I want to get this into standard v14

sonicdoe · 2019-08-02T06:50:02Z

I’d like to help with this 👍 I’ve opened standard/eslint-config-standard#152 to add no-var to the ESLint config and included a list of all failing repositories in the description.

feross · 2019-08-02T06:51:00Z

@sonicdoe Thanks for taking on this task! 💪

This is added to this project as part of roll-out across GOV.UK of stylelint to replace scss_lint-govuk as the default linter. This project previously didn't have any JS linting installed so Standard is installed as per the GDS Way [1]. For SCSS linting stylelint-config-gds [2] is used which is a GDS standard based upon stylelint [3]. Both of these projects are then added to dependabot to have their versions maintained with the same approach we use for rubocop-govuk. Standardx is used instead of standard because Standard 16 introduced a rule that disallowed the use of `var`, instead preferring `let` or `const` [4]. This conflicts with the GOV.UK approach where we tend to not embrace features that we know will break old browsers even if they're not necessarily supported [5], disallowing var will mean that < IE 11 will be unable to run any of the JS. In order to customise standard rules this project has switched to using standardx [6] which allows us to disallow rules. I've used this so we can disallow the 'no-var' rule. [1]: https://gds-way.cloudapps.digital/manuals/programming-languages/js.html#linting [2]: https://github.com/alphagov/stylelint-config-gds [3]: https://stylelint.io/ [4]: standard/standard#633 [5]: alphagov/govuk_publishing_components#1611 (comment) [6]: https://github.com/standard/standardx

This applies standard for linting JavaScript as per the GDS Way [1]. Standardx is used instead of standard because Standard 16 introduced a rule that disallowed the use of `var`, instead preferring `let` or `const` [2]. This conflicts with the GOV.UK approach where we tend to not embrace features that we know will break old browsers even if they're not necessarily supported [3], disallowing var will mean that < IE 11 will be unable to run any of the JS. In order to customise standard rules this project has switched to using standardx [4] which allows us to disallow rules. I've used this so we can disallow the 'no-var' rule. [1]: https://gds-way.cloudapps.digital/manuals/programming-languages/js.html#linting [2]: standard/standard#633 [3]: alphagov/govuk_publishing_components#1611 (comment) [4]: https://github.com/standard/standardx

This is added to this project as part of roll-out across GOV.UK of stylelint to replace scss_lint-govuk as the default linter. As this project has very little JS and Sass I did um and ah about whether to add this but it felt best to be consistent. This project previously didn't have any JS linting installed so Standard is installed as per the GDS Way [1]. For SCSS linting stylelint-config-gds [2] is used which is a GDS standard based upon stylelint [3]. Both of these projects are then added to dependabot to have their versions maintained with the same approach we use for rubocop-govuk. Standardx is used instead of standard because Standard 16 introduced a rule that disallowed the use of `var`, instead preferring `let` or `const` [4]. This conflicts with the GOV.UK approach where we tend to not embrace features that we know will break old browsers even if they're not necessarily supported [5], disallowing var will mean that < IE 11 will be unable to run any of the JS. In order to customise standard rules this project has switched to using standardx [6] which allows us to disallow rules. I've used this so we can disallow the 'no-var' rule. [1]: https://gds-way.cloudapps.digital/manuals/programming-languages/js.html#linting [2]: https://github.com/alphagov/stylelint-config-gds [3]: https://stylelint.io/ [4]: standard/standard#633 [5]: alphagov/govuk_publishing_components#1611 (comment) [6]: https://github.com/standard/standardx

This applies standard for linting JavaScript as per the GDS Way [1]. Standardx is used instead of standard because Standard 16 introduced a rule that disallowed the use of `var`, instead preferring `let` or `const` [2]. This conflicts with the GOV.UK approach where we tend to not embrace features that we know will break old browsers even if they're not necessarily supported [3], disallowing var will mean that < IE 11 will be unable to run any of the JS. In order to customise standard rules this project has switched to using standardx [4] which allows us to disallow rules. I've used this so we can disallow the 'no-var' rule. [1]: https://gds-way.cloudapps.digital/manuals/programming-languages/js.html#linting [2]: standard/standard#633 [3]: alphagov/govuk_publishing_components#1611 (comment) [4]: https://github.com/standard/standardx

This is added to this project as part of roll-out across GOV.UK of stylelint to replace scss_lint-govuk as the default linter. This project previously didn't have any JS linting installed so Standard is installed as per the GDS Way [1]. For SCSS linting stylelint-config-gds [2] is used which is a GDS standard based upon stylelint [3]. Both of these projects are then added to dependabot to have their versions maintained with the same approach we use for rubocop-govuk. Standardx is used instead of standard because Standard 16 introduced a rule that disallowed the use of `var`, instead preferring `let` or `const` [4]. This conflicts with the GOV.UK approach where we tend to not embrace features that we know will break old browsers even if they're not necessarily supported [5], disallowing var will mean that < IE 11 will be unable to run any of the JS. In order to customise standard rules this project has switched to using standardx [6] which allows us to disallow rules. I've used this so we can disallow the 'no-var' rule. [1]: https://gds-way.cloudapps.digital/manuals/programming-languages/js.html#linting [2]: https://github.com/alphagov/stylelint-config-gds [3]: https://stylelint.io/ [4]: standard/standard#633 [5]: alphagov/govuk_publishing_components#1611 (comment) [6]: https://github.com/standard/standardx

This is added to this project as part of roll-out across GOV.UK of stylelint to replace scss_lint-govuk as the default linter. This project previously didn't have any JS linting installed so Standard is installed as per the GDS Way [1]. For SCSS linting stylelint-config-gds [2] is used which is a GDS standard based upon stylelint [3]. Both of these projects are then added to dependabot to have their versions maintained with the same approach we use for rubocop-govuk. Standardx is used instead of standard because Standard 16 introduced a rule that disallowed the use of `var`, instead preferring `let` or `const` [4]. This conflicts with the GOV.UK approach where we tend to not embrace features that we know will break old browsers even if they're not necessarily supported [5], disallowing var will mean that < IE 11 will be unable to run any of the JS. In order to customise standard rules this project has switched to using standardx [6] which allows us to disallow rules. I've used this so we can disallow the 'no-var' rule. [1]: https://gds-way.cloudapps.digital/manuals/programming-languages/js.html#linting [2]: https://github.com/alphagov/stylelint-config-gds [3]: https://stylelint.io/ [4]: standard/standard#633 [5]: alphagov/govuk_publishing_components#1611 (comment) [6]: https://github.com/standard/standardx ---

This is added to this project as part of roll-out across GOV.UK of stylelint to replace scss_lint-govuk as the default linter. This project previously didn't have any JS linting installed so Standard is installed as per the GDS Way [1]. For SCSS linting stylelint-config-gds [2] is used which is a GDS standard based upon stylelint [3]. Both of these projects are then added to dependabot to have their versions maintained with the same approach we use for rubocop-govuk. Standardx is used instead of standard because Standard 16 introduced a rule that disallowed the use of `var`, instead preferring `let` or `const` [4]. This conflicts with the GOV.UK approach where we tend to not embrace features that we know will break old browsers even if they're not necessarily supported [5], disallowing var will mean that < IE 11 will be unable to run any of the JS. In order to customise standard rules this project has switched to using standardx [6] which allows us to disallow rules. I've used this so we can disallow the 'no-var' rule. [1]: https://gds-way.cloudapps.digital/manuals/programming-languages/js.html#linting [2]: https://github.com/alphagov/stylelint-config-gds [3]: https://stylelint.io/ [4]: standard/standard#633 [5]: alphagov/govuk_publishing_components#1611 (comment) [6]: https://github.com/standard/standardx

feross closed this as completed Sep 26, 2016

feross reopened this May 15, 2018

feross added the enhancement label May 15, 2018

feross added this to the standard v12 milestone May 15, 2018

feross modified the milestones: standard v12, standard v13 Aug 28, 2018

jbottigliero mentioned this issue Mar 30, 2019

chore: Updates var usage to let conventional-changelog/standard-version#321

Merged

feross modified the milestones: standard v13, standard v14 Jul 5, 2019

sonicdoe mentioned this issue Aug 2, 2019

Enable no-var rule standard/eslint-config-standard#152

Closed

feross modified the milestones: standard 14, standard 15 Aug 15, 2019

christianbundy mentioned this issue Sep 5, 2019

Disallow statements starting with continuation characters #1402

Open

dependabot bot mentioned this issue Mar 15, 2021

Bump standard from 14.3.1 to 16.0.3 Crown-Commercial-Service/digitalmarketplace-supplier-frontend#1379

Merged

github-actions bot locked as resolved and limited conversation to collaborators Nov 18, 2021

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Require let or const instead of var (no-var) #633

Require let or const instead of var (no-var) #633

albinekb commented Sep 19, 2016 •

edited

dcousens commented Sep 19, 2016

albinekb commented Sep 19, 2016 •

edited

dcousens commented Sep 19, 2016 •

edited

albinekb commented Sep 19, 2016 •

edited

dcousens commented Sep 20, 2016

feross commented Sep 20, 2016

dcousens commented Sep 20, 2016 •

edited

feross commented Sep 26, 2016

ForsakenHarmony commented Oct 13, 2017

redbmk commented Apr 4, 2018

LinusU commented Apr 4, 2018

feross commented May 15, 2018

harrysarson commented Mar 9, 2019

LinusU commented Mar 11, 2019

feross commented Jul 29, 2019

sonicdoe commented Aug 2, 2019

feross commented Aug 2, 2019

Require let or const instead of var (no-var) #633

Require let or const instead of var (no-var) #633

Comments

albinekb commented Sep 19, 2016 • edited

dcousens commented Sep 19, 2016

albinekb commented Sep 19, 2016 • edited

dcousens commented Sep 19, 2016 • edited

albinekb commented Sep 19, 2016 • edited

dcousens commented Sep 20, 2016

feross commented Sep 20, 2016

dcousens commented Sep 20, 2016 • edited

feross commented Sep 26, 2016

ForsakenHarmony commented Oct 13, 2017

redbmk commented Apr 4, 2018

LinusU commented Apr 4, 2018

feross commented May 15, 2018

harrysarson commented Mar 9, 2019

LinusU commented Mar 11, 2019

feross commented Jul 29, 2019

sonicdoe commented Aug 2, 2019

feross commented Aug 2, 2019

albinekb commented Sep 19, 2016 •

edited

albinekb commented Sep 19, 2016 •

edited

dcousens commented Sep 19, 2016 •

edited

albinekb commented Sep 19, 2016 •

edited

dcousens commented Sep 20, 2016 •

edited