console: Add common aliases for console methods

[aks-]

• add common aliases as per following spec
• https://github.com/DeveloperToolsWG/console-object/blob/master/api.md

[silverwind]

I suggest using the same signature as in the spec: console.debug(object [,object, ...]).

[aks-]

Then we might want to change the complete doc for console. Because the signature to methods like log are also not matching with spec we are following https://github.com/nodejs/node/blob/master/doc/api/console.markdown#consolelogdata- Do you think I should change them as well?

[silverwind]

Yes, I think changing the signatures to match the spec would be preferred. Can you do so in a second commit?

[silverwind]

Please also add some tests for these methods here.

[jasnell]

Wouldn't this need to be semver-major as it could break anyone who has monkey patched these? (same issue as nooping the undefined ones)

[silverwind]

I don't think we have to support broken polyfills or monkey patches, and I think it's unlikely anyone would do so in the case of these methods.

[aredridel]

Hm. on the converse side, this lets things designed for the browser run as is, even if they use these features.

[chrisdickinson]

@jasnell:

Wouldn't this need to be semver-major as it could break anyone who has monkey patched these? (same issue as nooping the undefined ones)

It seems low risk. If code is monkeypatching the console object, that code's new debug/exception methods should (continue to) win. I might be missing something though — do you have a link to the problems arising from noop'ing handy?

[aks-]

@silverwind I have added the tests for this and changed the signture of newly added aliases as per browser spec. Besides that I created a new PR #3584 in which I changed the method signatures which were there previously in doc to follow the browser spec. I haven't changed the signatures of methods which have differnent behaviour in node than in browser (e.g. console.assert). Now I am in confusion if we are actually following the spec?

[silverwind]

@aks- thanks, I'll have a look later.

This PR is LGTM.

[silverwind]

@aks- on second look, what's up with tools/v8-prof/tick-processor-tmp-1221 in this patch?

[aks-]

@silverwind sorry I added that by accident I guess...removed it now

[silverwind]

CI: https://ci.nodejs.org/job/node-test-pull-request/665/

[silverwind]

LGTM, will likely land this tomorrow.

[Fishrock123]

I'm not sure this is a good idea actually if we aren't going to follow the rest of the "spec".

@aks- Sorry, my fault for pulling you into this console-spec thing. :/

[silverwind]

@Fishrock123 I think we should implement as much of this "spec" as is feasible. What's wrong with these alias methods?

[Fishrock123]

Nothing particularly but it's more things we need to keep supporting. Seems low-value -- stuff works fine if it follows this part but not another part kind of thing.

[aks-]

@Fishrock123 it's okay, I am actually learning a lot about the structure of node codebase :)

[silverwind]

Guess I'll put this on hold then.

Interestingly, MDN says debug is deprecated and it lists exception only as _exception. I think some more research into browser behaviour is needed.

[Fishrock123]

See #3584 (comment), the situation is generally very silly.

[jasnell]

Closing due to lack of activity. Seems like there's very little interest in this in general.