Inline parameter name hints in front of literal values

[balta2ar]

I wonder whether it's possible to implement "parameter hints" as it's implemented in IntelliJ's IDEs.

[brettcannon]

Closing as a duplicate of #70

[balta2ar]

@brettcannon This is not a duplicate. I believe this issue and #70 have different scopes. What I'm talking about is adding those small visual greyed out keywords (result, text, key, number). This request is not about displaying something in a tooltip.

[balta2ar]

Hm, this may not be possible before microsoft/vscode#16221 is implemented.

[balta2ar]

Hm, or maybe it is possible already: jacobdufault/cquery#85 (comment)

This can be done in vscode by using decorators (similar to how semantic highlighting works).

[balta2ar]

FTR: here is a relevant example that demonstrates how to use decorators: https://github.com/Microsoft/vscode-extension-samples/tree/master/decorator-sample

[brettcannon]

Thanks for the suggestion, @balta2ar , but we think this takes the extension too far into IDE territory for us to implement.

[balta2ar]

Well, let me disagree. The whole idea of the python plugin (to my opinion) is to make Python developer's experience better. Otherwise we could've stayed with plain vim with no extensions.

And it IS one of those things that is helpful to users, that makes their experience better. Look at the issue in vscode: microsoft/vscode#16221, it's got so many upvotes, users want it. If this is too far into IDE, then how is this different from "Peek definition" feature? They both save time on jumping around, they provide context.

I understand the resources of the team are limited but at least keeping the issue open could make it more discoverable for potential contributors.

[brettcannon]

We closed this specifically because we wouldn't accept a PR implementing this feature (otherwise it would have been left open and we would have added the 'help wanted' label). I realize you would like to see this and that some others might also find it useful, but we just don't think the cost of the feature and how it fits into the general design we are going for warrants the cost of implementing it. (Obviously people can free to implement this as a separate extension since there is nothing in our extension that would prevent it from working.)

As for the upvotes on the VS Code issue, there are 28, but that is actually rather small for the VS Code repo (e.g. it doesn't show up until page 6 when you sort by 👍 votes). And the fact this issue has been open for almost five months on our issue tracker and has not received any upvotes from another person means we are not seeing a high enough demand to reconsider our decision.

So I'm sorry but we won't be re-opening this.