cider-browse-ns.el (cider-browse-ns--doc-at-point): Fixup

[abo-abo]

• cider-browse-ns.el (cider-browse-ns--doc-at-point): It was a malformed form
  previously.

[abo-abo]

Also, I see one instance of when-let from subr-x has sneaked in place of -when-let from dash. If it's OK to use it, probably all of them need to be updated.

[expez]

I see one instance of when-let from subr-x has sneaked in place of -when-let from dash.

That would be a mistake. IIRC when-let was added in 24.4 or 24.5 and we're still supporting at least 24.3.

Could you update the commit message so it doesn't mention the file? Something like "Fix a malformed call-site" would be adequate.

[bbatsov]

Btw, as this fixes a user-facing bug, it probably warrants a changelog entry as well. You can fix the when-let problem in a separate commit.

[abo-abo]

OK, updated. Btw, what's up with not mentioning the file in the commit message? I think it's a good practice.

[expez]

OK, updated. Btw, what's up with not mentioning the file in the commit message? I think it's a good practice.

I'm sure @bbatsov has an opinion on this too but the biggest problem I had with the previous commit message was that the first line (the summay which appears in git log and all other tools) was essentially useless because it only contained file and function information along with the word "fixup".

[abo-abo]

essentially useless because it only contained file and function information along with the word "fixup"

That's kind of the point. The commit message for such a minor fixup must look benign so that I can safely skip it when browsing the log. On the other hand, "Fix a malformed call-site" is ambiguous and sounds important enough to be tripping through it each time I'm looking for some change.

Of course, that's just my opinion. I hope that you see that there's at least some logic to it:)

[bbatsov]

That's kind of the point. The commit message for such a minor fixup must look benign so that I can safely skip it when browsing the log. On the other hand, "Fix a malformed call-site" is ambiguous and sounds important enough to be tripping through it each time I'm looking for some change.

Well, that's always subjective. :-)

I'm a big believer in meaningful commit messages and clear commit titles. I don't like file mentions in commit messages because file names change and they don't add a ton of value - I'm generally interested in the problem and solution; the location is just details (and is often implied by the nature of the problem).