Make the 'q' map actually :quit

[benknoble]

:cclose and :lclose will refuse to close the quickfix window if it's the last window. :q doesn't care.

I'll also note that I discovered this because Ack was overriding maps I had set up in ~/.vim/after/ftplugin/qf.vim, which doesn't seem like a good idea. I think we could fix that with a hasmapto check or the <unique> modifier. Please let me know if I should open an issue for this.

[ches]

This seems reasonable, although I imagine there will be some people who will prefer that it does not quit Vim so easily… I'd definitely not like to solve that by adding a configuration option, I'd rather it be doable easily by normal means of Vim mappings customization as was your first instinct.

I agree that there are a host of problems with the way the quickfix mappings are handled, in fact see #170 for my sense that they shouldn't be a feature of this plugin at all anymore. If the Vim plugin ecosystem and plugin managers had common version number conventions to make a major release in Semantic Versioning terms, I'd probably remove it now.

How about we go one of two directions:

• If your problem would be solved by a PR like you suggest to prevent clobbering a user's existing quickfix mappings (without forcing them to redundantly change g:ack_mappings), let's do that instead of this change to use :q. That'd be a general improvement with little argument against it, without any behavior change that some may not prefer.
• Rip off the band-aid and kill this feature of ack.vim. If you are content using your own customizations through after/ftplugin/qf.vim, I'll consider that plus the number of users on other threads who were happy to learn about QFEnter or vim-qf to be enough votes. I'll prepare loud documentation of the decision and the suggested alternatives for when users come to GitHub with pitchforks after they next run their plugin manager's update 😇

Apologies for the slow response to this issue and any others—I've been the last standing maintainer but I've not had time to give attention to ack.vim for awhile.

[benknoble]

@ches Thanks for the detailed and considerate response.

I'll investigate using hasmapto and <unique> and let you know. If it seems un-viable, then option number two may be the right move; I'll have to dig into the code a bit more (it's not completely fresh in my head).

I'll note post-investigation that some of the mappings seem like feature-creep (heck, I never use them), so I agree in spirit with your comments in #170 and above—it would be a pitchfork moment, though.

[benknoble]

@ches Are these PRs dead in the water? Is help needed?

[benknoble]

I'm closing this because #260 was supposed to supersede it, but I also don't use the plugin anymore.