Лёша Огоньков <lesha.ogonkov@xxxxxxxxx> writes: > On Tue, 27 Oct 2020 at 11:32, Stefan Haller <lists@xxxxxxxxxxxxxxxx> wrote: >> >> On 25.10.20 4:29, Felipe Contreras wrote: >> >> The other hunk (adding compinit) is not so important to me; I suppose it >> >> was not in the original version because most zsh users already have this >> >> in their .zshrc anyway. But it's not wrong, and doesn't hurt to have >> >> here, I guess. >> > >> > If you don't have compinit, then how is the '_git' script being loaded >> > in the first place? >> >> I didn't say it's unnecessary to have it in your .zshrc. I just said >> it's maybe unnecessary to document it here because most zsh users have >> it in their .zshrc already anyway. > > As an inexperienced zsh user it took me ages to understand why the > whole thing is not working. So,... even though it may look to more experienced zsh users that it is unnecessary to document it in this file, in your opinion, it is a good idea to mention "compinit" to help less experienced users? In any case, the patch in question is the only thing in flight that conflicts with Felipe's 29 patch series, and the change to zstyle line is common between both efforts, so it is just between adding the "autoload -Uz compinit && compinit" near the fpath=(...) thing or leaving it out. I'll let those who know zsh to figure it out. I suspect that the resolution would be either to ask Felipe to rebase his on yours, rebase yours on top of Felipe's, or just drop yours (if "autoload" thing is unnecessary as Stefan suspects---I cannot tell what Felipe wants to suggest by his response, between "unless you have compinit you wouldn't be using this script so why write it?" and "your users need to be taught about the need _somehow_, so why not make this the place to do so?"). Thanks.
