Philip Kaludercic writes: > writes: > >> Thanks, now done. > > Great, it builds now. > > Looking at the file, there seem to be a few checkdoc issues that could > be easily fixed. I guess that would be nice to see. Along these lines it > might also be cool to mark the Commentary sections "Instillation" (that > could be updated when the package is added), "Configuration", ... as > subsections of Commentary by having them begin with four > semicolons. That way when using outline-minor-mode, you'll see all of > them when the section is folded. Thanks for suggestions, now done. > > A general question I have is if this could be generalized to a general > completion system? You of course don't have to, I'm asking more out of > personal curiosity. It already is, enabling 'capf-autosuggest-mode' in an ordinary buffer displays an auto-suggestion for the first completion at point candidate as you type. By default, it is only displayed if point is at EOL. However, I don't advertise this feature that much mostly because I find it a bit intrusive. For more information, including how to lift the EOL restriction, you can read the ";;;; Details" section. History auto-suggestion in comint and eshell is actually just a special case of this general mechanism for displaying the first completion at point candidate. In comint and eshell, point is on the last line most of the time which means that the auto-suggestions aren't obtrusive, in fact, they are quite helpful. That is why this package mostly advertises this sub-feature. > > Otherwise, I can add it the package to ELPA with autosync'ing, unless > there are any objections. Okay, thanks. > > -- > Philip Kaludercic