Re: Standardizing more key bindings?

[Nikolay Kudryavtsev]

Yeah, having such protocols would result in having to struggle to define 
the protocol for each case to be just robust enough. Well if Helm has 
some completely unique command, we can't really standardize it, since we 
need to classify it first. In the end it's a (hypothetical)tool for 
keybinding standardization with generic customization being the bonus.

As for being able to keytheme Emacs in a more intentional(declarative?) 
manner, I agree that it would have been great, I just don't see any 
workable solution to accomplish that. Lets say in my theme package I 
want some utility menu for commands in the current major mode, let's say 
on C-c C-u. Then I'll have to defer to mode developers to classify their 
commands somehow. Lets say elisp-mode has byte-compile-file command 
classified as utility. Then my declarative bindings go through all 
commands classified as utility and byte-compile-file gets automatically 
bound to C-c C-u C-a for example. And we're stuck with pretty much the 
same problem as with the keybinding-protocols I described only in a more 
freeform fashion.

--
Best Regards,
Nikolay Kudryavtsev