Re: [Matlab-emacs-discuss] [matlab-emacs:src] New commit [93a747] by Eric Ludlam

[Eric L. <eri...@gm...>]

On 12/1/20 3:55 AM, Uwe Brauer wrote:
>>>> "MiEGr" == MATLAB in Emacs Git repository <no...@sr...> writes:
>     > ## Branch: usage1
>
>     > matlab-complete.el:
>
>     > (matlab-complete-symbol): Check if matlab-shell was loaded before
>     > checking if the shell is active.
>
>     > By Eric Ludlam on 11/30/2020 18:59
>     > [**View Changes**](https://sourceforge.net/p/matlab-emacs/src/ci/93a74784861ccaa390274ad51e32f56ce824b988/)
>
>
> I just tried out the latest commit, now after the shell is loaded, in a
> .m file matlab-complete-symbol, works now as expected. Thanks!
>
> However I don't understand the default keybinding
>
>      (define-key km "\M-\t" 'matlab-complete-symbol)
>
> That is a quite old binding, so on modern PC that translates to escape
> tab, at least on my Thinkpad X1, which is a bit cumbersome to use.
> Should we change that to alt tab?
>
> Any opinions?
On newer Emacsen M-TAB is already bound to complete-symbol or variant 
that uses 'completion-at-point-functions' configuration variable.

I think the right solution is to remove `matlab-complete-symbol', and to 
augment completion-at-point-functions so that users who customize their 
completion engine (such as by using company mode) will get that effect 
for MATLAB buffers also.  I'm not sure how big a change that is though, 
which is why I was thinking closing out some open stuff and making a new 
branch for that experiment would be worth while.

I'd also like to better understand Karthik's patch which I haven't had a 
chance to look into yet.  I haven't used the xref system enough to know 
how it works yet.

Eric