Re: Emacs terminology (not again!?) [was: Apologia for bzr]

[David Reitter <david.reitter@gmail.com>]

[-- Attachment #1: Type: text/plain, Size: 3075 bytes --]

On Jan 7, 2014, at 3:31 PM, Drew Adams <drew.adams@oracle.com> wrote:

> Concrete suggestions about that might well be helpful.

I had some suggestions or questions back in 2005 or thereabouts, when I was less familiar with Emacs,
and I asked questions on this mailing list.  Perhaps now the time is better (although I have less time).

Many things are vastly improved from what they were back then - better default key bindings, ELPA, copy&paste with other apps working out of the box, and so on.  

At this time, my concrete suggestion I would have is to make semantic, CEDET and etags work out-of-the-box for all major programming paradigms, and make them work without noticeable performance penalty.  Standard IDEs like Netbeans, Eclipse or Xcode give me that.  Emacs 24 has come a long way, but it's not quite there yet.  For instance, I can enable Semantic from the menu, but then I'd try C-M-/ for a completion, next to [win ...] in nsterm.m, and that doesn't give me all the members of the class of "win", in the correct capitalization.  In .c files, I get at least some helpful echo area messages, but scrolling gets sluggish.  Netbeans doesn't do that.   And it just-in-time compiles, marks syntax errors automatically right in the source code, and so on, and so forth.

The second suggestion is better project management, integrated with Semantic.  Some ingredients, like Speedbar and Desktop, are there, but they do not provide an integrated experience. 

Other people will have other suggestions...

>  What you suggest
> is not a bad goal, but the starting point should not be to short-sell
> newbie Emacs users (not suggesting your approach does that; I don't
> know).  They are as bright as past newbies, no doubt.

Indeed, that's not what I meant.  It's just that even intelligent people prefer to spend their cognitive resources on the actual tasks they are employed to do, or the tasks they intend to do, rather than to learn the tool.  We're in a world where software has become obvious to use, or at least very discoverable.

> You
> cannot just pick up Emacs expecting it to do what you are used to,
> without being disappointed or missing the boat.

The idea in Aquamacs is that you can do that, and transition to faster routines as you go along.
It's probably true that the proportion of highly proficient users is smaller, but the overall number of users is greater.

> One has only to look at the questions on a site such as
> StackOverflow, not especially those about Emacs, but generally
> (and apparently about PHP in particular), to see that some users
> expect instant familiarity with a product without any effort - not
> even a cursory look at the doc or a Google search.  The SO site
> leaders are continually lamenting the poor quality of (some)
> questions and askers. 

Yes, programming has become very much a copy&paste (sorry, kill&yank) effort.  For many programming projects, that is sufficient.  You won't get to hack away at Google, Inc., or at Oracle Corp that way, but that's probably OK. 


[-- Attachment #2: smime.p7s --]
[-- Type: application/pkcs7-signature, Size: 4151 bytes --]