Re: No automatic tabs in Emacs?

[Elena]

On Dec 6, 10:24 am, Miles Bader <mi...@gnu.org> wrote:
> > Miles, CEDET could have the most advanced project-management system of
> > this world, but if nobody knows or can figure out easily how it works,
> > it doesn't matter.
>
> I'm confused; are you implying that CEDET's project-management is
> somehow hard to use or figure out?  Have you used it?

Yes, I have been burned by it in the past.  I thought I should have
given it another try now, but reading both Alex Ott's tutorial and
some comments around - like yours at the end of your post - convinced
me that things are still not that approachable.

> > That's why I think Emacs can't be an IDE, not even a code-editor,
> > especially for beginners.  If Emacs users say it isn't so, well, I
> > just can say they don't know what an IDE is.  [2]
>
> You can say that, but of course it isn't very convincing -- you're
> simply asserting that anybody who disagrees with you must not understand
> the argument.
>
> [and remember that many Emacs users (including myself) also use IDEs (I
> only dabble with VS -- it's bad enough watching others use it -- but
> I've used Eclipse a lot).]

Well, then it could just be a matter of different experiences.  I
myself use both VS and Emacs, and watch other people use either an IDE
or an editor to accomplish tasks, so I think I'm entitled in sharing
my opinion to help newcomers save time and effort.  If everyone
thought alike, then I guess either Emacs or IDEs would be extinct.

In the past, people watched me develop in Perl and amazed at my
productivity, but behind my smile I *knew* using an IDE would have
been way more productive.  I didn't let the powerful text-editing
capabilities of Emacs fool me into thinking I was using a code editor.

> Maybe you're right, and adding much[*] "IDE type" functionality Emacs is
> doomed due to lack of manpower, but really, it's not really the _goals_
> that are difficult (to understand), it's the implementation.
>
> [*] One must remember, however, that IDEs as we know them are a bundle
> of often rather orthagonal functionality, so it may be somewhat
> misleading to talk about them as if they're a unit.  Some of this
> functionality is easier to implement than other, and it may well be
> possible to get quite far without being perfect.

Could you please expand on this?

My opinion is that GNU Emacs is doomed as an IDE because it tries to
implement all functionality from scratch, for both portability and
licensing reasons.  If GNU Emacs relied on external tools to build the
functionality of an IDE (look at GCCSense, CScope, etc. for C/C++
development), it would be very likely outperform any IDE out there.