help-gnu-emacs
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: What's the differences b/t M-x eshell and M-x shell.


From: Xah Lee
Subject: Re: What's the differences b/t M-x eshell and M-x shell.
Date: Wed, 08 Dec 2010 15:31:22 -0000
User-agent: G2/1.0

On Oct 9, 2:27 pm, Sean McAfee <eef...@gmail.com> wrote:
> p...@informatimago.com (Pascal J. Bourguignon) writes:
>
> > eshell is implemented in emacs lisp, and gives you direct access to
> > emacs lisp, which is a much better language than any shell language (but
> > scsh of course).
>
> It's a pity, then, that eshell's documentation is worse than any other
> shell's, by several orders of magnitude.

yeah... pretty bad.

it's soooo verbose. It assumes people are highschool students.

here's the url:
http://www.gnu.org/software/emacs/manual/html_node/eshell/index.html

it start with a elaborate intro
http://www.gnu.org/software/emacs/manual/html_node/eshell/What-is-Eshell_003f.html#What-is-Eshell_003f
telling people about the concept of a “shell”.

then it goes with the “verbs” concept.

i spend 20 sec trying to see what commands are actually there... my
effort just got to this dead end:
http://www.gnu.org/software/emacs/manual/html_node/eshell/Built_002dins.html#Built_002dins

and, like, 80% of the nodes are all empty.

little to no actual examples. If the doc actually just show 20
examples of the most common use of ls, cd, mkdir, cp, grep, etc
that'll immediately improve the doc some 100 fold.

pretty sad.

But what a great tool it is. The doc issue i wouldn't blame the author
John Wiegley.
Rather, i blame FSF... for various reasons, the culture of FSF... made
it that way.

recently been thinking more about this and scanning FSF dev list
again, just incredible inefficiency. I don't know even where to begin,
and if i just keep typing, within a min i'll probably be tagged a
destructive troll...

to begin, people are forced into using texinfo ... 30 old tech, today,
really nobody knows it, while really tons of alternatives even
counting org and muse mode's markdown/html features within emacs. But
to say texinfo should be phased out is immeditaly a stone wall that
cant go anywhere...

then, the whole gnu emacs manual is written in a elaborate verbose
style, as the mother of examples of how new dev should follow...

forgive me, to say this on a FSF owned forum, but recently i've been
thinking that it'd be a good experiment if Richard Stallman would
withdraw from the emacs dev list. Why not? The venerable Richard has
selected emacs maintainers, Stefan & Yidong. So Ricard, why not leave
the list and let them wholly run the show for a while? maybe with a 1
year agreed time that you come back and monitor it or back to the
helm. Am saying this because with you there, due to your status and
venerability, anything you say, has strong influence on the whole
list. People are afraid to say things or tech opinions if you happened
to have a word on xyz in your email. You (and me too) are getting old,
out of touch. The young devs, those make the world's tech google,
facebook, youtube, github, twitter... are running the show.

when i look at gnu emacs dev list, it's arguments, arguments,
arguments, among its developers, many i've been acquainted over the
years. The arguments are not really the sort that's purely technical
and respectfull, and it's not just between a few devs. It's like all
over, and some with bad words thrown about. Of course on a logical
level probably most still say it's still civil, but it really do hurt
feelings and exacerbate the debates. Tens, hundreds, of words throwing
back and forth. “You!”, “No YOU!” It's pretty ill. (and i thought i
being the troll am the only one having verbal fights with some emacs
developers)

most other open source projects, including many using GPL, on Google
Code or github, with integrated bug db, integrated comment for any
public joe, integrated web-interfaced svn/git, integrated with instant
msg, dev identies/profile, history, email, dev's blog/site/twitter,
all web 2.0 communication tech, instaneous and inter-connected down to
their cellphone, iphone, ipad, blackberry or whatnot, and all
typically so easy to use that one doesn't need to know any unix sys
admin skill to start a coding project and enjoy all these features.

Emacs is still 1990's plain text email list, manually hosted
maintained chron job'd version control system, static website, really
bad bug databe, really bad email archive (plain text search)... as i
can see these are really becoming major drag.

recently, the dev list discussed ibuffer. Reading it is like watching
a circus show. Long time developers can't find old discussions, keep
asking other to find it, then argue, then asking what was decided
before, and it turns out this has been discussed back in 2001... and
still pretty much no decision or no nothing is actually being done.
And this is not just isolated incidence. I read the dev list now and
then since 3 years ago... More than half of the issues almost went
like that. Quite a miracle that emacs 23.x is released and 24 coming
out.

am not sure other open source dev team such as maybe python, ruby,
linux, apache etc are like this... maybe most older tech projects are
like this... but am thinking perhaps Stefan & Yidong can start fresh
and change things a bit to move things faster. Maybe as a experiment
for a year.

this post is quickly typed. All errors and bad thoughts are just me.

 Xah ∑ xahlee.org ☄


reply via email to

[Prev in Thread] Current Thread [Next in Thread]