[Sussex] Straw Poll

[Geoff Teale]

Steve wrote:
------------
> Geoff
> 
> It's good to be back in rant mode - hope you're enjoying the
> ride as much as I.

Yeah, I am.  The world has been spinning on a different axis for a while,
for a couple of weeks at least I have the time to talk a bit more. 

> I for one have come to dislike IDEs or any large package that
> tries to be all things to all people.  I find that they often
> fail at the edges, that they limit the ways in which I can 
> solve a problem.  I much prefer lots of little tools that do
> one job, but do it well, that I can plug together using make.

Hmm, I would agree with this.   When I was doing java stuff (we're talking 3
years ago) I found some of the IDE's nice, but ultimately I wanted the
control that doing it all at the command line gave me.

The power of make is that it allows you pull all those little tools
together, but emacs is very similar.  As you well know, modes are e-lisp
programs.  The RMail mode for example is a mail front end written in e-lisp.
What Emacs provides is a framework in which to write front ends - that
interface is principly textual (though it does also support graphics).
Emacs can be an IDE (it could be argued that it was the first IDE), but it
can also be pretty much anything else you want.  Some people even argue that
Emacs is much less a text editor and much more an OS (or more accurately a
virtual machine) in it's own right.

Lets continue with the example of email in emacs.  Emacs RMail/Sendmail is
only a frontend to to sendmail (or compatible) in the same way mutt is.
However, for the emacs user it means that:

1. They don't have to learn a new tool.
2. If they are working at the terminal they can view and send mail in a
single session (yes, I know you could use screen or promote things using fg
and visa versa, but Emacs is simpler!).
3. Again, in the terminal environment it is trivial to copy and paste
between their e-mail environment, their calendar, their text editor, etc.
etc.

Emacs is a configurable front end for many, many tasks, it gives you the
power to do anything you want as a power user (up to and including the
trivial authoring of new functionality), but it doesn't _require_ you to
operate at that level.  It is fair to say that the framework does limit the
ways in which you can approach the solution - but this is also a strength,
and ultimately anything you can do at the command line you can do in emacs
(plus a whole bunch more).

<snip> 

> But isn't this very similar argument to "We'll do this in VB - 
> it'll be easier than learning C++/Java/...".  I grant you that
> WYSIWYG editors are simpler to learn (even use cold), but they
> never offer the same power.  

That is true of VB, but not of Emacs.  Emacs lets you operate at a trivial
level if you so wish and at a more complex level if that is your need.  I
think it is fairly unique in achieving that without one getting in the way
of the other.

>But WYSIWYG should be spelt WYSIAYG:
> What You See is _All_ You Get.  [X]Emacs does try to provide both.
> The menu system provides an intuitive interface while elisp and 
> the command interface is there for real power.  But how many
> people ever get beyond the menu?

I wondered when we'd get to WYSIAYG ;).  You've already made the other side
of the argument, Emacs is not limited in that way, in fact I would say
WYSITTOTI, a rather less catchy acronym meaning "What You See Is The Tip Of
The Iceberg".  

In answer to the question, well, anyone who needs the power finds it, for
the same reason they do in vi.

> By being less intuitive vi forces you to go out and find out more
> about it.  

Only for those stubborn enough to put up with it.  Again, 10, 15 years ago
you'd accept it and fight through, why would you now?  By being less
intuitive it forces people away from it to less powerful editors and they
end up without the functionality they really need.

This is of course where gvim and kvim come in (not vim itself though).  They
provide menus so the average user can get into vim.  Even so, the constant
mode switching is something that most people are going to question
initially.

> Most vi users (and I include myself in this) learnt
> from other people.  Tips are passed on from user to user.  Judging
> from the number of Word documents that I've had to edit where the
> original author has not used section number tags but done section 
> numbering by hand I think intuitiveness is over rated.  

Not at all.  Doing it by hand is ituitive, working out that the feature
exists because it's hidden in some menu is not.  If you open notepad, word,
nano or emacs under X you can (assuming you've seen WIMP environment before)
open a file, do some work on it and save it without ever once consulting a
manual or asking for help - in vi, even if you worked out how to open a
file, how long would it take you to work out that you need to press `i` to
go into insert mode.

>They 
> encourage one never to get beyond the level of novice.

That may be true.  

> Thanks, my big point in this rant thread.  Non-intuitive interface
> give rise to more advanced users.

There is a nice phrase for that arguement:
Post hoc ergo propter hoc.

People who have the character to stick with vi and find that functionality
would do so whether or not you gave them icons and menus.  

The people who don't use the automated features of Word do so because they
don't look for them, not because they're hard to find.   Those people would
walk away from vi well before they ever got to the power (probably before
they ever worked out how to get out of vi!).

-- 
GJT
Free Software, Free Society. 
http://www.fsf.org   http://www.gnu.org



The above information is confidential to the addressee and may be privileged.  Unauthorised access and use is prohibited.
 
Internet communications are not secure and therefore this Company does not accept legal responsibility for the contents of this message.
 
If you are not the intended recipient, any disclosure, copying, distribution or any action taken or omitted to be taken in reliance on it, is prohibited and may be unlawful.
 
Claybrook Computing Limited is a subsidiary of Claybrook Computing (Holdings) Limited
Registered Office: Abbey House. 282 Farnborough Road, Farnborough, Hampshire GU14 7NJ
Registered in England and Wales No 1287205
 
A Hogg Robinson plc company