Skip site navigation (1)Skip section navigation (2)
Date:      Fri, 2 Jun 2000 18:42:16 +0000 (GMT)
From:      Terry Lambert <tlambert@primenet.com>
To:        tms2@mail.ptd.net (Thomas M. Sommers)
Cc:        freebsd-chat@FreeBSD.ORG
Subject:   Re: Why encourage stupid people to use *BSD WAS:Re: IE
Message-ID:  <200006021842.LAA24897@usr09.primenet.com>
In-Reply-To: <393739CF.2115B00C@mail.ptd.net> from "Thomas M. Sommers" at Jun 02, 2000 12:36:31 AM

next in thread | previous in thread | raw e-mail | index | archive | help
> > > I think the important thing is not to create false expectations (of ease
> > > and simplicity) in the minds of those who have never been exposed to a
> > > real operating system.
> > 
> > I think the important thing is to create TRUE expectations of
> > ease of use and simplicity in the minds of those who have
> > never been exposed to FreeBSD.
> > 
> > Of course that means changing the code to have those as
> > attributes.
> 
> But can that be done without radically changing the nature of the OS? 

Who cares?  The point is to have the most effective human-computer
interaction (HCI) possible.

Only part of this is having the system stay up, for which FreeBSD
has a better track record than Windows, so far.


> Home users and small businesses have been accustomed to a
> single-user OS.  Multi-user OSes are necessarily more complex
> and (in some ways) more limiting than single-user OSes.

This is an implementation detail, and it is based on a false
premise: that what people are currently accustomed to is what
an average human, with no prior experience, would expect the
system to behave.

This premise has led to a hell of a lot of bad design.  See the
"Interface Hall of Shame" at:

	http://www.iarchitect.com/shame/htm

for good examples of people trying to make software running on
computers resemble everyday objects (e.g. a representation of a
"thumb-wheel" on a graphical representation of a device, with
no physical thumb-control mapped to the representation).


> People are not prepared for, and may not put up with, these
> complexities and limitations.

Most of these complexities are artifacts of substantial design
flaws, which should be corrected, instead of glossed over as
"that's the way it works; it's better, trust me".


> For example, people will say: "What do you mean I have to login?
> I didn't have to do that with Windows."

Windows 3.1, perhaps.  Probably it should be called "unlocking",
not "logging in".  Certainly, it should be possible to turn on
a FreeBSD box and just get a graphical desktop or shell prompt
with a particular users credentials as an active default.  It's
the user's choice, not the OS designers.  The "login problem"
is trivial to overcome.

> or "What do you mean I can't undelete a file? I could do that
> with Windows."

Well, the inability to undo an "oops" is a moronic point about
FreeBSD.  Humans have accidents; you must accomodate this fact
about humans, rather than trying to suppress it.  You must
design systems which tolerate faults.

Why is is that people are all happy about hardware fault tolerance,
but steadfastly deny the need for operator fault tolerance?  Are
humans somehow more predictable than solid state physics?


> This is not to say that FreeBSD can't be made simpler, but
> there is a limit, and that limit is more complex than Windows.

I think you are confusing the complexity of the system with the
complexity of abstraction presented to the operator of the system.

A system of differing blocks which can be arranged to completely
fill a box can require very complex ordering constraints to be
able to completely fill the box.  But once closed, all you see
is a three dimensional cube.  How complex is that presentation
of the interior complexity of the system?  Almost everyone can
relate to a simple cube.


					Terry Lambert
					terry@lambert.org
---
Any opinions in this posting are my own and not those of my present
or previous employers.


To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-chat" in the body of the message




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200006021842.LAA24897>