comp.unix.user-friendly
Chuck Luciano
Just kidding.
Chuck
--
==============================================================================
Name : Chuck Luciano Witticism : Remember, in theory, there is no
Email : ch...@csn.org difference between theory and practice.
Phone : 303-421-9113 However, in practice, this is not true!
Tom O Breton
>Could this be called, comp.oxymoron?
>
>Just kidding.
Well, joke or not, I'd like to see it done. Say, add a utility that
intelligently plucks switches and options and stuff from man pages and
shows them to you, in accessible hypertext style, for starters.
Tom
--
The Tom spreads its huge, scaly wings and soars into the sky...
(t...@world.std.com, TomB...@delphi.com)
thi
> In article <CEDvz...@csn.org> ch...@teal.csn.org (Chuck Luciano) writes:
> >Could this be called, comp.oxymoron?
> >
> >Just kidding.
>
>
> Well, joke or not, I'd like to see it done. Say, add a utility that
> intelligently plucks switches and options and stuff from man pages and
> shows them to you, in accessible hypertext style, for starters.
oh no.... this reminds me of my boss showing me tops-20 command
completion. nifty but what about third party software?
actually, it reminds me of jurassic park where lex, navigating flight-
simulator-style through directory farms, sez "i know this -- this is
unix!" :-)
do we really want this? (on my vt100?)
thi
--
------------------------------------------------------------------------
t...@eclu.psu.edu db
+1 814 865 1444 __
Jim Clarke
t...@world.std.com (Tom O Breton) writes:
$IBSUBQUOTE
>In article <CEDvz...@csn.org> ch...@teal.csn.org (Chuck Luciano) writes:
>>Could this be called, comp.oxymoron?
>>
>>Just kidding.
$ESUBQUOTE
>Well, joke or not, I'd like to see it done. Say, add a utility that
>intelligently plucks switches and options and stuff from man pages and
>shows them to you, in accessible hypertext style, for starters.
$QUOTESIG
> Tom
>--
>The Tom spreads its huge, scaly wings and soars into the sky...
>(t...@world.std.com, TomB...@delphi.com)
$EQUOTE
$IBREPLY
I'm starting to feel old. UNIX unfriendly? You ain't seen nothing!
$SIGNATURE
--
Jim Clarke -- Dept. of Computer Science, Univ. of Toronto, Canada M5S 1A4
cla...@csri.toronto.edu or cla...@csri.utoronto.ca // (416) 978-4058
$EDATA
todd
: Could this be called, comp.oxymoron?
No way, dude. There is a massive effort going on right now
to develop an X-based interface, with on-line help and drag-
and-drop capabilities to the ever-popular 'soelim' command.
Todd.
--
Todd Bradley -- Supreme Ruler of The Galaxy | Visual Numerics, Boulder, CO
"If your OS doesn't use MB2 and MB3, you | 303-581-3293
might as well be using a Mac!"--OS/2 bigot | Email: to...@boulder.vni.com
Dale Harris
>Chuck Luciano (ch...@teal.csn.org) wrote:
>: Could this be called, comp.oxymoron?
>
>No way, dude. There is a massive effort going on right now
>to develop an X-based interface, with on-line help and drag-
>and-drop capabilities to the ever-popular 'soelim' command.
>
I have no problem with this, so long as I can always get back to a command
line that has at least the capability of the window interface.
--
__ _ Dale A. Harris {rod...@ecst.csuchico.edu}
/ ) // / Government embodies waste and
/ / __. // _ , , , __. _ /_ _ __ _ bureaucracy, and you
/__/_(_/|_</_</_ (_(_/_(_/|_/_)_ / /_</_/ (_</_want it to run Health Care
.
>: Could this be called, comp.oxymoron?
>
>No way, dude. There is a massive effort going on right now
>to develop an X-based interface, with on-line help and drag-
>and-drop capabilities to the ever-popular 'soelim' command.
UNIX has all sorts of capabilities. That's not what being
user friendly is all about.
Is this group a joke, or is there actually an organized effort
out there? User friendly UNIX wouldn't be UNIX. To be user friendly,
it must be easy to use, administer, repair, backup and install. An
effort aimed at just one of these areas ("use") wont accomplish
anything.
--
The big mistake that men make is that when they turn thirteen or fourteen and
all of a sudden they've reached puberty, they believe that they like women.
Actually, you're just horny. It doesn't mean you like women any more at
twenty-one than you did at ten. --Jules Feiffer (cartoonist)
John Nash
>In article <1993Oct4.1...@pvi.com>, todd <to...@pvi.com> wrote:
>>Chuck Luciano (ch...@teal.csn.org) wrote:
>>: Could this be called, comp.oxymoron?
>>
>>No way, dude. There is a massive effort going on right now
>>to develop an X-based interface, with on-line help and drag-
>>and-drop capabilities to the ever-popular 'soelim' command.
>
>
> UNIX has all sorts of capabilities. That's not what being
> user friendly is all about.
>
> Is this group a joke, or is there actually an organized effort
> out there? User friendly UNIX wouldn't be UNIX. To be user friendly,
> it must be easy to use, administer, repair, backup and install. An
> effort aimed at just one of these areas ("use") wont accomplish
> anything.
We use lots of services in molecular biology, where a DOS or Mac box
just won't cut it. In addition, there are lot of very nice mol. bio.
software out there based on UNIX. It is pure arrogance to expect the
DOS/MAC or even VAX user to come across to UNIX and be expected to be
fully conversant with new applications, *and* an arcane operating
system with all its attendant folklore (incl. the "give 'em enough
rope to hang themselves" mindset) - and this statement is coming from
a command-line UNIX fan! We cannot sneer at their platforms and then
not give them the opportunity to learn ours (yes - in the last
statement, I'm oversimplifying, generalising and being elitist and
bloody-minded, but there's some truth in the sentiment, please take it
in the humour with which it's intended).
In many cases, these new UNIX users are new to computers. These are
often not fresh-faced grad. students who deserve any hazing they get
;-) but older scientists who have to become computer proficient, as
part of the chain of having to learn a new technology (i.e. molecular
biology).
Often, as beginners, X is not an option to them. These sort of users
are not only found in mol. biol., and I think they deserve the first
priority lest UNIX keep its "techie" reputation.
--
John Nash (na...@nrcbsa.bio.nrc.ca)
Institute for Biological Sciences, National Research Council of Canada,
*** Disclaimer: All opinions are mine, not NRC's! ***
Jonathan C Patrick
I'm not sure scientists or other "technologists" are the best sample for
considering the friendliness of UNIX. Maybe the business user (gasp!)
might be a more useful data point. Having come very late to UNIX via an
ALPHA OSF/1 server and the internet itself, I confess that I'm having
trouble going back to the command line. The argument in corporate America
has been MAC or Windows vs the DOS command line. Usually the response is
in favor of the GUI and point and click and similar menu bars and icons
and the like. Corporate users expect things to be consistent and simple.
UNIX is powerful, semi-consistent, incredibly rich all the things that
user-friendly doesn't mean to someone using the system at the same time it
means everything to someone building the system. Am I off track here, or
do we need a clarification of who the customer of the user friendly
discussion is. As a technician, I love this stuff; as a manager of
technology I want to run and hide when people talk of open-systems and
UNIX. That fundamental ambiguity is at the heart of any user-friendly
discussion regardless of operating system.
Tom O Breton
> completion. nifty but what about third party software?
What are you talking about? Man pages are standard, and those odd
programs that ain't got them are no worse off than before.
thi
In article <CEEH7...@world.std.com> t...@world.std.com (Tom O Breton) writes:
>
> > oh no.... this reminds me of my boss showing me tops-20 command
> > completion. nifty but what about third party software?
>
> What are you talking about? Man pages are standard, and those odd
> programs that ain't got them are no worse off than before.
maybe i meant tops-10 (is there such a thing?). regardless, this
particular operating system allowed command (ie, name as well as op-
tions) completion by hitting '?' during the command line editing. it
listed out relevant options if current input was ambiguous, as well.
my point above was that the feature was nifty, but that for this o/s
(note that o/s == shell in this case), it was a hassle if not impos-
sible to implement the same features for new software that wasn't
bundled with system software.
chill out, i'm not attacking man pages.
Matt Ackeret
>Well, joke or not, I'd like to see it done. Say, add a utility that
>intelligently plucks switches and options and stuff from man pages and
>shows them to you, in accessible hypertext style, for starters.
Sounds like you want something like the "Commando" interface in MPW.
You can do things via command-line-like interface, or end the command name
with an ellipses(...) and then a little modal window comes up where you can
use checkboxes and the like to do the whole command.. All it's literally
doing (and you can see it doing it) is changing command-line switches and
such, then it gives the command to the shell...
Not that I'm in any way comparing MPW to UNIX... the thing that I like
most about MPW, Projector (the project management tool), has no relevance to
UNIX as far as I know.. The things that bug me about MPW is that they take
a lot of things from UNIX [stated in the manual], yet appear to do things
differently _just for the sake_ of being different... e.g. option-d instead
of \, and a few other "this character instead of that character".. The
biggest thing is simply a major mindset difference.. I don't like the fact
that you can do every command in any window.. because of this fact, you
have to use openapple-return (or the enter key) instead of just return...
annoys the hell out of me.
BTW, if you're curious about UNIX (or at least UNIX while conforming
to GS/OS) on a "small" computer, check out comp.sys.apple2.gno, which discusses
the third party GNO multitasking shell for the GS.
--
unk...@apple.com Apple II Forever
unk...@ucscb.ucsc.edu These opinions are mine, not Apple's.
John Wroclawski
From: t...@ECLU.psu.edu (thi)
maybe i meant tops-10 (is there such a thing?). regardless, this
particular operating system allowed command (ie, name as well as op-
tions) completion by hitting '?' during the command line editing. it
listed out relevant options if current input was ambiguous, as well.
my point above was that the feature was nifty, but that for this o/s
(note that o/s == shell in this case), it was a hassle if not impos-
sible to implement the same features for new software that wasn't
bundled with system software.
As a point of interest, this is untrue. Tops-20's interactive command
parser, complete with prompting, interactive help, command completion,
type-sensitive parsing, etc., etc. was implemented as a system call
which could be used by any program. All one needed to do was supply a
little set of tables which told the parser what to do. A huge number
of third-party programs used this stuff, and the result was one of the
friendliest command line environments around. Aside from ease of use,
the existance of this system call led to standardization of the
interface, which made it much easier to learn new programs. Apple
reinvented this concept some years later...
When the PDP-10s started to fade away, some folks at Columbia wrote an
equivalent library for unix, so that they could give their users the
same look and feel. Since it wasn't universally available, it didn't
have quite the same impact. Still a nice way to write a program with a
command-line interface, though.
-john
John Wroclawski
j...@lcs.mit.edu
Andy
/andy
--
--------------------------------------------------------------------------
Bill Gates should limit his salary to the | PI9...@pt.hk-r.se is:
number of bytes addressable by the latest | Andy Eskilsson
version of MS-DOS, and be taxed based on | Tranbaersv. 25:12
the number of bytes of RAM needed by the | s-372 38 Ronneby
latest version of MS-Windows | SWEDEN
--------------------------------------------------------------------------
"Hey, it's Unix! I know this!"
Lex, Jurassic park.
Apples has been a problem ever since eden.
Don't walk in front of me, I might be unable to follow you.
Don't walk after me, I might be unable to lead you.
Just walk by my side and be my friend.
Wes Groleau x1240 C73-8
could with VMS. I have also used X-windows tools extensively. There are a
lot of things I do with shells that I can't do with X-windows. But what I can
do with windows, I can do much more efficiently than with shells. It's kind
of like K-Mart vs. John Doe Ford. I can go to the former every week for
a quart of oil, cheap and quick. But when I'm finally ready to fix the leak,
I go to "the right place" for a head gasket set, etc. and get out the fancy
tools. Some people I work with give me the impression they'd get out the fancy
tools every week, to build a refinery to add some oil. I've seen three ways
to delete a section from a document or program:
1. Go into vi and type "dd" a hundred times. (They know how to type "100dd"--
why don't they do it ?!?!?
2. Go to a shell and type a rather cryptic and rather long awk or perl command.
(Two minutes to minutes to figure out the command, 30 seconds to type it,
two seconds for it to execute, five seconds to survey the results and see
that you screwed up big-time, the rest of the day to recover.)
3. Drag the mouse across it, and press "Cut"
----------------------------------------------------------------------------
Real programmers type
set OM = ( `who am i | nawk '{$0=$6;sub ("^.","");sub ("[)]","");FS=".";if (NF==0 )[node=""} else {if (NF==1){node = $1 "." $2}};print node}'` )
But WIMPs (Window-Icon-Mouse-Persons) get the job done.
-- someone who does both
--
Wes Groleau gro...@e7sa.epi.syr.ge.com
137 Ruth Avenue
Syracuse, NY 13210
Michael Salmon
|> >Could this be called, comp.oxymoron?
|> >
|> >Just kidding.
|>
|>
|> Well, joke or not, I'd like to see it done. Say, add a utility that
|> intelligently plucks switches and options and stuff from man pages and
|> shows them to you, in accessible hypertext style, for starters.
Many years ago I heard of an experimental ai shell that tried to do
what you meant rather that what you said. It used the man page
descriptions to determine which command you needed to use and then used
some form of negotiation with the user to determine how to map the
parameters. I had heard that it was quite successful but I haven't
heard anything more.
Besides Unix is techo friendly, who cares about users :^).
--
Michael Salmon
#include <standard.disclaimer>
#include <witty.saying>
#include <fancy.pseudo.graphics>
Ericsson Telecom AB
Stockholm
peter da silva
> As a point of interest, this is untrue. Tops-20's interactive command
> parser, complete with prompting, interactive help, command completion,
> type-sensitive parsing, etc., etc. was implemented as a system call
> which could be used by any program. All one needed to do was supply a
> little set of tables which told the parser what to do.
At this point I'd like to put in a plug for Eric Allman's "parseargs"
routine. It does all the argument parsing and is a handy hook for stuff
like this, all based on a little table. Brad Appleton at Harris and I have
enhanced it to the point where you can take a program that uses parseargs
and compile it on DOS, Amiga, VMS as well as UNIX and it'll provide the
native command line interface for each. Adding a user-friendly popup to
fill in the blanks under X or Windows or MacOS would be a piece of cake,
for someone used to that interface.
--
Peter da Silva `-_-'
Network Management Technology Incorporated 'U`
1601 Industrial Blvd. Sugar Land, TX 77478 USA
+1 713 274 5180 "Hast Du heute schon Deinen Wolf umarmt?"
Andy Newman
"User Friendly" is a stupid phrase. It is totally subjective as it depends
on the user. I'm a programmer and I use UNIX. I think UNIX is friendly
therefore UNIX is user friendly.
--
Andy Newman (an...@research.canon.oz.au)
Grant Edwards
: In article <TTN.93Oc...@ECLU.psu.edu> t...@ECLU.psu.edu (thi) writes:
: maybe i meant tops-10 (is there such a thing?). regardless, this
: particular operating system allowed command (ie, name as well as op-
: tions) completion by hitting '?' during the command line editing. it
: listed out relevant options if current input was ambiguous, as well.
: When the PDP-10s started to fade away, some folks at Columbia wrote an
: equivalent library for unix, so that they could give their users the
: same look and feel. Since it wasn't universally available, it didn't
: have quite the same impact. Still a nice way to write a program with a
: command-line interface, though.
Columbia... command line completion... sounds like kermit! Are you
talking about the wart library that ckermit uses?
--
Grant Edwards |Yow! Where does it go when
Rosemount Inc. |you flush?
|
gra...@rosemount.com |
Bryan Walls
(Michael Salmon) wrote:
> Many years ago I heard of an experimental ai shell that tried to do
> what you meant rather that what you said. It used the man page
> descriptions to determine which command you needed to use and then used
> some form of negotiation with the user to determine how to map the
> parameters. I had heard that it was quite successful but I haven't
> heard anything more.
> Besides Unix is techo friendly, who cares about users :^).
>
> Stockholm
I think your talking about Xerox's DWIM (Do What I Mean) facility. It was
pretty cool, and worked pretty well. The version I used was based in
InterLisp-D on their Lisp workstation. Xerox really developed (develops)
some awesome stuff, but their marketing is so amazingly bad for
computers...
--
Bryan Walls My opinions are not NASA policy nor, in general, vice versa
bwa...@msfc.nasa.gov
Ninja
> Well, joke or not, I'd like to see it done. Say, add a utility that
> intelligently plucks switches and options and stuff from man pages and
> shows them to you, in accessible hypertext style, for starters.
try tkman for tcl/tk, not perfect, but still really good
Ninja
Walt Sullivan
Tom> NNTP-Posting-Host: eclu.psu.edu
Tom> In article <CEDzo...@world.std.com> t...@world.std.com (Tom O Breton) writes:
> In article <CEDvz...@csn.org> ch...@teal.csn.org (Chuck Luciano) writes:
> >Could this be called, comp.oxymoron?
> >
> >Just kidding.
>
>
> Well, joke or not, I'd like to see it done. Say, add a utility that
> intelligently plucks switches and options and stuff from man pages and
> shows them to you, in accessible hypertext style, for starters.
Tom> oh no.... this reminds me of my boss showing me tops-20 command
Tom> completion. nifty but what about third party software?
In a previous life, I developed 3rd-party software for TOPS-20. The
COMND JSYS (which does all the nifty completion tricks, noise words,
defaults, ...) is available to every programmer. Unfortunately, it
means that TOPS-20 has to run your program to be able to parse the
command for your program. The advantage it has is that every program
has the same style command interface, with the same features.
Part of the perceived "unfriendliness" of Unix is the maze of twisty
little options, all different that the novice must contend with. Of
course, he/she can set up shell aliases to provide the most commonly
used option sets, but once he/she can do that well, he/she is no
longer a novice.
I'd oppose any attempt to give easy access to the powerful (dangerous)
commands to novices who might not understand the consequence of their
actions.
--
Walt Sullivan
INTERNET: wa...@bnr.ca (work)
UUCP: wa...@orbit.ocunix.on.ca (home)
FIDO: 1:163/109.4
Thomas Dunbar
user-friendly to the novice involves making it impossible for
the novice to do stupid things. but, as ritchie has said, he
didnt want to do that because it would interfere with doing clever
things.
if a feature ceases to be a benefit when one ceases to be a novice,
it's a burden to all that can/will learn.
thomas (tdu...@vtaix.cc.vt.edu)
Budi Rahardjo
...
> Is this group a joke, or is there actually an organized effort
> out there? User friendly UNIX wouldn't be UNIX. To be user friendly,
> it must be easy to use, administer, repair, backup and install. An
> effort aimed at just one of these areas ("use") wont accomplish
> anything.
The term user-friendliness refers to normal/ordinary users.
So we can take administer, repair, backup, and install out.
Leat the `easy to use' though.
Heck, any OS is difficult to administer.
-- budi
T.Dunn
>I'd oppose any attempt to give easy access to the powerful (dangerous)
>commands to novices who might not understand the consequence of their
>actions.
Hehehee... Sure, why not? We could have an envar NOVICE that would
trigger warnings on such obviously silly things such as "rm * -R".
Anyway, trust me, the aforementioned novices will learn very quickly
what the consequences of their actions are. And generally, they won't
be able to harm much.
You did back up, didn't you?
--
Mollison's Bureaucracy Hypothesis:
If an idea can survive a bureaucratic review and be implemented
it wasn't worth doing.
T.Dunn
>"User Friendly" is a stupid phrase. It is totally subjective as it depends
>on the user. I'm a programmer and I use UNIX. I think UNIX is friendly
>therefore UNIX is user friendly.
You're not keeping up on the jargon. The term "user friendly" isn't
something that we just made up for this group. It has a specific
meaning, and doing a semantic breakdown of it isn't going to change
the meaning of the term. Go look in a magazine post 1990 if you still
don't understand the generally accepted meaning of the term.
David McNicol
>Walt Sullivan <wa...@bcarh2ec.bnr.ca> wrote:
>>I'd oppose any attempt to give easy access to the powerful (dangerous)
>>commands to novices who might not understand the consequence of their
>>actions.
>Hehehee... Sure, why not? We could have an envar NOVICE that would
>trigger warnings on such obviously silly things such as "rm * -R".
Why not just:
alias rm 'rm -i'
and the like.. That's what our sys-admins do for new users.. when they
know enough UNIX to change it in their .cshrc, then they are judged
to be wise enough not to type ``rm -rf ~''.
Dave.
+----------------------------------------------------------------+
| |
| David McNicol, 4th year laser physics, Strathclyde University. |
| |
| e-mail: cab...@cc.strath.ac.uk |
| |
+----------------------------------------------------------------+
Jean-Luc Bernard
>
>Well, joke or not, I'd like to see it done. Say, add a utility that
>intelligently plucks switches and options and stuff from man pages and
>shows them to you, in accessible hypertext style, for starters.
>
> Tom
This is possible ! If you are a Sun user and run Open Windows, I found a few
years ago a wonderful program named 'catcher' which purpose is to generate an
simple Open Windows graphical interface for commands with many nifty arguments :
Host geocub.greco-prog.fr (192.33.148.2)
Last updated 04:09 6 Aug 1993
Location: /pub/X11
FILE -rwxr-xr-x 109517 bytes 00:00 21 Jun 1991 catcher.tar.Z
Note : you will have to write the simple scripts that are needed to generate the
GUI for a given command - Documentation is included in the tar file.
Enjoy it !
---
Unix is the answer, but only if you phrase the question very carefully.
_________________________________________________________________
Jean-Luc BERNARD Computer Center
Advanced Computer Research Institute Email: jlbe...@acri.fr
1 Boulevard Marius Vivier-Merle Voice: (+33) 72.35.84.03
69443 LYON CEDEX 03 - FRANCE Fax (+33) 72.35.84.10
Chris Gray
Andy Newman <an...@research.canon.oz.au> wrote:
>"User Friendly" is a stupid phrase. It is totally subjective as it depends
>on the user. I'm a programmer and I use UNIX. I think UNIX is friendly
>therefore UNIX is user friendly.
And td...@ecst.csuchico.edu (T.Dunn) replied:
>You're not keeping up on the jargon.
Inevitably, someone has to post an excerpt from the Jargon File:
user-friendly: adj. Programmer-hostile. Generally used by hackers in
a critical tone, to describe systems that hold the user's hand so
obsessively that they make it painful for the more experienced and
knowledgeable to get any work done. See {menuitis}, {drool-proof
paper}, {Macintrash}, {user-obsequious}.
So it might as well be me *8^>.
__________________________________________________________________________
Chris Gray cg...@se.alcbel.be Compu$erve: 100065.2102
Ignore my broken mailer - the addresses above are the only truth
__________________________________________________________________________
Alle menschen sind Auslaender. Fast ueberall.
Morgan J Ryan -- Neil Patterson Publishers
I'm new to unix, studying hard and learning slowly because I'm pretty
dumb. I have a dialup account for internet access, figured I'd start to
learn unix because it was there. The first thing a new guy like me sees is
the editor. And it was vi. Of course I said, "Very funny, but I need to
write some email, where is the unix word processor?" And helpful people
said "vi! A superb little program!" So for a few weeks I actually
*programmed* my email. I got to say "Fixed that typo. I guess I'm a hacker
now."
I ransacked the net looking for a real word processor. I knew there
had to be one. Finally I turned up pico. It ain't much, but at least it
has linewrapping. At least when I try to correct a typo it doesn't blow my
text to lowercase confetti.
So I ask you. Why do unix pros use an editor that was clunky in
1971? Is vi a test? Beat vi and you can learn the rest?
I'm starting to appreciate unix. I alias all over the place. After
many months I'm starting to understand man pages enough to browse for fun
and learn new ways to screw around.
And all along I wonder, have these guys who mock Macs seen one in
the last eight years? Do they realize that anything I can think of to do,
I can do on a Mac without reading directions, or asking on the internet,
or communing with the kernel or whatever it is wizards do?
A booster of vi is mocking my Mac? I don't get it. Really, I don't
see much to debate about *whether* unix is unfriendly. Unix is
mumbo-jumbo. If I turned on my Mac and it said mumbojumbomumbo the way
unix does, I would check the case for bulletholes.
Thank god that unix people are so generous with help. I'm still
asking 100 questions a week and getting gentle patient answers.
How *do* I get plaintext man pages?
Would I have to ask this with a dopey old (l)user-friendly os?
I have an inkling of what unix can do, so I'm muddling along
learning. But I swear it seems that an interface could only *stay* this
savage on purpose.
Morgan
Paul Lalonde
I'm not sure you know just how much flamage you are inviting, but I
think I'll actually try to tell you why I like vi, and maybe some of
the reasons why you don't.
In comp.unix.user-friendly you write:
> I ransacked the net looking for a real word processor. I knew there
>had to be one. Finally I turned up pico. It ain't much, but at least it
>has linewrapping. At least when I try to correct a typo it doesn't blow my
>text to lowercase confetti.
vi has line wrapping - the command is :set wrapmargin=8
(change the 8 to the left margin size you want)
To re-format a paragraph try being at the start of the paragraph
and typeing !}fmt followed by return.
> So I ask you. Why do unix pros use an editor that was clunky in
>1971? Is vi a test? Beat vi and you can learn the rest?
Because it's not clunky :-) The catch is that if you expect to use
it like most people use word processors it's not going to work for you.
Vi requires a bit of a different mind set. You don't cruise around the
text with the arrow keys for more than a few strokes in a row. Rather,
you search for a string using the / command, which brings you right
there. You don't dig around everywhere for a function, you use tags
(man ctags for more details). You don't delete each character, but
rather the whole word with the dw command. You change words by using
the cw command instead of deleting and re-typing. The list continues,
but the deal is that you have to learn how vi does things before you
become productive with it. But once you know them you can go like
greased lightning. Challenge an old-time vi user to a race on some
editing jobs. Vi will win in their hand ahead of you on a WP or a
point and click nearly every time.
And I've not even mentionned what you can do with the regular expression
language for searches and substitutions. I've yet to see a WP or a
point and click that provides that much useful power.
> And all along I wonder, have these guys who mock Macs seen one in
>the last eight years? Do they realize that anything I can think of to do,
>I can do on a Mac without reading directions, or asking on the internet,
>or communing with the kernel or whatever it is wizards do?
Yep, we do. But we can do it faster and more efficiently, once we've
learned how to use the tool. Look at it this way: You can use a hand
saw all you like. It's use is simple, intuitive, and easy. I'll
use a table saw. I have to know how to adjust the fences, set depth
of cut, and find the power switch, and even worry about cutting my
fingers. You might be faster on the hand saw for one cut, but any more
than that and my extra learning about the tool will make me faster.
> How *do* I get plaintext man pages?
nroff -man /usr/man/man1/man.1
(substitute the path and name of the manpage file for the man.1 example)
> Would I have to ask this with a dopey old (l)user-friendly os?
Maybe, but then again you wouldn't be actually learning how to use
a table saw :-)
> I have an inkling of what unix can do, so I'm muddling along
>learning. But I swear it seems that an interface could only *stay* this
>savage on purpose.
Well, tradition, and a different mind-set.
Paul
--
Paul A. Lalonde Internet: lal...@cs.ubc.ca
"On ne voit bien qu'avec le coeur, l'essentiel est invisible aux yeux"
- Antoine de St.-Exupery
whi...@oberon.meakins.mcgill.ca
: So I ask you. Why do unix pros use an editor that was clunky in
: 1971? Is vi a test? Beat vi and you can learn the rest?
No. Mastering vi does not turn one into a unix pro.
At best you can say that you have mastered an editor from 1970's. groovy.
: And all along I wonder, have these guys who mock Macs seen one in
: the last eight years? Do they realize that anything I can think of to do,
: I can do on a Mac without reading directions, or asking on the internet,
: or communing with the kernel or whatever it is wizards do?
A wizard is someone with extrordinary powers. It is sad but true that
if you master the goofy set of unix tools you have a very powerful
machine at your disposal. That is of course assuming that it is
connected to a network - handling mail, news, running nfs, etc.
If it isn't it is on a network unix is probably used as program launcher
( a 400 MB version of command.com and edlin ) that came for free with
your great number crunching hardware.
: I have an inkling of what unix can do, so I'm muddling along
: learning. But I swear it seems that an interface could only *stay* this
: savage on purpose.
Writing better tools takes too much time. Worse yet is that
everyone wants them for free. No market.
Whitney
Morgan J Ryan -- Neil Patterson Publishers
>I'm not sure you know just how much flamage you are inviting,
in this group. And I'll be the first to admit I wish I knew what unix
people know. But does your comment on line-wrapping make my point?
> vi has line wrapping - the command is :set wrapmargin=8
>(change the 8 to the left margin size you want)
>To re-format a paragraph try being at the start of the paragraph
>and typeing !}fmt followed by return.
I read the vi man pages a couple times a long time ago. How did I forget
:set wrapmargin=8? (Incidentally, thanks. Once again I get help...)
Could an oldtime vi user really do a complex editing job faster than me in
Word or Wordperfect or any of those suboptimals? Is there any function in
vi that I don't have in Word? Search, change, sort, I can do all that
stuff. Fast. I think the handsaw analogy breaks down. Maybe they are both
powersaws. But mine I can walk right up to and use. If I walk right up to
vi and start sawing, ho boy, call an ambulance. Ease of use seems to me a
Big Virtue. Engineers make just about everything easy to use. Music
systems are as complex as computers, but they've been made easy to use.
I'm wondering why unix isn't easy to use already? Ease of use
happens naturally with everything else. But not unix. Why? Morgan
Fazal Majid
In article <293rui$k...@inxs.concert.net>, mor...@rock.concert.net (Morgan J Ryan -- Neil Patterson Publishers) writes:
|>
|> Could an oldtime vi user really do a complex editing job faster than me in
|> Word or Wordperfect or any of those suboptimals? Is there any function in
|> vi that I don't have in Word? Search, change, sort, I can do all that
|> stuff. Fast. I think the handsaw analogy breaks down. Maybe they are both
|> powersaws. But mine I can walk right up to and use. If I walk right up to
|> vi and start sawing, ho boy, call an ambulance.
Sure, when I first started using Unix (Ultrix 4.2 on a Vax), I would rather use
cat as a text editor than vi... After that I used the Wysiwyg X editor dxnotepad
that comes with DECstations. The real break is when I discovered the light of
emacs :-)
Now, I can't go back to any of those so-called "user-friendly" editors. And I am
definitely sure my productivity is at least double my former one. Why do you
think professional programmers pay hundreds of dollars on Emacs-style editors
such as Brief and Epsilon on PCs, even though most compilers come with supposedly
user-friendly integrated development environments. Not just programmers at that,
the secretaries here at Yale use epsilon too.
--
Fazal MAJID | /\ /\ /\ /\ 0- Sarff, the serpent
ma...@math.yale.edu | /\ / \ / \ / \ / \ / with the terrible
ma...@inf.enst.fr | \/ \/ \/ \/ \/ sting.
Thomas A Fine
>Could an oldtime vi user really do a complex editing job faster than me in
>Word or Wordperfect or any of those suboptimals? Is there any function in
>vi that I don't have in Word? Search, change, sort, I can do all that
>stuff. Fast. I think the handsaw analogy breaks down. Maybe they are both
>powersaws.
As compared to Suntools textedit, they are both powertools. Nevertheless,
vi is even more powerful than Word or Wordperfect. There are hundreds
more commands in vi than in Wordperfect. Even the most experienced
vi user only uses a fraction of what vi can do.
>But mine I can walk right up to and use. If I walk right up to
>vi and start sawing, ho boy, call an ambulance. Ease of use seems to me a
>Big Virtue. Engineers make just about everything easy to use. Music
>systems are as complex as computers, but they've been made easy to use.
> I'm wondering why unix isn't easy to use already? Ease of use
>happens naturally with everything else. But not unix. Why? Morgan
A computer is different than everything else in the world. A car, a
toaster, these are products. They were designed and built with a set
of tools that engineers and manufacturers use. These tools are complex,
varied, and obscure. But you never see the tools, only the product.
You don't care how they get the hardened steel cylinder lining into
the aluminum engine block. You just drive.
With a computer, the computer can be either the tool or the product.
In the case of the Mac, its a product. In the case of UNIX, its
almost always a tool. But once I've used that tool, and made a product,
I could sell UNIX machines with my software on them, and this could be a
nice user-friendly product, if I wanted it to be. Underneath it all,
the tool would still be there, shipped right along with the product.
So, once again, I'll say what other people have already said: The price
of power and flexibility is complexity. Maybe it doesn't have to be
quite as complex as UNIX, but, it still has to be complex.
tom
Jeff Alsip
It helps a new user figure out the correct command line entry for
a given command... just by answering a set of yes/no questions.
As every option flag is resolved you can watch the command line
grow at the bottom of the screen. The package came with all the
tools necessary for an administrator to create extra screens for
commands not already covered. Very efficient, very powerful...
but then thats AT&T!
One thing that has to be realized is that adding alot of support
scripts to help the beginner will eat up ALOT of disk, and may
cause an in-efficient draw on system services.
Jeff Alsip
jef...@vpnet.chi.il.us
Jeff M Younker
Computers do not exist to be used by programmers and system administrators.
They exist so that other people can accomplish their jobs. They exist to
make things easier for the 'users'.
These 'users' shouldn't have to worry about How to get the computer to help
them with their jobs. If users are spending time worrying about how to
read their mail, the computer is getting in the way. They should be
reading mail, not searching through man pages.
If people are worrying about setting up environment variables to
make their typesetting software work, something is wrong. They should be
worrying about how their document looks.
If the professor down the hall is worrying about how to incorporate the
plots from mathmatica into his grant proposal, something is wrong. This
person should be worrying about the content of the proposal, not the
mechanics of adding a picture.
If an inventory worker is worrying about path names, he's not taking
inventory.
Th computer should not pervent a person from completing his/her job. The
person's job is not there so that the person can use a computer. The
computer is there so that the person can perform their job more
efficiently. The computer should make it easier to do the job. If it
doesn't, the computer is not worth having.
If I own a company and I have to make a choice between a system
that every idiot in the company can use with no training, and a cryptic
text orient environment that requires months of training to be used
effectively, I'm going to pick the easy to use environment.
Some may argue that their are other reasons for choosing a computer system
for company, and their are. But in the future, these technical issues are
going to vanish into woodwork. All the things that gave unix it's
technical edge over the years are becoming standard. Any new operating
system is going to compare favorabley to unix in technical aspects. They
will all have advantages and disadvantages. The distinguishing factor is
going to be ease of use. Unless unix boxes hide all that crypticness and
are made easy to use, unix is going to die.
--
jeff younker je...@math.uh.edu
- These are my opinions. If they resemble those of the -
- University of Houston it is possible that I am very sick. -
Mike Mussington
: actually, it reminds me of jurassic park where lex, navigating flight-
: simulator-style through directory farms, sez "i know this -- this is
: unix!" :-)
: do we really want this? (on my vt100?)
*I* DO!!!
--
Gregory Owen
je...@karazm.math.uh.edu (Jeff M Younker) writes:
> Computers do not exist to be used by programmers and system
> administrators. They exist so that other people can accomplish their
> jobs. They exist to make things easier for the 'users'.
have forgotten.
But as a side question, and one which is becoming the basis of
this group -- is user-friendliness worth the lack of power it implies?
One of the best things on a Unix system (I know shells are not Unix,
but they sure as hell aren't DOS and WINDOWS) is the way quick and
easy control structures can be used to perform repetitive tasks. I
use "foreach" probably a minimum of 100 times in an 8-hour work day.
Then I go to a DOS box with Windows, and have to do something
repetitive (like convert a slew of tiffs to pcx format). It requires
me to press the same buttons, repeatedly, and with proper timing
(i.e., no doing things ahead of time and waiting for the computer to
catch up.) If I push ahead to the hot keys -- "for the power user" --
I still end up pressing the same keys, repeatedly, and with proper
timing.
As a side note, what Windows needs to be user friendly is a
macro utility, sort of like 'start-kbd-macro' that would let you make
a bunch of clicks or keystrokes and then repeat them as many times as
you wanted.
Back to the original point, how can Unix be made easy to use?
Good system administration. If the system administrator makes sure
every package is installed correctly, and that every user who needs to
use a package has the correct environment, then the user won't have to
worry about those fiddly things. I've trained a user to use a limited
subset of the system -- a word processor, an image viewer and a file
tool -- and it didn't take long at all for her to be up and running.
(Granted, she had troubles with textedit's quirks -- I would not raise
a finger against someone who tried to write a _GOOD_ X text editor for
the masses, a la word or wordperfect.)
A lot of work for the sysadmin? well, a bit more. Assume that
this only needs to be done for the "every idiot" who just has to do
this or that little thing, and that "power users" can fend for
themselves.
Truth to tell? I'm not 100% convinced of everything I just
said. Let's see some good discussion. Gawd, I gotta watch out I
don't ramble too much on this group.
Greg Owen { go...@allegro.cs.tufts.edu, go...@xis.xerox.com }
1.01 GCS/GO d++ p+ c++ l++ u++ e+ -m+ s++/- n- h !(f)? g+ -w+ t+ r-- y?
"These fragments I have shored against my ruins/Why then Ile fit you.
Hieronymo's mad againe./Datta. Dayadhvam. Damyata."
Peter da Silva
> Word or Wordperfect or any of those suboptimals?
Probably, with little user-friendly things like automatic marks and delete
buffers and 'last search' and consistent context-based short commands.
Just not having to move off the home keys to hit F1... that saves time.
> Is there any function in vi that I don't have in Word?
Pipe block of text through arbitrary external program? Regular expression
based complex search and replace?
g/^[ ]*int[ ]*foobuf/s/BUFSIZ/FOOBUFSIZ/
changes every declaration of foobuf to the new size, even if the declaration
isn't *quite* the way you thought.
> I'm wondering why unix isn't easy to use already? Ease of use
> happens naturally with everything else.
You think? I find most DOS editors REALLY frustrating. Even the best, Brief,
has the idiocy of 'search matches shortest pattern'. Ease of use depends
on what you're using it for.
DOS. "more file". Oops, just sits there.
MacOS. I keep forgetting that closing the last window doesn't terminate
the application. I've seen people try to double-click on menu items
after months of use.
VMS. Just what *is* the 'ftp' command in *this* version of TCP/IP?
Pat Fitzpatrick
that has not been mentioned, perhaps because it is "common knowledge."
It's an editor called "emacs." Many systems have it installed. Very
nice editor!
Pat Fitzpatrick
Juergen Nickelsen
kj...@typhoon.atmos.colostate.edu (Pat Fitzpatrick) writes:
[Disclaimer: I am one of those people who would like to change their
login shell to Emacs. I am able to work with vi, though.]
I would not consider Emacs a user-friendly editor for novice users.
Both Emacs and vi share the same disadvantage: If you get into them by
accident, you don't know how to get out. (At least Emacs offers a
tutorial which tells you how to get out of it, I know, but this offer
is usually gone after the user pressing keys in panic.)
If arrow keys work depends on proper installation in both.
Learning Emacs is hard in the beginning and can be frustrating
sometimes (it was for me), and you have even more ways to get stuck in
the beginning that with vi. I would even say you have more and easier
ways to screw up your text than in vi. (Imagine a large C program,
where the region spans the whole buffer by accident, and the user
types ESC g (fill-region))
Of course there are significat differences: Tutorial, multi-level
undo, online help. But (except the tutorial) you can only get real
advantage of these if you already have gained some proficiency.
But you *can* turn Emacs into a user-friendly editor with moderate
effort:
- make sure arrow keys work
- configure all keys with meaningful labels (like "PgDn") to behave
as expected
- put the most important commands on function keys
- etc.
I saw a setup like this in a beginner's undergraduate programming
class, and this was actually good. They even began with telling the
students nothing about the function keys except one, which invoked
"execute-extended-command", and made them use "find-file" and
"save-buffer" from there. Function keys were introduced later, and
even later things like C-X C-f etc. Good work.
--
Juergen Nickelsen
John Hascall
}I would not consider Emacs a user-friendly editor for novice users.
}But you *can* turn Emacs into a user-friendly editor with moderate
}effort:
Our "student development group" came up with a setup for emacs
(invoked with "ezemacs", which IMHO does the right things with
arrows, help, dangerous commands, etc.). Even though I'm a
diehard vi'er, I think it is quite a nice effort.
sample screen below -- some highlighting lost in the translation to net news :-(
Hello world
[...a couple blank lines deleted here so it fits better on the news screen...]
*Edit (above)------------------------------------------------------(below) Help*
.(File) (Cursor Movement) (Delete) (Miscellaneous)
Get File: C-x i Page Up: C-h v Delete a Char: C-d Word Wrap: C-h r
Save File: C-x C-s Page Down: C-v To End of Line: C-k No Word Wrap: C-h e
Exit: C-x C-c End of Line: C-e Backward: backspace Undo: C-x u
------------------Start of Line: C-a --------------------- Cancel: C-g
(Arrow Keys) Start of File: C-h < Page Up: Prev Screen Repaint: C-l
Up Down Left Right End of File: C-h > Page Down: Next Screen
-------------------------------------------------------------------------------
C-x i: Press [Ctrl] and x, release both keys, then press i
The meta key, M, is C-h [VT240]
To remove Help Screen: press [Help] or C-h x [EasyEmacs]
John
--
John Hascall ``An ill-chosen word is the fool's messenger.''
Systems Software Engineer
Project Vincent
Iowa State University Computation Center + Ames, IA 50011 + 515/294-9551
Peter A. Bidian
>In article <Oct08.225...@yuma.ACNS.ColoState.EDU>
>kj...@typhoon.atmos.colostate.edu (Pat Fitzpatrick) writes:
>> By the way, for those who dislike vi (like me) there is an alternative
>> that has not been mentioned, perhaps because it is "common knowledge."
>> It's an editor called "emacs." Many systems have it installed. Very
>> nice editor!
>[Disclaimer: I am one of those people who would like to change their
>login shell to Emacs. I am able to work with vi, though.]
>I would not consider Emacs a user-friendly editor for novice users.
>Both Emacs and vi share the same disadvantage: If you get into them by
>accident, you don't know how to get out. (At least Emacs offers a
>tutorial which tells you how to get out of it, I know, but this offer
>is usually gone after the user pressing keys in panic.)
>If arrow keys work depends on proper installation in both.
>Learning Emacs is hard in the beginning and can be frustrating
>sometimes (it was for me), and you have even more ways to get stuck in
>the beginning that with vi. I would even say you have more and easier
>ways to screw up your text than in vi. (Imagine a large C program,
>where the region spans the whole buffer by accident, and the user
>types ESC g (fill-region))
How right you are!! Just yesterday I opened emacs for fun, and was impressed
that it opened an extra window in X11, but then I was completely lost.
I thought it was a great idea to have the tutorial, but I didn't find it
very helpful. Sometimes it gives you command names but no Key sequence,
or I have a problem understanding the way the tutorial works.
I tried search, and by intuition I typed C-s which seamed to be the right stuff, but after I typed the string to be searched and hit return, I got
a ^J in the control window and the cursor moved to the next line.
And finally when I tried to get out... no luck. Kill window did the job then.
For now I'll stay with vi. At least it works on all terminals, you don't
need function keys (if you have to move from Sun to Iris to Mac terminals
you know that relying on keys can be a big hazard). Had enough trouble
to use vi after using the VMS editor.
Just my poor expierence...
Peter
Bill Vermillion
>
>Could an oldtime vi user really do a complex editing job faster than me in
>Word or Wordperfect or any of those suboptimals?
Probably. The fact that vi is modal - the same keys are used for
commands as entry - means no moving the fingers from the home keys to
function keys etc. But there are things in which WP will be much faster
and others that vi will be faster.
>Is there any function in
>vi that I don't have in Word? Search, change, sort, I can do all that
>stuff. Fast. I think the handsaw analogy breaks down. Maybe they are both
>powersaws.
But you can't compare vi with Word, WordPerfect, etc. vi is a text
editor the others are word processors.
You don't do font changes, underline, etc in vi. You would pass that
to something like nroff.
But when you are writing code in any language the ability to use such
things as the showmatch option to find/show parenthesis balancing is
nice.
You can take your text in vi and pass it out to/through something
external and read it back in. You can execute programs from within vi
and read their output into the program.
In short, if you need a word processor, use one. If you need a text
editor then use one of those. The Unix philosophy is not one
all-purpose to that does everything, but many samller special purpose
tools that can be combine in user defined order to accomplish the
specific task at hand.
The vi on this machine is 159k of code. I've seen them smaller. The
Unix version of Word Perfect that I have seen are all about 1.8 megs of
code. That can also be a big factor in useage.
> Ease of use seems to me a
>Big Virtue. Engineers make just about everything easy to use.
The problems with making something 'easy to use' right out of the box
means that you are requiring a lot of additional code and often
keystrokes that an experienced user finds cumbersome, and gets in the
way.
I do some work for a site that recently re-organized their computer
systems and now is bringing up a network for all users - count is about
990 machines on the net.
Prior to that the group I did the most work for were using Xenix and
Unix machines.
The casual users, those who did a few letters per day love Word Perfect
because it's like their home machine, and it has tons of help.
There are about 2 or 3 'power word processing users' who hate some of
the function in Word Perfect because it is so much harder to do things
than their previous word processor. It was Lex.
(Lex is a word processor from England - originally written on a PDP-8
in 1975. It is supremely flexible. You can interface virtually
anything to it, if you have the specs for the device. It is written in
a language called filetab-D (if I remember correctly). And because of
that you can write programs in the word processor that do word
processing functions. One of the Lex principals left and wrote
Uniplex - a lot of the Lex functionality is in Uniplex. And later SCO
introduced Lyrix, which was a decendant of Uniplex from what I can see.
Many of the editing function and keystroke sequences are indentical
among all three of those word processors.)
WP is easy to learn and it's fairly easy to produce decent documents by
just sitting down and using the help as often as needed.
But there are also production environments. By that I mean places
where the user is actually in the word processor 6 to 8 hours per day.
If you can give them something that is more efficient though hard to
learn it can be a plus.
For example. If a hard to learn word processor was 10% more efficient
than an easy to learn program, and the person used it 4 hours per day,
at the end of the week that person gained two extra hours of
production.
Of couse this efficient word processor is not easy to use so they go to
school for a week to learn it. (This was quite common a few years ago
when all secretaries had to go to school to learn to use the IBM
Displaywriters).
In 5 months you break even on the time lost for the school. And after
that you gain 5 weeks of extra production time per year. If you had a
group of people doing this, that means 10 people do the work that would
take 11 in another environment.
The world is full of tradeoffs. And in most instances that I have
seen the trade-off between user friendly and hard-to-learn is
efficiency and quite often computer-horsepower.
>Music
>systems are as complex as computers, but they've been made easy to use.
You must have a very poor computer. The music system I worked on had
over 3000 controls on it's front panel. The master tape recorder had
over 800 preliminary adjustments for proper setup.
Yet all it could do was play or record music. It couldn't control the
air conditioning, make coffee when I woke up, place a phone call to any
phone in the world in a matter of seconds or launch the space shuttle.
The computer is the only general purpose machine that has even been
invented. You can teach it to do anything. A hammer is good for
pounding nails and won't make coffee. Yet a friend of mine taught a
computer to nail boards together to assemble roof trusses. Try THAT on
your 'complex music system' :-)
--
Bill Vermillion - bi...@bilver.uucp OR bi...@bilver.oau.org
Tim Smith
>By the way, for those who dislike vi (like me) there is an alternative
>that has not been mentioned, perhaps because it is "common knowledge."
>It's an editor called "emacs." Many systems have it installed. Very
>nice editor!
There's another alternative, MicroEmacs, which is usually *smaller* than
vi, and can handle multiple files well, either in split windows, or by
easily flipping between buffers.
--Tim Smith
Sourcerer
: >I'm not sure you know just how much flamage you are inviting,
:
I'm not a Unix expert. This is just to second you about Word and
Wordperfect. Unless it's a capability specific to writing code, I haven't
read anything in this group about the capabilities of vi that can't be
done in a major DOS/Windows word processor -- or for that matter in a
dinky little thing like Transwrite for the Amiga. Nor does vi sound any
more capable than text editors on other platforms -- say Cygnus Ed on the
Amiga, which is very rich and capable and supremely easy to use even for
the loathed "end user".
In the case of Windows at least, the dislike of pointy/clicky GUI's is
misplaced since Windows is fully keyboard-only operational as are all well
written Windows applications (unlike Mac & Amiga). If you don't want to
use the mouse, then don't. The consistent interface makes learning a new
app much easier and gives newbies a shot at learning by actually doing
work rather than staring in frustration at some cryptic statement that
translates as "you can't do that. Nevermind why, or how you can".
But Unix is deeply satisfying to me for what I am sure are perverse
reasons. I just thank God my paycheck doesn't depend on using it.
--
(}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}})__
({{{{{{{{{{{{{{{{{{ At the back of the blue bus{{} /(**)\
({dje...@telerama.pgh.pa.us{{{{Sourcerer{{{{{{{{{{) \../
(}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}) ||
Sourcerer
: Computers do not exist to be used by programmers and system administrators.
Amen, Jeff, Amen. I do have these issues facing me in the real world,
with 70 users who are about to be transistioned off of an antique Wang
network onto something else, which I guarantee will not be UNIX based for
the reasons you have listed. The pain comes because I also recognize that
UNIX would handle the network more efficiently and flexibly than would DOS
but we can't live with the collapse in productivity we would experience.
Our issues are totally -- I mean absolutely -- opaque to the UNIX people
we've talked to. It is incredible. Consider the several hundred posts to
this group, which is purportedly here to discuss the needs of the "user",
subtract the "user? user? we doan nee no stinkin users!", the UNIX/VMS
comparisons, and the text editor worship, and what's left? The user
exiting stage right and purchasing MS Windows, that's what's left. Oh,
and UNIX sysadmin types wondering if Congress will extend unemployment
benefits.
Michael Lemke
Peter A. Bidian <pe...@aem.umn.edu> wrote:
>nic...@prz.tu-berlin.de (Juergen Nickelsen) writes:
>
>>In article <Oct08.225...@yuma.ACNS.ColoState.EDU>
>>kj...@typhoon.atmos.colostate.edu (Pat Fitzpatrick) writes:
>
>>> By the way, for those who dislike vi (like me) there is an alternative
>>> that has not been mentioned, perhaps because it is "common knowledge."
>>> It's an editor called "emacs." Many systems have it installed. Very
>>> nice editor!
>
>>[Disclaimer: I am one of those people who would like to change their
>>login shell to Emacs. I am able to work with vi, though.]
>
>>I would not consider Emacs a user-friendly editor for novice users.
>
Absolutley.
>
>How right you are!! Just yesterday I opened emacs for fun, and was impressed
>that it opened an extra window in X11, but then I was completely lost.
>I thought it was a great idea to have the tutorial, but I didn't find it
>very helpful. Sometimes it gives you command names but no Key sequence,
>or I have a problem understanding the way the tutorial works.
>
>I tried search, and by intuition I typed C-s which seamed to be the right stuff, but after I typed the string to be searched and hit return, I got
>a ^J in the control window and the cursor moved to the next line.
>
>And finally when I tried to get out... no luck. Kill window did the job then.
This is how I usually get out of vanilla emacs too. I perceive the exit
CTRL/xyz as an exit password. The problem with its help facility is you
need a manual for the help. Example: help splits the screen. Now you
need to know how to get rid of the second window. It tells you, I think
but that is gone of course as soon as you got help on something. Now you
are stuck. And the whole system with info etc is so complex I really
can't remember this. And I don't see why I have to. Help must be easy
to use and not require as much learning as the rest of the editor.
>For now I'll stay with vi. At least it works on all terminals, you don't
>need function keys (if you have to move from Sun to Iris to Mac terminals
>you know that relying on keys can be a big hazard).
The latter point is true of course. But if you look at ergonomic
issues, function keys, especially the keypad, are *much* better than
pressing two keys at the same time. Very tricky indeed with your coffee
cup in one hand. Ehm, make that phone not coffee cup.
>Had enough trouble
>to use vi after using the VMS editor.
>
>Just my poor expierence...
>
You are not alone. But you might try emacs -l tpu-edt. GOLD/E will
really do exit. And GOLD/KP 8 will fill your paragraph/select range.
If now emacs could highlight a region on the screen... (X doesn't
count!)
Michael
--
Michael Lemke
Institute of Astronomy, Cambridge UK
(mic...@io.as.utexas.edu or UTSPAN::UTADNX::IO::MICHAEL [SPAN])
John Hascall
}Jeff M Younker (je...@karazm.math.uh.edu) wrote:
}: Computers do not exist to be used by programmers and system administrators.
}: They exist so that other people can accomplish their jobs. They exist to
}: make things easier for the 'users'.
}Our issues are totally -- I mean absolutely -- opaque to the UNIX people
}we've talked to. It is incredible. Consider the several hundred posts to
}this group, which is purportedly here to discuss the needs of the "user",
}subtract the "user? user? we doan nee no stinkin users!", the UNIX/VMS
}comparisons, and the text editor worship, and what's left? The user
}exiting stage right and purchasing MS Windows, that's what's left. Oh,
}and UNIX sysadmin types wondering if Congress will extend unemployment
}benefits.
What makes you think we are all sysadmins? I am a programmer, a
programmer *is* a user. Unix is my tool. If my tool is not suitable
for your users, fine, use something else -- but kindly don't f*ck up
MY tool.
Peter A. Bidian
>>For now I'll stay with vi. At least it works on all terminals, you don't
>>need function keys (if you have to move from Sun to Iris to Mac terminals
>>you know that relying on keys can be a big hazard).
>The latter point is true of course. But if you look at ergonomic
>issues, function keys, especially the keypad, are *much* better than
>pressing two keys at the same time. Very tricky indeed with your coffee
>cup in one hand. Ehm, make that phone not coffee cup.
Not being american, I do not drive with a mug in a hand and I don't type with
one in my hand either. When I have to do lunch, I just read news... :-))))
What I was refering to, is the case where there are no function keys!
>>Had enough trouble
>>to use vi after using the VMS editor.
>>
>>Just my poor expierence...
>>
>You are not alone. But you might try emacs -l tpu-edt. GOLD/E will
Is good to hear it.
Peter
Sourcerer
: What makes you think we are all sysadmins?
If the shoe don't fit don't wear it.
: I am a programmer, a programmer *is* a user. Unix is my tool. If
: my tool is not suitable for your users, fine, use something
: else -- but kindly don't f*ck up MY tool.
What gave you the idea I'd do anything to your tool?
The intention of my posts to this newsgroup is to focus on the concerns of
the only computer users for whom "user-friendly" has a solid, crystal
clear and urgent meaning, i.e., the end-users of apps and the people who
employ them. I have indicated elsewhere that Unix meets the needs of
programmers and sysadmins very well, perhaps better than any OS, and for
the small use I make of it, Unix satisfies me as well.
The general attitude expressed on this subject in this group is that
(non-programmer) end-users are lazy, computer illiterate bottom-feeding
wankers sucking on trashy GUI tits. This is not helpful.
If I interview someone for an administrative position and ask the question:
We have an annual turnover of "x" number of employees in entry level
postions who need to be quickly brought up to competence in using our
network. How would you assist the administration group in facilitating this?
And the answer I get is:
"Duhhoiii"?
I will not hire that person.
Do not doubt the question is asked.
Tim Maroney
(Andy Newman) writes:
>"User Friendly" is a stupid phrase. It is totally subjective as it depends
>on the user. I'm a programmer and I use UNIX. I think UNIX is friendly
>therefore UNIX is user friendly.
You're quite correct. UNIX is designed for friendliness to a
particular user community, and it achieves those goals quite well. It
makes those users happy and content, it allows them to do what they
want to do, and it makes them feel good about themselves for having
mastered its subtleties and intricacies.
However, few people fit the criteria for membership in that user
community or would ever wish to. In an age when computers are copming
to every desk, every home, every counter, and every briefcase, UNIX is
destined to grow in absolute numbers and to shrink in terms of relative
market share. It may live forever as a combination hobbyist system and
guru-maintained server architecture, but it will become relatively less
and less significant as time goes on, compared to computer systems that
appeal to a wider audience.
--
Tim Maroney, Communications and User Interface Engineer
{apple!sun}!hoptoad!tim, t...@toad.com
"The above opinions and suggestions have absolutely nothing to do with
the little fat man putting crisp $100 bills in my pocket."
-- Alan Vymetalik
Tim Maroney
>>"User Friendly" is a stupid phrase. It is totally subjective as it depends
>>on the user. I'm a programmer and I use UNIX. I think UNIX is friendly
>>therefore UNIX is user friendly.
In article <290igq...@charnel.ecst.csuchico.edu> td...@ecst.csuchico.edu
(T.Dunn) writes:
>You're not keeping up on the jargon. The term "user friendly" isn't
>something that we just made up for this group. It has a specific
>meaning, and doing a semantic breakdown of it isn't going to change
>the meaning of the term. Go look in a magazine post 1990 if you still
>don't understand the generally accepted meaning of the term.
I'd like to see some citation to establish such a claim. So far as I
can tell as an experienced usability engineer, Andy is using the term
correctly. Usability engineering attempts to define the
characteristics of the users of a system and maximize its usability by
that user community. UNIX was developed by programmers for programmers
and does very well in achieving programmer-friendliness.
It is thoroughly unsuited to any other class of users and would require
such extensive modification for non-programmers that it would no longer
be recognizable as UNIX. It would also be incapable of running the
millions of lines of UNIX software which have been written by
programmers for programmers. Face it, any operating system which
requires you to change #defines and run a C compiler just to install a
software package is not for anyone except programmers, and never will be.
--
Tim Maroney, Communications and User Interface Engineer
{apple!sun}!hoptoad!tim, t...@toad.com
"I was brought up in the other service; but I knew from the first that the
Devil was my natural master and captain and friend. I saw that he was in
the right, and that the world cringed to his conqueror only from fear."
- Shaw, "The Devil's Disciple"
John Hascall
}The intention of my posts to this newsgroup is to focus on the concerns of
}the only computer users for whom "user-friendly" has a solid, crystal
}clear and urgent meaning, i.e., the end-users of apps and the people who
}employ them.
I don't understand this. For your typical "application user",
the O.S. is barely relevant. For example:
% sas
how is that less friendly than:
$ sas
or C:\> sas
???
}The general attitude expressed on this subject in this group is that
}(non-programmer) end-users are lazy, computer illiterate bottom-feeding
}wankers sucking on trashy GUI tits. This is not helpful.
No, *my* attitude is: we can teach almost anyone here what they
need to know to get their work done on our Unix system in 3 hours,
why can't you<1>.
}If I interview someone for an administrative position and ask the question:
}We have an annual turnover of "x" number of employees in entry level
}postions who need to be quickly brought up to competence in using our
}network. How would you assist the administration group in facilitating this?
}And the answer I get is:
}"Duhhoiii"?
}I will not hire that person. Do not doubt the question is asked.
How about:
``Good bye, thank you for your time. I'm sorry, but I won't work for
a place where they "facilitate", or "empower" or other
management-by-buzzword''
--------
<1> That's a generic, rather than personal "you".
John Lovell
>
..text deleted...
>
>And finally when I tried to get out... no luck. Kill window did the job then.
>For now I'll stay with vi. At least it works on all terminals, you don't
>need function keys (if you have to move from Sun to Iris to Mac terminals
>you know that relying on keys can be a big hazard). Had enough trouble
>to use vi after using the VMS editor.
>
>Just my poor expierence...
>
>Peter
Yes, suffering from the same transition, the simplest thing I found was to just install EDT on the Sun/Iris/etc. It's expensive (about $1000, from various companies), but at least it's painless.
John
Peter da Silva
> I'm not a Unix expert. This is just to second you about Word and
> Wordperfect. Unless it's a capability specific to writing code, I haven't
> read anything in this group about the capabilities of vi that can't be
> done in a major DOS/Windows word processor
Oh, cool. How do I sort a list of names and addresses in Wordperfect. Can
I do it in one line? It's not just that you can do these things. They're
also easy to do.
Oh, by the way, I use Elvis on my Amiga. It's a VI clone.
> app much easier and gives newbies a shot at learning by actually doing
> work rather than staring in frustration at some cryptic statement that
> translates as "you can't do that. Nevermind why, or how you can".
On Windows, DOS, Mac, Amiga, I get a lot of 'you can't do that'. How about
looking at the contents of a file that's in the process of being downloaded?
On the Amiga, it's locked. On DOS, the last time I tried that I screwed
something up and lost the file. Sure, you could modify the modem program to
let you do it. But that is a bit beyond the naive user.
Peter da Silva
> Some may argue that their are other reasons for choosing a computer system
> for company, and their are. But in the future, these technical issues are
> going to vanish into woodwork. All the things that gave unix it's
> technical edge over the years are becoming standard.
Cool. When it's got all the features of UNIX it'll effectively be UNIX. I
can't wait. I'm sick of running into 'you can't get there from here'.
Peter da Silva
> Face it, any operating system which
> requires you to change #defines and run a C compiler just to install a
> software package is not for anyone except programmers, and never will be.
The operating system doesn't do that. The application programmer does. There
are plenty of canned applications for System V that just slide in. Sun
software seems a bit less well put together.
And even if you have to run the compiler to do the install, that's not a
disaster. I do a bit of support for a local BBS, and I just installed a
new version of the MUD client, tinyfugue. I unpacked it, typed "make"
and it came up working. Autoconfiguration and everything, even on a SCO
system (which is about as oddball as you get these days).
Sourcerer
: % sas
: ???
: How about:
Ah, well.
I suspect that those of us who subscribed to this group with some hope
that it would address the concerns of the users for whom 'friendliness' has
some meaning have found each other and will manage to discuss those
concerns via e-mail in the context of getting through the business day.
Thereby leaving you (pl) to wrestle with the great issue: whether
Spiderman is greater than Batman -- er, whether emacs is greater than vi.
Lucio Chiappetti
|> In article <293rui$k...@inxs.concert.net> mor...@rock.concert.net (Morgan J Ryan -- Neil Patterson Publishers) writes:
|> >Could an oldtime vi user really do a complex editing job faster than me in
|> >Word or Wordperfect or any of those suboptimals? Is there any function in
|> >vi that I don't have in Word?
|>
I use Word for production of documents, and either dxnotepad or
EVE for source code writing. Word is more suited for the former,
and yes, it's true that dxnotepad (quite similar to textedit) is
not so good for repetitive operations.
Old time command-line-mode-like editors like VMS EVE or EDT, or IBM VM/CMS
XEDIT are definitely better for that.
|>
|> >But mine I can walk right up to and use. If I walk right up to
|> >vi and start sawing, ho boy, call an ambulance. Ease of use seems to me a
|>
I agree, vi is a nightmare.
The very fact that whatever you type on the keyboard are commands
and not things which go into your text makes it horrible.
I refused to learn it but the two or three commands I need for
csh command-line editing. I use the dxnotepad/textedit stuff
although they are not optimal, I can go along with them, or do
my editing on a Vax with EVE or an IBM with XEDIT (I've even
installed an XEDIT clone on Unix).
These feeling are shared by all my colleagues (astronomers)
coming from the IBM or VAX world.
|> > I'm wondering why unix isn't easy to use already? Ease of use
|> >happens naturally with everything else. But not unix. Why? Morgan
|>
|> A computer is different than everything else in the world. A car, a
|> toaster, these are products. They were designed and built with a set
|> of tools that engineers and manufacturers use. These tools are complex,
|> varied, and obscure. But you never see the tools, only the product.
Yes, but IBM mainframes, VAXes etc. are computers too !
And they provide EDITORS which are much more USER-FRIENDLIER
(and equally powerful) than vi.
vi might have been great in 1969 when Unix was invented and all
other were editing on a teletype or card punch.
But IBM was providing XEDIT in line mode at least in 1979, and in
full screen mode soon afterwards, and DEC was providing EDI on PDP
in line mode and EDT in full screen at the same time, and HP was
providing EDIT/1000 in the same frame of time, and ALL OF THEM ARE
A GREAT STEP FORWARD in friendliness then vi.
Why did Unix (and in particular vi) NOT make the same leap forward
that those other systems did (without of course loosing the good
things it has, pipes, i/o redirection etc.) ?
VI : the VIsceral editor (one that induces GUT feelings, of hatred or
love, in those who use it)
--
-----------------------------------------------------------------------------
A member of G.ASS : Group for Astronomical Software Support
-----------------------------------------------------------------------------
Lucio Chiappetti - IFCTR/CNR Milano | U N I C U I Q U E
via Bassini 15 - I-20133 Milano - Italy | System Manager
Internet: LU...@IFCTR.MI.CNR.IT | FORTUNAE SUAE
Decnet: IFCTR::LUCIO (39610::LUCIO) |
Bitnet: <please do not use any more> |
-----------------------------------------------------------------------------
John Hascall
[...quotes my entire article and ignores it (one or the other please)...]
}Ah, well.
}I suspect that those of us who subscribed to this group with some hope
}that it would address the concerns of the users for whom 'friendliness' has
}some meaning have found each other and will manage to discuss those
}concerns via e-mail in the context of getting through the business day.
Are you the pot or the kettle? I've posted at least 3 examples of how
we've made our Unix system friendlier (ezemacs, easyVincent, tcsh);
all I've seen from you is bitching...
John ``don't work no business day'' Hascall
Peter da Silva
> I suspect that those of us who subscribed to this group with some hope
> that it would address the concerns of the users for whom 'friendliness' has
> some meaning have found each other and will manage to discuss those
> concerns via e-mail in the context of getting through the business day.
Well now, I've made a few suggestions as to how to make the system friendlier.
Both in this group and via email. Stuff that I have found useful.
But you have to understand that your (generic you) sniping at the UNIX fans is
going to put them on the defensive. How about instead of targeting the folks
who respond with "who needs readable options" and beating on them, you target
the folks suggesting ways to *get* readable options? Same for anything else.
My big gripe is error messages. The gods know 'perror' is pretty minimal, but
at least it gives you the hook to find out what went wrong. Messages like
"Can't open input file" are useless.
This isn't a UNIX problem, by the way. Even VMS programs (and VMS generally
does a better job of this than most) will give you high level error state
only, when what you really need is a trace. Mac and Windows programs are
pretty bad about this: when something gets messed up your options are to
locate a guru who can grovel around in PIF files or resource forks, or to
reinstall from scratch.
Craig Bergren
> I can go along with them, or do
> my editing on a Vax with EVE or an IBM with XEDIT (I've even
> installed an XEDIT clone on Unix).
Great, see Unix is user friendly. Lucio can install one of the worst
editors known to man on his Unix system. Maybe you can get HP to give
you the source for the second worst editor, EDIT/1000. :-)
> Yes, but IBM mainframes, VAXes etc. are computers too !
> And they provide EDITORS which are much more USER-FRIENDLIER
> (and equally powerful) than vi.
> vi might have been great in 1969 when Unix was invented and all
> other were editing on a teletype or card punch.
Your history is a little flawed here, vi is a screen editor invented in
Berkeley after they started using VDTs. ed was the first line editor on
Unix (circa 1969), ex was a Bezerkely invention as was vi, ex is the
line mode for vi (when you use the : vi command it prompts you for an
ex command).
> But IBM was providing XEDIT in line mode at least in 1979, and in
> full screen mode soon afterwards, and DEC was providing EDI on PDP
> in line mode and EDT in full screen at the same time, and HP was
> providing EDIT/1000 in the same frame of time, and ALL OF THEM ARE
> A GREAT STEP FORWARD in friendliness then vi.
You've brought back some really dark memories of life in Big Iron. I've
used all the editors you mention and come to exactly the opposite
conclusion. I even run vi on my PC.
The first time I used XEDIT was after I matriculated to the real world.
Before then I card punches, line editors, and full screen editors on CDC
Cyber systems (some NOS editor), a line editor which came with a
time-sharing system for MVT (milton/wylbur) and a screen editor derived
from super wylbur, the MVS successor to wylbur (I think milton/wylbur
came from University of Waterloo). The IBM site I worked in used
milton/wylbur on MVT which TSO wouldn't run on and ran super wylbur on
MVS because it wasn't as big of a pig as TSO. It also had a better
editor.
After matriculation, I was exposed to EDT on RSX/11 and VAX, a bunch of
other editors for Motorola development systems (Exormacs), the Xerox
Sigma 9 line editor and later XEDIT for sending and receiving mail. For
the first few years I worked primarily on HP 1000s using EDIT/1000 and
HP instrument controllers using the UCSD P-system editor. The XEDIT and
Sigma 9 editors were to access a corporate mail system. I've also used
a pee cee in and out of Windows (2.x and 3.x). Now I work almost
exclusively in Unix with vi.
Earlier this year I even tried emacs for a week. For those who like
mode-less editors, emacs is not for you.
Of all these editors, the notepad editor in Windows was the easiest to
use and the most useless for doing real work (it doesn't handle very
large files). The worst was any IBM line editor (take your pick
XEDIT/CMS, EDIT/TSO, or EDLIN/DOS).
All the line editors are pretty cryptic. The best line editor (easiest
to use) was the one in wylbur and super wylbur the one in NOS wasn't too
bad either. I think because wylbur's line editor was the least cryptic
and equals XEDIT in power. The most powerful editors were on DEC
systems (RSX/11, VMS, Unix). ex/ed in Unix pick one depending on your
preference. tico on RSX/11 looked as powerful, but I didn't work with
RSX/11 long enough to learn it.
As for screen editors, EDT and emacs I find totally confusing because of
all the $%&QQ! special function keys that only exist on certain
keyboards (where do I find the EDT "gold" key on an IBM terminal) or
exist on no keyboard (emacs nonsense control sequences: ^h means help
but ^v means move up, or is it down). XEDIT/CMS and EDIT/TSO are almost
identical and are mediocre from a user friendly standpoint, but can be
made to do powerful things when using macros. EDIT/1000 is brain dead.
> These feeling are shared by all my colleagues (astronomers)
> coming from the IBM or VAX world.
All my PDP/11 and VAX friends wish Unix had tico. I'll bet someone has
ported tico to Unix. Which would again prove that Unix is user
friendly.
> Why did Unix (and in particular vi) NOT make the same leap forward
> that those other systems did (without of course loosing the good
> things it has, pipes, i/o redirection etc.) ?
IF THEY'RE SO GOOD ....
Why do you think HP doesn't sell EDIT/1000 on HP-UX?
--
CB
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
! The views I hold are not mine, nor those of my employer. !
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
John Gibson
> I agree, vi is a nightmare.
> The very fact that whatever you type on the keyboard are commands
> and not things which go into your text makes it horrible.
etc.
If all your complaints about Unix are complaints about vi, I have good news
for you: vi is not Unix. There are other editors available, emacs, for example,
and word processors, such as WordPerfect. Emacs is not subject to the complaint
you raise about vi, it's free, it's extremely powerful, and at least where I
come from, it, not vi, is considered the standard Unix editor. Odds are 100 to
1 that it's installed on your system.
Give emacs a try. Now let's get on with discussing Unix.
John
--
__________________________________________________________________
gib...@gorn.tam.cornell.edu
Theoretical and Applied Mechanics (607) 255-4077 office
Cornell University, Ithaca NY 14853 (607) 273-9189 home
------------------------------------------------------------------
Operator
--
In article <2934u3...@chaplin.cs.ubc.ca>, lal...@cs.ubc.ca (Paul Lalonde) writes:
|>
|> I'm not sure you know just how much flamage you are inviting, but I
|> think I'll actually try to tell you why I like vi, and maybe some of
|> the reasons why you don't.
|> In comp.unix.user-friendly you write:
|> > I ransacked the net looking for a real word processor. I knew there
|> >had to be one. Finally I turned up pico. It ain't much, but at least it
|> >has linewrapping. At least when I try to correct a typo it doesn't blow my
|> >text to lowercase confetti.
|>
|> vi has line wrapping - the command is :set wrapmargin=8
|> (change the 8 to the left margin size you want)
|> To re-format a paragraph try being at the start of the paragraph
|> and typeing !}fmt followed by return.
In emacs yu can use Esc-g or Esc-q .
|>
|> > So I ask you. Why do unix pros use an editor that was clunky in
|> >1971? Is vi a test? Beat vi and you can learn the rest?
|>
|> Because it's not clunky :-) The catch is that if you expect to use
|> it like most people use word processors it's not going to work for you.
|> Vi requires a bit of a different mind set. You don't cruise around the
|> text with the arrow keys for more than a few strokes in a row. Rather,
|> you search for a string using the / command, which brings you right
|> there. You don't dig around everywhere for a function, you use tags
|> (man ctags for more details). You don't delete each character, but
These capabilities are available in emacs. Plus emacs does not have the
obnoxious feature, that if you realize you made a typo 2 characters earlier
you are forced to change mode 3 times to fix it.
|> rather the whole word with the dw command. You change words by using
|> the cw command instead of deleting and re-typing. The list continues,
|> but the deal is that you have to learn how vi does things before you
|> become productive with it. But once you know them you can go like
|> greased lightning. Challenge an old-time vi user to a race on some
|> editing jobs. Vi will win in their hand ahead of you on a WP or a
|> point and click nearly every time.
Bullshit, an emacs guru will beat the shit out of any vi user. Plus emacs
has on-line help, allows you to edit rectangles and files with vey long
lines such as gene sequences.
|>
|> > How *do* I get plaintext man pages?
|>
|> nroff -man /usr/man/man1/man.1
|> (substitute the path and name of the manpage file for the man.1 example)
|>
Actually, a better way of doing it would be in a shell buffer in emacs, then you
can search it for key words.
I think Computer Companies should offer emacs , jove, tcsh on their computers.
Then if 1 likes they always ftp from a remote site and install the standard
shit that's usually given (vi, csh, sh)
Paul, tu es le Roi des Imbeciles. vi est 1 programme qui vaux 1 kilo de merde.
peter da silva
> And they provide EDITORS which are much more USER-FRIENDLIER
> (and equally powerful) than vi.
I work on VAX/VMS and UNIX both. I don't find EDT any better than vi, in
fact back when I was working on RSX I used a version of ex (the line-mode
subset of vi) that I wrote myself in preference to EDT.
EDT macros, and definitely TPU macros, are way more powerful than the poor
macro facilities of vi, but once you start using regular expressions for
your repetitive operations you don't want to mess around in macros any
more.
This is the basic reason for vi's dominance: it's got this facility that
other editors lack. This is due to its line-mode origins: in line mode you
generally use regular expressions for most editing (as opposed to text
entry), and once you're familiar with them they're incredible tools. Just
bend your cerebellum the right way and you'll never go back... it only
hurts for a little while.
> Why did Unix (and in particular vi) NOT make the same leap forward
> that those other systems did (without of course loosing the good
> things it has, pipes, i/o redirection etc.) ?
Regular expressions.
--
Peter da Silva `-_-'
Network Management Technology Incorporated 'U`
1601 Industrial Blvd. Sugar Land, TX 77478 USA
+1 713 274 5180 "Ja' abracas-te o tey lobo, hoje?"
Lezz Giles
I just want to tell everybody that you've all got it wrong - the greatest
editor ever created was called ecce!
ecce was the Edinburgh Context Compatible Editor. It was created at
Edinburgh University and when I used it it was a line-editor in their
home-grown operating system for ICL mainframes (EMAS). At a superficial
level it was similar to ed, but it had a full language which
was powerful enough to be turing-machine-equivalent. It was very small
(originally written for a PDP-11 with a ridiculously small amount of
memory), fast, and easy to use - there wasn't much you couldn't do easily.
Vecce was a visual, screen-orientated version of ecce.
I did port ecce to C/Unix in my early years of employment after University,
but sadly I succumbed to the standard Unix environment and started using
vi and emacs... :-(
So anyway everybody should just remember that all those other editors are
just mere bit-twiddlers compared to the power and the glory of ecce. If
you can't write a tower of hanoi program in the normal command language
then it ain't a man's editor!
Lezz "what was this thread about again....?" Giles
F. Scott Ophof
..[that IBM-XEDIT is the worst editor (cryptic)]..
>Lucio Chiappetti (lu...@ifctr.mi.cnr.it) wrote:
..[that he likes IBM-XEDIT a lot]..
>I've used all the editors you mention and come to exactly the opposite
>conclusion. I even run vi on my PC.
..[much appreciated history lesson]..
Craig, what exactly is it about IBM-XEDIT that you find so cryptic,
and how do you mean that? Lucio, could you explain what it is about
IBM-XEDIT that makes you feel it's such a good editor?
If such things are explained in more detail, maybe then we can start
finding out what makes an editor "good" resp. "bad".
Maybe we'll discover that it's very much a question of what a person
thinks an editor should (not) do.
BTW, maybe this also has to do with editing, at least it does relate
to ease-of-reading; the following paragraph is formatted so that the
right edge is (almost) straight. That in itself may look nice, but
the extra spaces between words makes it quite hard for me to read.
Personally I seem to prefer the "form follows function" mindset, and
prefer to reduce "useless gadgets & chrome" to the minimum.
>Of all these editors, the notepad editor in Windows was the easiest to
>use and the most useless for doing real work (it doesn't handle very
>large files). The worst was any IBM line editor (take your pick
>XEDIT/CMS, EDIT/TSO, or EDLIN/DOS).
Let me say that I find CMS-XEDIT the (close-to-)ideal editor for
*me*, because the commands are plain words that do what they imply.
The easy-to-learn/use macro language (REXX) makes designing macros
fun, and I can still understand them years later (without having to
refer to the REXX or XEDIT manuals).
The "vi" editor is one I find difficult to learn/use because the
commands are NOT words, and even their "names" often do not imply
what they do. I quoted the word "names" because of the fact that
they are linked directly to a single keystroke. This also means
that any other product that links functions directly to keystrokes
is difficult for me to learn/use (especially when the meaning of a
keystroke changes from product to product). This may also be why I
have no interest in learning to use a mouse, or any system based on
the use of a mouse.
I wouldn't be surprised if the above paragraph contains much of what
Lucio wanted to say.
BTW, Unix may be user-friendly in that it allows you to install
(almost) any product one wants, but there's a bad side; one needs to
be able to understand what compiling is all about, because most (at
least those I've tried) installation manuals of freeware products on
Unix are written by and for people who are cognizant with compilers.
This means that the authors take much pre-knowledge for granted on
the part of the installer, much more than they could ever begin to
realize! This doesn't make Unix itself unfriendly, but when one
takes into account the types of replies to installation questions
from novices, then the Unix people become a rather unfriendly lot.
The Unix people become exasperated with "silly questions which THEY
could look up the answer to in no time flat", not realizing that
they are experts (relatively seen), and the novices are NOT.
Heck, often the novice (or less-experienced) user doesn't even know
where to *start* looking, for the moment assuming that the relevant
help file *is* available (which it quite often isn't).
Add to that the fact that Unix seems to be more behept with phrases
that mean special things not used in day-to-day language, and it is
no wonder to me that Unix is called "unfriendly". Any idea how many
people don't even know what a "source file" is?
Back to the main subject.
>> Why did Unix (and in particular vi) NOT make the same leap forward
>> that those other systems did (without of course loosing the good
>> things it has, pipes, i/o redirection etc.) ?
>IF THEY'RE SO GOOD ....
Another good example of an area where Unix did NOT make any real
advance is in the area of pipes; IBM's CMS-Pipelines is definitely
an improvement on the basic concept of pipes.
One thing may be of importance in the Unix world, and much more so
than in the others (especially the IBM mainframe world), and that is
that most IBM mainframe software costs money and is meant for use in
the commercial world. In the commercial world there simply is much
less time, incentive, and interest to spend more than the absolute
minimum effort on learning a product.
The academic world however is by definition an environment where one
learns new things. And Unix is used there the way IBM mainframes
are used in the commercial world.
Where people use computers as a tool to simply DO things (accounting
and such stuff), one will find the accent on the experience and
knowledge of the users, NOT on the tools. An example is the
accounting departments of universities. Research on the other hand
(be it at academic sites or commercial ones) will often see Unix in
use.
Regards.
$$\
F. Scott Ophof
>If all your complaints about Unix are complaints about vi, I have good news
>for you: vi is not Unix.
>you raise about vi, it's free, it's extremely powerful, and at least where I
>Give emacs a try. Now let's get on with discussing Unix.
But "vi" is part of the Unix environment for most people, and as
such definitely discussable. For example, pressure to rewrite the
relevant man pages into something more useful for ALL types of users
would be a Good Thing.
That a product is free is not a valid reason in my book. But the
quality of the help file *is*. Nor is power, if the basic learning
curve is steep enough to make it difficult to learn for this or that
user type. As I said elsewhere, if one has no compiling knowledge
and the relevant support personnel is not friendly enough (which
includes the time to do it), then one is up the creek without the
paddle, right?
Regards.
$$\
F. Scott Ophof
>In article <2934u3...@chaplin.cs.ubc.ca>, lal...@cs.ubc.ca (Paul Lalonde) writes:
>|> vi has line wrapping - the command is :set wrapmargin=8
>|> (change the 8 to the left margin size you want)
>|> To re-format a paragraph try being at the start of the paragraph
>|> and typeing !}fmt followed by return.
That's a good example. I can't remember any section of the "vi" man
page which is entitled "Paragraph commands" or such. Nor is there a
section called "Formatting".
Looking closer at the command, which seems to be:
!}fmt
it might be that this is an external command (due to the "!") and
not part of "vi" at all. So you're asking of the user that heesh
know about commands outside the editor itself, in this case a
command which many users will feel to be one which belongs in the
editor, not outside...
Thanks for showing a prime example of why Unix is called
"unfriendly". (grin)
>|> Because it's not clunky :-) The catch is that if you expect to use
>|> it like most people use word processors it's not going to work for you.
>|> Vi requires a bit of a different mind set. [...]
How right you are!
>|> but the deal is that you have to learn how vi does things before you
>|> become productive with it.
Again, how true! However, the problem is that the information
available online to the user is not conducive to easy reading, let
alone understanding. Also, it does NOT explain the "vi" mindset.
To be told to "buy book XYZ" is imho a close-to-sadistic comment,
especially when those books also do not explain the mindset of "vi".
The "vi" mindset was explained to me in 5 minutes, and did make
sense to me back then. Sadly enough, it didn't remain in-memory, so
now I've got to do it the hard way... :-(
>|> > How *do* I get plaintext man pages?
>|> nroff -man /usr/man/man1/man.1
>|> (substitute the path and name of the manpage file for the man.1 example)
Isn't this a rather ridiculous thing to have to do? Especially when
the user might not even know what path to substitute? And how far
back into the path you show to substitute? Why assume that the user
knows where to find the files? A bit more verbosity might be quite
useful, and even reduce the number of times the same question is
repeated in the future...
>Paul, tu es le Roi des Imbeciles. vi est 1 programme qui vaux 1 kilo de merde.
Ivan, la meme pour le mentalite d'Unix. Pauvre gens qui doivent
l'utiliser... ;->
A bientot.
$$\
Joshua Geller
> EDT macros, and definitely TPU macros, are way more powerful than the poor
> macro facilities of vi, but once you start using regular expressions for
> your repetitive operations you don't want to mess around in macros any
> more.
> This is the basic reason for vi's dominance: it's got this facility that
> other editors lack. This is due to its line-mode origins: in line mode you
> generally use regular expressions for most editing (as opposed to text
> entry), and once you're familiar with them they're incredible tools. Just
> bend your cerebellum the right way and you'll never go back... it only
> hurts for a little while.
> Regular expressions.
emacs lacks regular expressions?
news to me.
josh
Kjetil Torgrim Homme
>> How *do* I get plaintext man pages?
> nroff -man /usr/man/man1/man.1
> (substitute the path and name of the manpage file for the man.1 example)
F. Scott Ophof:
> Isn't this a rather ridiculous thing to have to do? Especially when
> the user might not even know what path to substitute? [...]
The solution doesn't even work, the file will contain lots of ^H to
do bold face etc. The same result can be had by writing
man command > file
To remove underlining and bold:
man command | sed 's/.^H//g' > file
(use Ctrl-V Ctrl-H to type in the ^H)
However - why do you want plaintext man pages? If you want to read
them on the computer, use the man-program, if you want to print it
out, use the -t switch (for "typeset" :-) to man.
Oh, and if you see a "#"-prompt some day, do:
# cp /usr/bin/less /usr/ucb/more
;-)
Kjetil T.
Morgan J Ryan -- Neil Patterson Publishers
>However - why do you want plaintext man pages? If you want to read
>them on the computer, use the man-program, if you want to print it
>out, use the -t switch (for "typeset" :-) to man.
The question is addressed to me. I have a dialup account. I found that
flashing man pages into my eyes wasn't getting me any better at using vi.
My solution was to download man pages into my laptop so I could stare at
them and wonder what they meant at leisure. But I couldn't figure out how
to do it. Someone emailed me with "man pages > file" and it worked just
fine. How was I supposed to figure out that simple command? Well, by
becoming an expert. I mean other than that. Oh. I use the > thing now to
learn other commands and programs, one at a time. Not vi though. I gave up
on vi a long time ago and switched to puny, friendly pico.
Lots of people have told me how wonderful vi is and why they love
it. And several people have told me how to get plaintext man pages.
Several different ways. Good old unix. Some didn't work. Ah, unix.
Lots of people have told me they can edit text far faster with vi
than with, say, a Mac program. I'm trying to understand what they mean.
To delete a block of text, I scratch over it with a mouse, bonk delete, go
on to the next thing. With vi, I seem to be required to change from mode
to othermode and then #$%:/d and then do that some more, et cetera. Is
that really faster? I don't need to be told that vi has great virtues. But
some of the virtues I've been told about seem a bit overstated.
It's clear unix has great tools. The tool I wanted was one that
let me type a letter from beginning to end without learning anything. I'm
willing to learn what I need to know, but I didn't feel the need to
relearn how to type. Obviously the tool I needed was pico. But the tool I
got was vi. That was when I had zero worldliness about unix. For a good
long time I thought unix was pretty unfriendly. To the people who say vi
is not unix, I say: longitudinal bisection of the hair. I logged on to
type a letter and logged off thinking vi was the damndest thing I ever saw.
Thank goodness unix people were so generous with assistance.
(Thanks Jane.)
To all the people who are inventing easy tools and easy shells,
the Gods keep you in hand. Morgan
Einar Indridason
>[lots of stuff on different editors deleted]
>So anyway everybody should just remember that all those other editors are
>just mere bit-twiddlers compared to the power and the glory of ecce. If
>you can't write a tower of hanoi program in the normal command language
>then it ain't a man's editor!
As far as I know, you can get towers of hanoi both for vi and emacs :-)
John Hascall
}le...@merlin.dev.cdx.mot.com (Lezz Giles) writes:
}>So anyway everybody should just remember that all those other editors are
}>just mere bit-twiddlers compared to the power and the glory of ecce. If
}>you can't write a tower of hanoi program in the normal command language
}>then it ain't a man's editor!
}As far as I know, you can get towers of hanoi both for vi and emacs :-)
I also have a copy of a turing machine done in vi macros
(pretty cool to watch, and almost as pointless as psychoanalize-pinhead...)
F. Scott Ophof
>However - why do you want plaintext man pages? If you want to read
>them on the computer, use the man-program, if you want to print it
>out, use the -t switch (for "typeset" :-) to man.
One reason (valid for me) is to download the file to my PC so I can
read the file comfortably with a pager or editor I feel comfy with
and has the functionality I expect from such a utility.
Then funny stuff like page headers/footers should preferably be
removed, and those funny backspace-underscore sequences and tabs
should also be removed (that being about the only way to make man
pages normally readable on a PC.
It's simply ridiculous to include page headers/footers in a file to
be displayed on-screen, when the page length of the file isn't the
same as the number of screen lines. These separators should be
inserted by whatever utility is used to actually PRINT the file, in
other words that sends the file to a REAL PRINTER.
>Oh, and if you see a "#"-prompt some day, do:
> # cp /usr/bin/less /usr/ucb/more
>;-)
Could you please explain this one (privately if you feel it's not
necessary to bug the whole audience)?
Regards.
$$\
Tim Smith
>Bullshit, an emacs guru will beat the shit out of any vi user. Plus emacs
>has on-line help, allows you to edit rectangles and files with vey long
>lines such as gene sequences.
The vi user will be done and out of the editor while the emacs user is
still waiting for emacs to load!
--Tim Smith
Budi Rahardjo
>On 13 Oct 1993 17:12:21 GMT John Gibson said:
>>If all your complaints about Unix are complaints about vi, I have good news
>>for you: vi is not Unix.
>>...
>>you raise about vi, it's free, it's extremely powerful, and at least where I
>>...
>>Give emacs a try. Now let's get on with discussing Unix.
>But "vi" is part of the Unix environment for most people, and as
>such definitely discussable.
Just like "edlin" on DOS ?
Look people, if you don't like `vi' or `emacs' get alternatives,
there's `joe', `pico', `ce', `fpted'. etc . Tons of good editors.
For a simple text editor ... use `pico'. (I still use `vi', even on
my pc :-)
-- budi
Bill Vermillion
> I agree, vi is a nightmare.
> The very fact that whatever you type on the keyboard are commands
> and not things which go into your text makes it horrible.
That is the differance between modal and non-modal programs.
The first computer driven word-processors that were succesful
made the typewriter to computer transition easy by beind
non-modal. Probably the first, or at least one of the most
popular (my computer useage doens't go back that far) was the
IBM Display writer.
Since most typists learned on a typewriter and KNEW that if you
type a letter or number key that key would appear on paper,
then the computer driven word-processors added extra keys for
formatting. A standard keyboard key was always document
input while the other did document manipulation - just like the
few document manipulation keys on a typewriter, shift and the
carriage return lever.
vi is very easy once you get used to the modal concept. You
are input mode most of the time. To get out you hit escape.
And now you are in command mode, and some of those commands put
you back into input mode.
Running vi in the 'novice' mode will show in the lower right
hand corner what mode you are in.
vi is just another document manipulation tool - that does not
try to be a word proccessor, and like any tool it is as easy or
as hard to use as the amount of effort you put into it.
The hardest transition for people first learning computers is
that neccesary for any program that is modal.
> I refused to learn it but the two or three commands I need for
^^^^^^^^^^^^^^^^ <- That's the heart of your problem :-)
But although I use vi all the time, you need to use
whatever tool suits YOU best. What I do doesn't mean a
damn thing. Something that too many people seem to
forget.
> Yes, but IBM mainframes, VAXes etc. are computers too !
> And they provide EDITORS which are much more USER-FRIENDLIER
> (and equally powerful) than vi.
And so were DOS and CPM and vi is so much better than the
editor provided with CPM - ed (no relation to the Unix ed) or
the supplied editor with DOS - edlin
Not all OS'es are supplied with an editor that is flexible an
powerful. I learned ed on Xenix because that is all there
were and the regular expression training made vi very powerful.
I learned ed on CPM because that's all there was. I learned
edlin on DOS because that's all there was.
When you bring up an OS for the first time you have to take
what is there.
--
Bill Vermillion - bi...@bilver.uucp OR bi...@bilver.oau.org
F. Scott Ophof
>In <931014.20...@MReXX-0.20> Op...@CS.UWindsor.Ca (F. Scott Ophof) writes:
>>But "vi" is part of the Unix environment for most people, and as
>>such definitely discussable.
The point is that for those who simply must make use of the software
available (editor or otherwise) and are not in a position to either
get any alternative at all, or something more decent (whatever that
may mean in their case), it behooves those of us in a position to
supply at least decently useable manuals, to do at least that.
Regards.
$$\
Lucio Chiappetti
|> In article <931013090...@poseidon.ifctr.mi.cnr.it> lu...@ifctr.mi.cnr.it (Lucio Chiappetti) writes:
|> > And they provide EDITORS which are much more USER-FRIENDLIER
|> > (and equally powerful) than vi.
|>
|> macro facilities of vi, but once you start using regular expressions for
|> your repetitive operations you don't want to mess around in macros any
|> more.
|>
editor. I found their usage worthwhile, although slightly cryptic.
The syntax resembled the ex one, but I found it easier to remember.
And the greatest thing of all was that you could do SE RE ON and
SE RE OFF to switch from using regular expressions not to use them.
I never caught up with doing these things in vi or ex or ed. I'll
either use the SET ARBCHAR facility in XC (a VM/CMS XEDIT clone),
or write a macro (until a fewe months ago, I felt it worth moving
the file to an IBM and editing it there ... now the IBM is no more)
|> This is the basic reason for vi's dominance: it's got this facility that
|> other editors lack. This is due to its line-mode origins: in line mode you
Hmm...
also EDT, EDIT/1000 >and< XEDIT started as line mode editors
(and the best of them retain such feature), so what ?
|> entry), and once you're familiar with them they're incredible tools. Just
|> bend your cerebellum the right way and you'll never go back... it only
^^^^
Exactly, BEND !
Kristian Koehntopp
> It's an unfortunate failing of Unix documentation that little or no
>activity documentation exists, especially stuff that's well integrated
>into the existing tool documentation system. Unix manpages do not fit
>everything; far too often these days they're being called on to do just
>that.
When I got my first UNIX-like System (a XENIX/386 from SCO) there
were two kinds of printed manuals that came with it. The first
set of manuals had the word "Reference" in their name. These usually
described tools, not activities, and their options. One could use
these manuals as a catalog of the building blocks for own tools and
scripts.
The other set of manuals had the word "Guide" in their name. These type
of manual usually described activities such as editing a text with vi,
writing a report-style document featuring the nroff/mm combo, compiling
linking and debugging a C program and so on.
Only the first set of manuals is available in the form of manual pages
on a normal (i.e. not AIX) system.
Kristian
--
Kristian Koehntopp, Harmsstrasse 98, 24114 Kiel, +49 431 676689
"I'm sorry, Dave. I'm afraid I can't post that."
DON
(or is it possible to) do a global change of case for a letter or a word
or a selected text under vi. I know for a letter or a word one may do
:b,es/case/CASE/g
but I feel there should exist a better way. Thanks for any reponse.
__________________________________________________________________________
Don Tang
Dept of Chemical Engineering, University of Rochester, Rochester, NY 14627