[comp.text] WYSIWYG = DIY

eykhout@kunivv1.sci.kun.nl (Victor Eijkhout) (08/09/89)

In article <14903@dartvax.Dartmouth.EDU> hugo@griggs (Peter Su) writes:
>I claim that WYSIWIG are overly concerned with form, and no concerned
>enough about with the logical operations that result in the form 
>that you want.

This is an essential point. The discussion so far has been mostly
on capabilities. Well let's grand that a virtuoso can do the same
things with a WYSIWIG system and with a mark-up language (TeX, troff).

Now how about if I am not the designer of the layout. 

Personally I feel that what I turn out is somewhat less execrable than
a lot of what I see, but I am dead sure that a professional designer
will make something that is still a whole lot better. I know, because
I have had the occasion to work with one a number of times.

How about this one: I come to this designer with a manual of which
I have already typed the first 40 pages, say that's 100 sections and
subsections, and she tells me 'Oh please do all your headings
in capitals'.

Or this one: I have keyed in a linear algebra course, hundreds of
exercises, and she says 'It would look nice if all your
exercises [that I did TeXbook style, first two lines indented]
were completely indented, with the number flush against the left
margin and a dotted line leading up to the first word'.

In both cases my texts were in TeX (with some provisory macros
so that I could at least print), and implementing those changes
took 5 minutes each.

Question: can someone tell me that with a wysiwig it is just
as easy to make a global design change?

Conjecture: wysiwig systems are for people who make their own
layout, and who have decided on the definitive layout
before they started keying in the text. This I think is a wrong
way of working. I think I have a right to say this, because I've
produced some 'master pieces of the printing art', and the design
was done by a pro, and only after I had finished the text.

Victor.

rh26@prism.gatech.EDU (Howard,Robert L) (08/11/89)

In article <387@kunivv1.sci.kun.nl> eykhout@wn2.UUCP (Victor Eijkhout) writes:
>In article <14903@dartvax.Dartmouth.EDU> hugo@griggs (Peter Su) writes:
>>I claim that WYSIWIG are overly concerned with form, and no concerned
>>enough about with the logical operations that result in the form 
>>that you want.

>This is an essential point. The discussion so far has been mostly
>on capabilities. Well let's grand that a virtuoso can do the same
>things with a WYSIWIG system and with a mark-up language (TeX, troff).

That seems fair.

>Now how about if I am not the designer of the layout. 

>How about this one: I come to this designer with a manual of which
>I have already typed the first 40 pages, say that's 100 sections and
>subsections, and she tells me 'Oh please do all your headings
>in capitals'.

[ Example 2 deleted ]

>In both cases my texts were in TeX (with some provisory macros
>so that I could at least print), and implementing those changes
>took 5 minutes each.

>Question: can someone tell me that with a wysiwig it is just
>as easy to make a global design change?

Yes, it is....if the person is smart enough to use styles (or style
sheets) in his document.

Using your first example this person would set up a style for all headings
and call it HEAD.  He makes HEAD do bold italics (well, why not :-).  Then 
every heading he types is formatted using the HEAD code.  It shows up on 
his screen in bold italics.  Now the designer really freaks and says to use 
caps.  All you need to do is go to wherever the HEAD style is defined, delete 
the bold and italic codes, and add the 'use all caps' code.  (Now I will 
admit you're in trouble if there is no such code but the point is that it is 
simple to do.)  What is the result?  In just a few seconds (depending on the 
speed of the computer) you have all the headings in caps AND you can SEE the 
change.

Sounds pretty good to me.

>Conjecture: wysiwig systems are for people who make their own
>layout, and who have decided on the definitive layout
>before they started keying in the text. This I think is a wrong
>way of working. I think I have a right to say this, because I've
>produced some 'master pieces of the printing art', and the design
>was done by a pro, and only after I had finished the text.

I think your conjecture is probably correct (for the most part) but 
my point is that you don't have to work that way if you are smart.
Why not enjoy seeing a reasonable facsimile of your document right
on the screen?

Just a few thoughts.
(They've worked for me. :-)

Robert

-- 
Robert L. Howard  (GTRI/STL/MSD)             (404) 421-7165
Georgia Institute of Technology, Atlanta Georgia, 30332
uucp:     ...!{allegra,amd,hplabs,ut-ngp}!gatech!prism!rh26
Internet: rh26@prism.gatech.edu

amanda@intercon.uu.net (Amanda Walker) (08/11/89)

[I haven't been following too much of this discussion, so please accept
my apologies if this is rehashing old ground]
 
As a number of people have pointed out, so-called "style sheets" can give
"WYSIWYG" text formatting systems many of the advantages of batch-style
formatters.  Whether or not the systems cease to be WYSIWYG at this point
is an interesting question in terminology, but there's another way of
looking at the issue which I find useful.

I often draw a parallel between text formatting systems and computer
languages.  In particular, I view things like TeX or LaTeX as document
compilers, and PageMaker or Interleaf as document interpreters.  The
feedback cycle is longer for the "compiler" style system, but it's easier
to get higher quality end results.  I think that the "dual-view" systems that
people have mentioned/wished for are the right way to go in the long run, just
as I think that good incremental compilers are the way to go for computer
languages.

--
Amanda Walker
InterCon Systems Corporation
--
amanda@intercon.uu.net    |    ...!uunet!intercon!amanda

pyoung@zaphod.axion.bt.co.uk (peter young) (08/11/89)

From article <387@kunivv1.sci.kun.nl>, by eykhout@kunivv1.sci.kun.nl (Victor Eijkhout):
> Question: can someone tell me that with a wysiwig it is just
> as easy to make a global design change?

Depends what you mean by WYSIWYG. My experience with FrameMaker leads
me to think that I could have made these changes even more quickly,
since the changes to type face, point size, indentation etc can be made
to all paragraphs with the same tag with one click of the mouse.
BTW I have no connection with Frame other than as a more than satisfied user.

Regards, Pete.
+---------------------------------------------------------------------+
| Pete Young                 pyoung@axion.bt.co.uk                    |
| British Telecom Research Labs,SSTF, Martlesham Heath IPSWICH IP5 7RE|
| Phone  +44 473 645054                                               |
+---------------------------------------------------------------------+
|       Old age and treachery will beat youth and skill every time.   |
+---------------------------------------------------------------------+

chris@mimsy.UUCP (Chris Torek) (08/11/89)

In article <1438@hydra.gatech.EDU> rh26@prism.gatech.EDU
(Howard,Robert L) writes:
[good points about using `style sheets' in IBM PC `word processors]

>Why not enjoy seeing a reasonable facsimile of your document right
>on the screen?

Mostly because, at the moment, that is not possible.  A reasonable
facsimile of the document would require a 2000x3000 pixel screen.
At lower resolutions, emulating the printed page produces something
which is sufficiently hard to read that I prefer not to look at it
while editing.
-- 
In-Real-Life: Chris Torek, Univ of MD Comp Sci Dept (+1 301 454 7163)
Domain:	chris@mimsy.umd.edu	Path:	uunet!mimsy!chris

eykhout@kunivv1.sci.kun.nl (Victor Eijkhout) (08/11/89)

In article <1438@hydra.gatech.EDU> rh26@prism.gatech.EDU (Howard,Robert L) writes:
>In article <387@kunivv1.sci.kun.nl> eykhout@wn2.UUCP (Victor Eijkhout) writes:
>caps.  All you need to do is go to wherever the HEAD style is defined, delete 
>the bold and italic codes, and add the 'use all caps' code.  (Now I will 
>admit you're in trouble if there is no such code but the point is that it is 
>simple to do.)  
 
But is it that simple? I know that by now I can program most everything
a designer asks in TeX, but the wysiwyg systems I've seen look like
somebody threw together a number of mechanisms until s/he was convinced
there were sufficient special cases to cover most of what someone
would want. There is hardly any programmability.

The dividing line is becoming thinner and thinner, with wysiwyg
systems getting style sheets, I admit.

Victor

dkelly@npiatl.UUCP (Dwight Kelly) (08/11/89)

In article <387@kunivv1.sci.kun.nl> eykhout@wn2.UUCP (Victor Eijkhout) writes:
>In article <14903@dartvax.Dartmouth.EDU> hugo@griggs (Peter Su) writes:
>>I claim that WYSIWIG are overly concerned with form, and no concerned
>>enough about with the logical operations that result in the form 
>>that you want.
>
>Question: can someone tell me that with a wysiwig it is just
>as easy to make a global design change?
>


Several WYSIWYG systems have implemented "style sheets" that define the
layout and look of a block of text.  By changing the style sheet you
can make a global change on all text using that sheet.  This is NOT anywhere
near as powerful as macros.  But it is a step in the right direction.

I have heard that Framemaker goes even farther towards global design, 
and still provides a screen preview and interactive layout. 

--
Dwight Kelly            UUCP: gatech!npiatl!dkelly
Director R&D            AT&T: (404) 962-7220
Network Publications, Inc    2 Pamplin Drive     Lawrenceville, GA  30245
             Publisher of "The Real Estate Book" nationwide!

t-rayc@microsoft.UUCP (Raymond Chen) (08/12/89)

In article <1438@hydra.gatech.EDU> rh26@prism.gatech.EDU (Howard,Robert L) writes:
>>Question: can someone tell me that with a wysiwig it is just
>>as easy to make a global design change?
>
>Yes, it is....if the person is smart enough to use styles (or style
>sheets) in his document.

The problem is that most people aren't smart enough to use styles.
Having worked in a university computer help center, I've seen some
amazingly ... "nonportable document conventions"... Since the system was
WYSIWYG, they figured that if it looked good on the screen (and on
the page), then their job was done.  Nevermind that they did such
amazingly inflexible things like using spaces to align tables or center
headings, inserting hard returns to force page breaks, even manually
typing the page number at the bottom of each page.  Imagine my grief
when they come in asking "I'd like to use font X instead of the
font Y I've been using so far.  But when I change the font, the formatting
goes really screwy."

The problem isn't whether WYSIWYG is "better" than batch processing.
The problem is that WYSIWYG makes ignorance acceptable.  (Followups
to sci.edu regarding the hip-ness of admitting that "Oh, I don't know 
anything about science/math/computers.") 

The capabilities of each are (to the extent that most people need) equivalent.
[Still, there are some things in TeX I have yet to see in a WYSIWYG
system.  Most of them involve carefully-calculated penalty and glue
settings to achieve a desired effect in a robust manner.]
--
Raymond Chen, mathematician by training		...!microsoft!t-rayc

edb@io.UUCP (Ed Blachman x4420) (08/12/89)

eykhout@kunivv1.sci.kun.nl (Victor Eijkhout) writes:

>In article <14903@dartvax.Dartmouth.EDU> hugo@griggs (Peter Su) writes:
>>I claim that WYSIWIG are overly concerned with form, and no concerned
>>enough about with the logical operations that result in the form 
>>that you want.

>How about this one: I come to this designer with a manual of which
>I have already typed the first 40 pages, say that's 100 sections and
>subsections, and she tells me 'Oh please do all your headings
>in capitals'.

>Or this one: I have keyed in a linear algebra course, hundreds of
>exercises, and she says 'It would look nice if all your
>exercises [that I did TeXbook style, first two lines indented]
>were completely indented, with the number flush against the left
>margin and a dotted line leading up to the first word'.

>In both cases my texts were in TeX (with some provisory macros
>so that I could at least print), and implementing those changes
>took 5 minutes each.

>Question: can someone tell me that with a wysiwig it is just
>as easy to make a global design change?

Ok, I know that tooting one's company's own horn is kind of frowned on
 hereabouts, but *someone* should say it: it all depends on the WYSIWYG
 system you're talking about.  Some WYSIWYG systems *are* well set up to
 allow you to make global changes of the kind you just described.  My
 personal experience with such systems has been that I have found them
 easier to learn than markup oriented systems like TeX.

Does this mean that everyone will find them easier to learn and to use?
 Of course not; people's mileage varies.  But on the WYSIWYG systems
 I've used, 5 minutes seems like about the right amount of time to make
 the changes you described.

>Conjecture: wysiwig systems are for people who make their own
>layout, and who have decided on the definitive layout
>before they started keying in the text. This I think is a wrong
>way of working. I think I have a right to say this, because I've
>produced some 'master pieces of the printing art', and the design
>was done by a pro, and only after I had finished the text.

I think the key point of WYSIWYG systems is the direct feedback.  Wanna
 know what your pages will look like if you narrow your column-width
 slightly, or go from one-column to two-column design, or choose a dif-
 ferent typeface?  In a good WYSIWYG system there's essentially no syn-
 tax to learn to make the change, and you can immdeiately see whether
 the effect is pleasing or not.  That feels to me like an easier way
 to validate a design than the trouble of twiddling markup, followed by
 proofing a document (either to a previewer or to paper).

As for design by pros vs. design by amateurs: I guess I agree that pros
 have a lot to contribute to design -- certainly the task of developing
 an effective and pleasing layout is not a simple one.  But that's inde-
 pendent of the implementation of that design.  I'm not a designer --
 but if I were, I think I'd find it easier to deal with a WYSIWYG sys-
 tem than a markup system, as the former would mean that I could concen-
 trate on design, without the cognitive load of having to learn a markup
 language.

>Victor.

Ed Blachman		edb@ileaf.com	(or)	...!mit-eddie!ileaf!edb

les@chinet.chi.il.us (Leslie Mikesell) (08/12/89)

In article <19001@mimsy.UUCP> chris@mimsy.UUCP (Chris Torek) writes:

>>Why not enjoy seeing a reasonable facsimile of your document right
>>on the screen?

>Mostly because, at the moment, that is not possible.  A reasonable
>facsimile of the document would require a 2000x3000 pixel screen.

"Reasonable" is a relative term.  Many times all you need is the
text, possibly with some modified attributes for fonts etc., but
you want to see the layout with the line and page breaks as you
go, or keep parallel columns together.  For example, I find it
difficult to compose something like a script or quoted text with
a comments in parallel paragraphs without seeing the context. Several
of the semi-wysiwyg wordprocessors handle this "reasonably" well. That
is, they calculate the character sizes on the fly but display the normal
screen fonts expanding the scale if needed to approximate the layout.

Les Mikesell

grunwald@flute.cs.uiuc.edu (Dirk Grunwald) (08/12/89)

Well, I know at leats one very good reason I use TeX rather than,
e.g., Word or somesuch.

I run a lot of simulations. For the last chapter I just wrote, I
converted the output of my simulation into a `vc' (or `sc') format
file using a Gnuemacs elisp function [using emacs as a glorified awk].

I then wrote out the visicalc file to a text file. This formats my
data, rounding just like I want it. I then use another Emacs function
to insert the LaTeX formatting commands to make it into a table.  It
actually uses a custom LaTeX environment to do this, built using
\newenvironment. I changed this \newenvironment approximately 10 times
before I liked what I saw [ using texx2 ].

Using this, I converted and formatted 16 tables of data.  It took a
couple of hours, but much of that was spent deciding how I want things
to look. The next 16 will be easier & faster.

How easy is it to import data into tables for WYSIWYG editors?
Admittedly, this is ``just another feature'' and probably isn't all
that difficult to add, but the point is that I did all of this using
relatively old software. Unless the WYSIWYG system has a textual
intermediate representation, I think that this would be very difficult
to do.

Of course, a two-view system would be able to deal with it.

les@chinet.chi.il.us (Leslie Mikesell) (08/13/89)

In article <GRUNWALD.89Aug11232039@flute.cs.uiuc.edu> grunwald@flute.cs.uiuc.edu writes:

>Well, I know at least one very good reason I use TeX rather than,
>e.g., Word or somesuch.
[...]
>I then wrote out the visicalc file to a text file. This formats my
>data, rounding just like I want it. I then use another Emacs function
>to insert the LaTeX formatting commands to make it into a table.
[...]
>How easy is it to import data into tables for WYSIWYG editors?

MS Word 4.0 and up can import directly from a Lotus or multiplan worksheet,
bringing the data in with tabs separating the columns.  You can select
the whole sheet, named regions or by row, column ranges.  You
do have to set the tab stops yourself to get the column widths
right, but the measurements can be stored in a style and easily applied
to similar tables.  If you like boxed tables, you can select paragraph
borders and add the vertical lines to the tab settings.  Again, you can
store the whole setup as a style, after tweaking one to suit your fancy.

BTW, these 3 functions of Word (.wks import, easy vertical lines between
columns, and styles by example) are about the only things that I find
lacking in Wordperfect.

Les Mikesell

cik@l.cc.purdue.edu (Herman Rubin) (08/16/89)

In article <387@kunivv1.sci.kun.nl>, eykhout@kunivv1.sci.kun.nl (Victor Eijkhout) writes:

> Conjecture: wysiwig systems are for people who make their own
> layout, and who have decided on the definitive layout
> before they started keying in the text. This I think is a wrong
> way of working. I think I have a right to say this, because I've
> produced some 'master pieces of the printing art', and the design
> was done by a pro, and only after I had finished the text.

WYSIWYG systems are the only reasonable thing for someone writing the 
paper.  When I compose my articles, I prefer to putting the stuff on
the screen instead of on pieces of paper.  I have written papers using
TeX, and it is a real pain.

There is no reason why a WYSIWYG system cannot be augmented into a TeX-
like system.  In any case, it should produce output which can be easily
and mainly mechanically converted into a typesetting language.
-- 
Herman Rubin, Dept. of Statistics, Purdue Univ., West Lafayette IN47907
Phone: (317)494-6054
hrubin@l.cc.purdue.edu (Internet, bitnet, UUCP)

phil@delta.eecs.nwu.edu (William LeFebvre) (08/16/89)

In article <1499@l.cc.purdue.edu> cik@l.cc.purdue.edu (Herman Rubin) writes:
>I have written papers using TeX, and it is a real pain.

Have you tried using LaTeX instead?

>There is no reason why a WYSIWYG system cannot be augmented into a TeX-
>like system.

Perhaps you don't know how much TeX really does.  Although it is
probably true that there are no reasons it can't be done, there are
*many* reasons why it is very difficult to do.  Just ask the people
that are working on "ITeX", an incremental version of TeX.  You need
something that works incrementally to do a decent WYSIWYG system.

		William LeFebvre
		Department of Electrical Engineering and Computer Science
		Northwestern University
		<phil@eecs.nwu.edu>

des@yatton.inmos.co.uk (David Shepherd) (08/21/89)

In article <1499@l.cc.purdue.edu> cik@l.cc.purdue.edu (Herman Rubin) writes:
>There is no reason why a WYSIWYG system cannot be augmented into a TeX-
>like system.  In any case, it should produce output which can be easily
>and mainly mechanically converted into a typesetting language.

With a decent dvi previewer you should be able to do all the proof checking
you would do in a WYSIWYG system before printing. 

But, you may say, you have the turn around time for the edit/TeX/preview cycle.
Using (Gnu-)Emacs in an adapted (La)TeX-mode you can quite easily select a
small section of text to independently process and preview (~10s to get screen
image) to get detailed layout right then process and preview a section
(LaTeX provides \includeonly for this) to get overall page placement right.

david shepherd
INMOS ltd