From: byron@cc.gatech.edu (Byron A Jeff) Subject: Re: A Word Processor for Linux Date: Tue, 10 Aug 1993 01:48:18 GMT
In article <SWEIN.93Aug9154849@nomad.etsd.ml.com>,
Scott Weinstein <swein@etsd.ml.com> wrote:
>
>I don't think this is my idea, but couldn't a visually based editor
>(WYSIWYG) have the capability to "compile" the document into tex, and
>save use tex as its document format? This would give both sides what
>they want. The power of tex, with the visual instant capabilites of a
>word processor. Possibly it could be written as an add-on, to possibly
>emacs?
>
Several folks have commented on this. In fact doc under interviews is a
stab at this as is a product called Scientific Word that runs under DOS/
Windows. I posted a detailed suggestion on Friday but I think it was on
the phantom thread. So here is a repost. Please feel free to comment.
========================== Begin Included Text ========================
This is my cut on the issues. It's my perception and no one else's:
- People gravitate to what they're most used to. That's why there is such
a big division between the TeX folk and the WYSIWYG folk.
- Different jobs have different levels of complexity and therefore require
different tools to accomplish said jobs effectively.
- Some folks have the time and energy to invest in deep understanding of
complex software. Some folks would rather die than crack the manual.
- There's always a tradeoff between power and complexity.
The bottom line is that there is no simple single solution to the problem. Both
formatting and WYSIWYG word processing tools are necessary and desirable.
They generally fall at opposite ends of the power/complexity line.
Given that what is the solution set? Here's my cut at it.
1) Support everything that already exists. Folks are already used to what
they have and most folks gravitate to what they already know.
DOSEMU is a huge step in that direction. Under Linux many of the DOS based
tools are supported especially WordPerfect. It sucks in my personal opinion
but millions upon millions of folks use it every day. Same to be said for
the Windows Based tools. That's why the WABI development is important for
Linux to hit the big time.
2) InterViews doc is a step in the right direction for new development.
It provides a WYSIWYG interface to TeX. Also a somewhat simple GUI is
available for modifying the documents. Lastly it does produce TeX output that
can be tweaked directly. What it needs however is a few ;-) things:
- extensability - both in the user interface and in key based macros.
- speed - Interviews has too much overhead. Probably coding in Tcl/Tk/C/C++
would give the speed and extensibilty needed.
- Interface templates - so that it look like many of the other popular
word processing products out there. Remember that folks generally
stick close to what they already know.
- New user painlessness and intuitiveness. Easily done with a stripped
down interface template for very basic stuff.
- Direct TeX/LaTeX interpreter with 2 modes.
1) Click compile and it displays your doc alongside the TeX text.
Basically just encapulating the current edit/TeX/xdvi cycle.
2) WYSIWYG as you go. As you type in the TeX window your document
shows up alongside as it will look when printed. Can this be done?
The real solution is to have a product that is actually all things to all
people. To do that it needs WYSIWYG, TeX, simplicity, complexity, extensibilty,
intuitiveness, and of course context based help! ;-)
So imagine a system with a TeX heart and bunches of WYSIWYG faces.
One face looks like WordPerfect, switch faces and it looks like
Word, switch again and it's TeX and xdvi side by side, switch again and it
has your custom interface with the look and feel you like best.
When it's first distributed all the standard faces should come with it.
Maybe we can even call it TeX/Faces! If that happens I want credit!
============================ End Included Text ======================
Let me know what you think.
Later,
BAJ