Re: Text processor

[Tim Allen <timallen godzilla zeta org au>]

On Sat, 19 Jun 1999, Robin Miller wrote:

> Okay, I got Gnome running. Now I need to be able to *work*
> with it. I write, like words, for a living, mostly online,
> and so far I haven't found a single Linux text processor
> that has the features I need, which are:

I admit - a good GPL'd word processor is still thin on the ground. AbiWord
might eventually get there (*please* let it be GNOMEified!), but it's not
there yet.

In short, what you want is a text-editor that is similar to the editors
you're familiar with, one with a very low learning curve, yet powerful and
efficient, right? Hehe.. Unfortunately, that's not the way Unix tends to
do things, as I'm sure you've noticed.. :)
 
As much as I'm sure you're right that such an application would greatly
help, I've got an idea why one hasn't been written...

> 1) Mousable Win/Mac-type interface

For starters, mice detract from keyboard time - it's easier to keep using
the keyboard than it is to keep switching between the two.

> 2) Ability to strip SGML and HTML tags from imported text (a
> la quotes from e-mail and Web sites)

For starters, email shouldn't have HTML tags. :)

Secondly, I'm sure it's the work of a moment to write a HTML-stripping
macro for emacs, and in vi I generally just play with a few global
search-and-replace commands. On the other hand, I'm sure there's already
some kind of filter that will remove HTML tags - vi can (and I'm sure
emacs would too) pipe a region of text through any arbitary command, and
replace it with the out put of the filter (sort(1) is a good example - a
few keystrokes and your lists are sorted alphabetically ;)

There exists a program called w3m which I think can (like lynx) parse HTML
into pure text, and I suspect it would be very good for this.

> 3) Ability to strip in basic HTML tags; links, <P>, <I>,
> etc.

What does that mean, "strip in"?

elvis, vim, and emacs can all syntax-highlight HTML (among other things)
and help you type it in..

> 4) Character Conversion; UC to lc, lc or UC to Capitalized,
> etc.

vi has the "~" command which changes the case of the letter under the
cursor and moves to the next character, so hold down "~" and change the
case of all your text.. :)

As for the more esoteric styles, again, an emacs macro, a commandline
filter, or a more interesting vi macro could do it.

> 5) Able to write and save in ASCII, Word, and RTF.

Hmm.. Well, OK, ASCII is the easy one. Again, all these filters are
writable, but I can't claim that any text editor can write these natively.
And I suppose as writer, it's a Microsoft World out there. Still, a HTML
-> RTF filter couldn't be too hard, could it? Say - there's probably
already a GPL'd bison/lex thingy parser for HTML, isn't there? From there
to RTF shouldn't be too hard.

> 6) A WORD COUNT FEATURE! NO PROFESSIONAL WRITER CAN LIVE
> WITHOUT ONE!!!

Oddly enough, Unix comes with a "wc" command just for this purpose :) See
the above comments on filters.

I've been trying to write an elvis alias that will pipe the current
selection (if any) through wc and show the results, or else pipe the
entire buffer through.. this would help you (and me) very much, I know,
but I can't get it working.. damn steep learning curve. :)

> GXedit comes fairly close, but it crashes more than Win98
> running three copies of Netscape on a 486, and the word
> count feature doesn't count words accurately. It is a toy,
> not a professional tool.

Never heard of it..

> Pico and Vi are no better than DOSEDIT -- and harder to use.
> If I wanted to use DOSEDIT I could have stuck with an 8086
> that took about 0.5 seconds to boot. :)  

OK, I must take issue with that. Yes, pico is a toy editor. It's simple,
and it lacks a whole lot of useful features. 

Secondly, edit.com (from the Win9x distribution) is a very nice editor -
it kicks the butt of Notepad easily. It also lacks a lot of nice features,
but it's a darn sight better than Notepad or the DOS 5/6 editor. 

Finally, vi and emacs are the two editing Titans, locked in eternal
combat, in a realm beyond that of mortals. vi has been called the most
efficient text editor ever written, and I'm inclined to agree. Yes, the
learning curves of both (especially vi) are steep, but they pay off.

As a side note, see http://www.csn.net/~bediger/knottenbelt.txt
There was a better research paper, but it seems to have gone offline. 
The vi-lover's homepage: http://www.cs.vu.nl/~tmgil/vi.html

> I haven't tried TED yet, but plan to shortly. WordPerfect
> and StarOffice are too locked-in on paper-oriented formats
> and eat a whole screen. Reporters and others who do online
> research need to work around open browser windows.

You'll be interested to know that elvis can read and render HTML.. and can
be used like a web browser - and it's still a few orders of magnatude
smaller than emacs! <g,r&d>

> We're not
> secretaries or novelists. We research and write at the same
> time, usually at high speed, because we have DEADLINES!!!

Both vi and emacs are highly efficient editors. And I should also note
that emacs has a built-in browser as well. With XEmacs, it'll display
images as well.

> I think one reason so many reporters write negative articles
> about Linux is that there is no way they can actually use
> Linux in their work. I'm sitting here running Linux,
> downloading information about Linux, and writing about Linux
> -- mostly in Windows, using NoteTab. (Through VMware)

Yeah, that's kind of sad, really. Well, here's to that great author's text
editor of the future. Unless you'd like to try emacs. 
 
> There are dozens of usable Windows text editors for writers.
> It would be nice to have one for Linux. (Nedit is the
> closest I've found so far...)

If you can be a little more informative about feature lists, or if there's
some good list on the notetab page, perhaps someone will write you a
NoteTab emulator for emacs. Not I, since I'm a vi partisian..

> I can promise any developer who comes up with a usable,
> non-crashing text/HTML editor for Linux at least a little
> bit of fame and fortune. My sites aren't as popular as
> slasdot, but a good word from me can overload most ftp
> servers. :)
> 
> -- Robin Miller
> Editor and Columnist
> Andover News Network
> http://www.andovernews.com
> http://www.techsightings.com
> http://www.davecentral.com
> 
> -- plus freelancing for all kinds of magazines and
> newspapers.
> (You have probably read things I wrote and didn't know it.)

Maybe I shouldn't have been quite so hasty about Ignoring The Press in my
last post. :)

Tim Allen