[ruby-talk:04087] Re: OT: Re: P. of least effort -- fragile typesetters

"Conrad Schneiker" <[e-mail]> writes:

> Hi,
> 
> "Hugh Sasse Staff Elec Eng" <[e-mail]> wrote:
> > On Thu, 13 Jul 2000, Conrad Schneiker wrote:
> >[...]
> > So what would you recommend for producing documents, papers etc?
> 
> I haven't looked around recently, so I don't know if there are any good
> answers.
> 
> > Having wrestled with Word, Star Office, eqn|tbl|troff,  and hand crafted
> > PostScript drawings (ouch!), I'd be interested.  Is Lout flexible enough?
> 
> This is the first time I've looked at it. Certainly looks interesting.
> Anyone else have any experience with it?

I asked a friend, Bennett Todd ([e-mail]), who was nice enough to
contribute the following along with the offer to correspond with those
who wish to take this topic further:

  Lout is pretty appealing; it's both simpler than TeX, and
  sufficiently powerful to allow nice-looking output from simple
  input. It only generates PostScript, but there aren't that many
  other targets of interest for a typesetter anymore, and if you do
  happen to have one, GhostScript will probably allow you to do
  beautiful output to them anyway, so that's really not much of a
  limitation.

  But like TeX, Lout is only for typesetting; it's not the choice for
  rendering to e.g. HTML, or plain ASCII text, or info format, or RTF,
  or LyX, or ....

  I dabbled with Lout some years back, but it just never grabbed me,
  because the markup was as cumbersome and clutterful as LaTeX, and it
  didn't open up all the output alternatives I like.

  I switched to using SGML for a while, with the Linuxdoc DTD, and did
  quite a few documents with that. It's a fine choice for supporting
  multiple outputs, but once again the source clutter is right up
  there, enough to put me off a little bit.

  At the moment I'm enjoying using sdf (listed in Freshmeat), which
  takes something very close to straight ASCII text and deduces most
  of the structure from it, so requiring less markup clutter; it can
  output various formats, and since SGML is one of 'em it can
  automatically produce many great output formats by multiple
  translation.

  I'm waiting for a new tool, currently under development; when it
  becomes available I expect to be able to use text with essentially
  no markup at all, pure straight ASCII text, and parse that to deduce
  doc structure and add SGML markup.

  Certainly when doing this sort of thing the first target to create
  is SGML, since it buys you the most, but I think it'll be worth
  building direct targets for other backends, including re-formatted
  text (making it a _brutally_ brilliant replacement for fmt(1) or
  par(1)), nice clean hand-editable portable HTML, and Lout. I
  continue to like lout for typesetting, even though I hope to not
  have to hand-write any.

/Lew
-- 
Lew Perin / [e-mail] / [e-mail]
www.panix.com/~perin/