Hello alls,
Deri James <d...@chuzzlewit.myzen.co.uk> wrote:
> If I have misunderstood Eric's intentions with regard to the purpose of
> introducing the .hygiene command, then it would be very helpful if he could
> elucidate further.
The .hygiene command is an interesting debate. I don't
exactly know what to think about it.
After writing several articles about troff on a french
linux & free software website, and reading the various
comments, I've got the feeling that:
1. people usually don't know that troff is able to produce
postscript documents.
2. people which use latex are happy with it and will not
switch to troff, even if it's typographic quality is
equivalent.
3. most people prefer using a markup syntax (txt2tag,
markdown...). Some of them are interested by another
backend than LaTex to produce pdfs.
4. some people are interested by the programmability of
troff, and want a tutorial to build their own macros.
I believe that the group 3. and 4. are the onliest group of
people which may join the community of groff users.
Those two groups of people need to be able to insert raw
troff commands in their documents: txt2tag, for example,
insert some of these raw troff commands while producing
troff documents, and should insert more of them to do a
better job (to deal with imbricated fonts for example). End
users of troff usually work with some handmade macros that
they need for their daily work.
Furthermore, I had a quick look at the evolution of txt2tag.
Its first release was a basic tag transformer. With time,
more and more complex commands where introduced, and there's
now a kind of txt2tag language that can be used to automate
some things and tweak the default export mechanism. Markdown
seems to follow the same evolution.
A versatile language is the core component of groff. And I
think that's its best feature for new users.
So... An hygienic mode could truncate groff, and effectively
decouplate a syntax and the troff backend. There would not
have any reason to use troff anymore...
But that also might be a nice part of groff to be able to
check that a document is strictly semantic. With that in
mind, instead of an hardcoded lock in the macro package, I'd
prefer and external tool, a kind of `checknr`, or a command
line option.
Saying that, I still don't exactly know what to think about
that idea.
Cheers,
Pierre-Jean.
