[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: A Modern Typesetting Language. (Was: .TQ to replace .PD 0)
From: |
John Gardner |
Subject: |
Re: A Modern Typesetting Language. (Was: .TQ to replace .PD 0) |
Date: |
Wed, 25 May 2022 23:00:36 +1000 |
Hi Ralph,
Or at best, gives it through some clunky ‘treat it as a string’ mechanism.
>
How is that clunky? Text is text. It's opaque, honest, and universal. The
foundation of the Unix Philosophy… you know this as well as I do. ;-)
One could look at shoehorning evermore complexity through to the
> post-processor, but that denies integration with the rest of troff of the
> expressiveness of those features.
>
I have a (currently hypothetical and unimplemented) solution that might
resolve the issue of integration and expressiveness. I've posted about it
before in this list
<https://lists.gnu.org/archive/html/groff/2020-09/msg00031.html> and won't
subject folks to another long-winded diatribe. ;-) But the premise is that
this:
.UR https://example.com/
Hyperlink example
.UE .
gets preprocessed into something this:
\X'meta: begin link'\c
.UR \X'meta: begin href'https://example.com/\X'meta: end href'
Hyperlink example
.UE .
\X'meta: end link'
… that's then used to demarcate regions of semantic or structural value for
a postprocessor that enhances the groff_out(5) source in some
driver-specific fashion.
I'm cognisant of the challenges this would involve, and I'm not even
claiming it's doable at this stage, but I'd still like to have a crack at
it nonetheless. 👍 It'll be fun when I get around to it.
— John
On Wed, 25 May 2022 at 21:43, Ralph Corderoy <ralph@inputplus.co.uk> wrote:
> Hi John,
>
> > > Support of modern font technologies and of course languages which
> > > aren't left-to-right.
> >
> > Agreed. But for everything else you've mentioned: it's just a matter
> > of writing another PDF postprocessor (or some other adapter for a
> > particular format). Postprocessors are where the real beauty of
> > Troff's staying power shines.
>
> One could look at shoehorning evermore complexity through to the
> post-processor, but that denies integration with the rest of troff of
> the expressiveness of those features. Or at best, gives it through some
> clunky ‘treat it as a string’ mechanism. Think more of a language were
> expressions can have these as first-class things with powerful
> operators.
>
> > > but the modern graphics model of PDF has moved on a lot from theirs
> > > and isn't targeted. Images and SVG as first-class objects.
> > > Transformation matrices. Advanced colour handling. Text-flow
> > > layout.
> >
> > PDF's graphics model hasn't changed
>
> From memory, PDF 1.3 added transitioning between colours, PDF 1.4
> introduced transparency, and PDF 1.7 gave us 3D artwork. There must be
> many more incremental improvements. :-)
>
> > and SVG isn't a first class object in PDF documents.
>
> No, I know. I was meaning they would be in a new document-layout
> language.
>
> --
> Cheers, Ralph.
>
>
- Re: .TQ to replace .PD 0, (continued)
Re: .TQ to replace .PD 0, Ralph Corderoy, 2022/05/23
Re: .TQ to replace .PD 0, Ingo Schwarze, 2022/05/24
Re: .TQ to replace .PD 0, Ralph Corderoy, 2022/05/25
Re: .TQ to replace .PD 0, John Gardner, 2022/05/25
A Modern Typesetting Language. (Was: .TQ to replace .PD 0), Ralph Corderoy, 2022/05/25
Re: A Modern Typesetting Language. (Was: .TQ to replace .PD 0),
John Gardner <=
Re: A Modern Typesetting Language., Ralph Corderoy, 2022/05/25