12 Jan
2022
12 Jan
'22
6:44 a.m.
This is clearly getting off track of TUHS. I'll stop
after this reply.
*From:* Blake McBride <blake1024(a)gmail.com>
*Date:* January 11, 2022 at 2:48:23 PM PST
*To:* Jon Forrest <nobozo(a)gmail.com>
*Cc:* TUHS main list <tuhs(a)minnie.tuhs.org>
*Subject:* *[TUHS] TeX and groff (was: roff(7))*
Although I'm not connected with the TeX
community, I don't agree with
much of what you said.
1. TeX source to C is fine - stable and works. It would be
impossible to rewrite TeX in any other language without introducing
bugs and incompatibilities. Leaving TeX as-is means that it can be
converted to other languages too if/when C goes out of style. TeX
as-is is exactly what it is. Anything else wouldn't be TeX.
I agree that Web->C works but it's a major obstacle in doing any
development work on TeX. Try making a major change in the Web source
that requires debugging.
Anything that can pass the TeX Trip Test can be called TeX. I know of
a full C reimplementation that passes the test but the author doesn't
want to make it free software.
There are other rewrites out there that could be candidates but someone
will enough power will have to proclaim one as the official TeX
alternative.
2. Drop DVI? Are you kidding me? Although PDF may
be popular now,
that may not be the case 20 years from now. A device-independent
format is what is needed, and that's what DVI is. TeX is guaranteed
to produce the exact same output 100 years from now.
And .PDF isn't?
.DVI was great until .PDF matured. .DVI has almost no penetration
these days, whereas .PDF is everywhere. I'm not saying that .PDF
will always be the proper alternative but a properly rewritten TeX
should make it much easier to replace .PDF will whatever comes
next.
3. I am curious about memory limitations within TeX.
TeX has various fixed sized memory pools, and contains clever code
to work around limited memory. Some of the newer TeXs,
like LuaTeX, use dynamic allocation but this isn't official.
Given how primitive things were when TeX was developed it's a
miracle it works as well as it does.
4. Knuth is getting up in age. Someone will have to
take over.
Exactly. I don't follow the TeX community so I don't know what
they're doing about this.
Jon Forrest
12 Jan
12 Jan
8 a.m.
New subject: TeX and groff (was: roff(7))
On Tue, Jan 11, 2022 at 6:45 PM Jon Forrest <nobozo(a)gmail.com> wrote:
No. It isn't. It is an Adobe product.
I know of
a full C reimplementation that passes the test but the author doesn't
want to make it free software.
Although it is possible, I find this hard to believe. I can't imagine
spending as much time as would be required to duplicate something that
already exists.
There are other rewrites out there that could be candidates but someone
will enough power will have to proclaim one as the official TeX
alternative.
Again, hard to believe.
2. Drop DVI? Are you kidding me? Although PDF
may be popular now,
that may not be the case 20 years from now. A device-independent
format is what is needed, and that's what DVI is. TeX is guaranteed
to produce the exact same output 100 years from now.
And .PDF isn't?
.DVI was great until .PDF matured. .DVI has almost no penetration
these days, whereas .PDF is everywhere.
DVI was never meant to have any penetration. It was always intended to be
an intermediary format.
I'm not saying that .PDF
will always be the proper alternative but a properly rewritten TeX
should make it much easier to replace .PDF will whatever comes
next.
Again, given the complexity of a proper TeX, and its declining popularity,
I find it ver hard to believe that someone would spend the time to
duplicate, with enhancements, it.
Blake McBride
8:10 a.m.
New subject: TeX and groff (was: roff(7))
<…>
No. It isn't. It is an Adobe product.
Check out ISO 32000-2:2020.
I think it’s ok in 2022 to say that PDF has escaped Adobe and is an open standard,
suitable for long-life documents, etc.
d
2. Drop DVI? Are you kidding me? Although PDF may
be popular now,
that may not be the case 20 years from now. A device-independent
format is what is needed, and that's what DVI is. TeX is guaranteed
to produce the exact same output 100 years from now.
And .PDF isn't? 
8:26 a.m.
New subject: TeX and groff (was: roff(7))
Knuth made an excellent point, maybe in _Coders At Work_ about literate
programming and why it didn't catch on: in general, about 1 out of 20
people can be a really good programmer. In general, one of 20 people can
be a really good writer. These talents are largely uncorrelated.
Sure, being competent at either is a teachable skill. But no one wants to
read either a program or a narrative written by someone who's merely OK at
it.
Adam
On Tue, Jan 11, 2022 at 7:20 PM David Arnold <davida(a)pobox.com> wrote:
<…>
2. Drop DVI? Are you kidding me? Although PDF
may be popular now,
No. It isn't. It is an Adobe product.
Check out ISO 32000-2:2020.
I think it’s ok in 2022 to say that PDF has escaped Adobe and is an open
standard, suitable for long-life documents, etc.
d
that may not be the case 20 years from now. A
device-independent
format is what is needed, and that's what DVI is. TeX is guaranteed
to produce the exact same output 100 years from now.
And .PDF isn't?
13 Jan
13 Jan
1:54 a.m.
New subject: TeX and groff (was: roff(7))
On Tue, Jan 11, 2022 at 9:01 PM Blake McBride <blake1024(a)gmail.com> wrote:
No. It isn't. It is an Adobe product.
Up to a point, Minister. PDF/A is an ISO standard that tracks PDF 1.4 or
PDF 1.7. It is meant for creating archivable PDFs, so it excludes linked
fonts (as opposed to embedded ones, which are allowed), JavaScript,
audio/video, encryption, external references, etc. For troff purposes, we
don't need any of that, so it's just a matter of setting the metadata
correctly. ISO standards can be withdrawn, but they remain available; I
doubt this one will be, since libraries are depending on it. There are
lots of FLOSS toolkits to generate PDFs.
2. Drop DVI? Are you kidding me? Although PDF may
be popular now,
Well, provided
there are DVI-to-whatever converters then. it's a systems
problem. What we really need is gcc support for some processor that is
easy to emulate (at least the userland). Historically that was MIPS; now
it's probably RISC/V. Or, I suppose, MMIX; there is a very partial Verilog
description at <https://github.com/tommythorn/fpgammix> that would make it
possible to create a hardware integer MMIX CPU using FPGAs.
And .PDF isn't?
that may not be the case 20 years from now. A device-independent
format is what is needed, and that's what
DVI is. TeX is guaranteed
to produce the exact same output 100 years from now.
4:13 p.m.
New subject: TeX and groff (was: roff(7))
nevertheless it's still an Adobe product.
--- Ursprüngliche Nachricht ---
Von: John Cowan <cowan(a)ccil.org>
Datum: 12.01.2022 20:54:34
An: Blake McBride <blake1024(a)gmail.com>
Betreff: Re: [TUHS] TeX and groff (was: roff(7))
an Adobe product.
14 Jan
14 Jan
2 a.m.
New subject: TeX and groff (was: roff(7))
Only in the sense that Fortran is an IBM product. (IBM's Fortran compilers
and runtimes are IBM products, but that's another matter.)
On Thu, Jan 13, 2022 at 5:13 AM Thomas Paulsen <thomas.paulsen(a)firemail.de>
wrote:
nevertheless it's still an Adobe product.
--- Ursprüngliche Nachricht ---
Von: John Cowan <cowan(a)ccil.org>
Datum: 12.01.2022 20:54:34
An: Blake McBride <blake1024(a)gmail.com>
Betreff: Re: [TUHS] TeX and groff (was: roff(7))
an Adobe product.
1074
days inactive
1075
days old
6 comments
6 participants
participants (6)
-
Adam Thornton -
Blake McBride -
David Arnold -
John Cowan -
Jon Forrest -
Thomas Paulsen