Re: [xml-dev] XML to troff still at work

[Peter Flynn <peter@silmaril.ie>]

On 04/27/2016 08:00 PM, Liam R. E. Quin wrote:
> On Mon, 2016-04-25 at 12:03 -0400, Simon St.Laurent wrote:
>> O'Reilly's old systems were largely DocBook and troff, so this
>> triggered some memories:
>>
>> https://rkrishnan.org/posts/2016-03-07-how-is-gopl-typeset.html
>>
>> Brian Kernighan, still fighting the devil he knows.
> 
> Yes, a lot of fun.
> 
> Unfortunate, too, that we've lost a whole bunch of good ideas in
> moving from troff to TeX 

Perhaps you might share these in private email. 'm always looking for
ammo :-)

> and to Web (even though we gained a whole bunch of new good ideas in
> the process).

I was intrigued by Ramakrishnan's comment:

>>> The Go programming language book is so beautifully typeset. To my
>>>  eyes, it is more beautiful than those LaTeX based ones. But
>>> that's just my perception..

Perhaps. I'd be equally interested to know if this is something that is
perceived to be intrinsic to [La]TeX, or something in the design.

Kernighan than says:

>> We tried Latex briefly but neither Alan or I care for its output,

I wonder does he mean here "its *default* output" which is indeed a
little cumbersome. Working to a proper designer's layout specification,
it would normally be impossible for the non-typographer to detect what
system was used to do any given piece of typesetting.

>> and I personally find it impossible to control.

It has its quirks, like any system, plus an unfortunately steep learning
curve. As I generate a large quantity of LaTeX code from XML via XSLT2,
I too find that having very robust control over the code formation
provides benefits far beyond any that exist natively in troff, LaTeX, or
any other typesetting system.

>> rewrites troff output to do vertical justification so all pages are
>> the same height.

Curious that this was felt necessary when LaTeX does it automatically,
but each to his own.

///Peter