Re: [xml-dev] The impact of data format selection on application develop

[Marcus Reichardt <u123724@gmail.com>]

Hey colleagues,

what's with this hatred towards simple publishing workflows based on CSV/TSV? I've used troff/groff, eqn, pic and co many times to prepare eg tutorial handouts, product documentation, and the like eg when LaTeX was overkill. That CSV might be underspecified compared to her complicatedness XML isn't a problem in these type of applications; we're talking about publishing where the final result is still inspected by a human, rather than enterprise integration where large data sets are automatically processed. TSV is as simple a data format as it gets - one ASCII char for record separation, another one for field separation, and that's it.

Incidentally, the SGML world, via parsing CSV-like streams into canonical angle-bracket form using SHORTREF, is much more open to these practical publishing workflows - you can integrate output from Unix tools via basic facilities eg entities into SGML data much more readily compared to XML where special tools are required. And although SGML came from IBM mainframes and has seen decades-long usage for preparing printed banking and insurance statements everybody has seen, a fun fact still is that groff, the GNU port of troff, was implemented by James Clark before he ventured into SGML big time.

Have a nice week,
Marcus Reichardt
sgml.io

> Am 10.07.2022 um 16:57 schrieb David Carlisle <d.p.carlisle@gmail.com>:
>