On 6 December 2010 22:31, Joshua Juran <jju...@gmail.com> wrote:
On Dec 6, 2010, at 7:19 AM, Aristotle Pagaltzis wrote:
* Peter da Silva <pe...@taronga.com> [2010-11-29 00:05]:
What wrong direction was chosen in 1969?
Leaving terminal support entirely up to userspace, via libraries?
The only worse clusterfuck that comes to mind from own experience
was printing under DOS.
In a case of premature computation gone horribly wrong, text editors for DOS
(and to this day, Windows) store text data as programs. Rather than go the
trouble of actually *running an app* to print a document, the user is free
to COPY HATE.TXT LPT1, since the purported text file is actually a sequence
of bytecode instructions for a printer. (Such files even used to end with
^Z as a further appeasement.)
And for this extraordinary convenience, it has only cost us negligible
trivialities, such as the assumption that \n codes a single character, and
the small requirement of writing a state machine to detect newlines instead
of a mere byte comparison. But surely, perpetually paying the interest on
this technical debt is worth having the marvelous convenience of sending a
file straight from disk to the printer without the CPU/memory bottleneck.
Even if we stopped using it fifteen years ago.
Imagine storing a shell script as PostScript, which the shell and your text
editor (and grep, etc.) are required to parse every time they open it, all
so you can netcat it to your printer without running enscript.
This reminds me of the interesting story of why WordStar failed... And
it all supposedly hinges on printer drivers...
Yves
--
perl -Mre=debug -e "/just|another|perl|hacker/"
