[reportlab-users] RE: Table sizing algorithms - comments welcome

[Robin Becker]

In article <PGECLPOBGNBNKHNAGIJHGECNELAA.andy@reportlab.com>, Andy
Robinson <andy@reportlab.com> writes
>> That's what I'm trying to say. There are *many* table environ-
>> ments for LaTeX users, some optimising different features. This
>> is totally ok. What I'm a little surprised with is the fact of be-
>> ing unable to find anything directly connected to Knuth himself,
>> which suggests that table rendering is a trivial issue! ;-)
>
>IMHO he was on a 10-year quest for the perfect paragraph,
>had a sense of how to do it, and little else concerned him.  
>Maybe he figured he could not afford 4 decades on tables 
>and left it to the user group :-)
>
>BTW, I was reading a cheap thriller (Trading Reality, Michael Ridpath)
>about a Virtual Reality company that gets it trouble.  The deceased
>boss's younger brother, a bond trader, takes over, and says to
>the head techie that he wants to learn about computers
>and try some programming.  She hands him TAOCP and says
>"I don't believe you. Come back when you've read that and 
>we'll talk!".  I doubt Mr. Ridpath has ever looked at TAOCP
>which made it doubly hilarious.
>
>- Andy
well as a matter of historical fact, the Knuth paragraph splitting
algorithm has been working fairly reliably since the mid 70's, it's also
one of the few algorithms with a prize for bug discovery. The fact that
the current lead implementation is written in the awful way it is
doesn't detract from its efficacy. 
-- 
Robin Becker