[tex-k] do not understand

[Wolfgang Helbig]

Hello,

thank you, Karl. I agree with you and with Don Knuth. In the preface on 
page vi, last paragraph, he admits, that he lies now and then. And this 
seems one of the places. But Don promises, that in later paragraphs, 
he'll tell the whole truth. But still, I do not know, when and why 
penalties after breaks are discarded. This makes me feel uneasy, but 
only very little so.

Greetings,
Wolfgang

I would not call it either a lie or an error. It seems to me that the
> explanation is in the sentence in the paragraph after the exercise:
>
>    a sequence ... will vanish as a unit ...
>    unless the optimum breakpoint sequence includes one or more of the
>    penalties.
>
> The "optimum breakpoint sequence" is being forced to include both (all)
> -10000 penalties, therefore they are not discarded.
>
> Another hint is in the answer to the exercise:
>    ... unless both of those values force breaks.
>
> So there are clues that -10000 penalties are special wrt discarding.
>
> I can imagine other ways to word this that might be clearer, but I can't
> justify adding another report to the pile for this. --best, karl.


-- 
Wolfgang Helbig