[OS X TeX] overview of what TeX is producing

Maarten Sneep maarten.sneep at xs4all.nl
Wed Mar 2 04:55:57 EST 2005

On 2 mrt 2005, at 2:06, Ross Moore wrote:

> Yeah; there needs to be more playing around with these parameters
> to get the best results.
> Even your final example of TeX has two **awful** Overfills
> at the top of the 2nd page. It also has one hyphenation
> of  Al-ice  --- quite horrible to my mind.

I've played with the parameters, and blocked hyphenation of Alice
(you wouldn't want to cut up a nice girl).

> That's the awful dilemma with TeX and LaTeX these days,
> especially with non-standard layouts:
>   You *can* get superior results with plain text --- but only
>   after lots of playing around with parameters, in a non-visual way.
> You always have to re-run and look at every page --- or pay
> attention to the .log file. (How many people really do that?)

I did for my thesis (only one underful, in a location that is bloody
hard to notice - an only by a tiny amount).

For the samples I filtered the output on "Over". There are still a few 
but that is unavoidable I think.

> We need a bunch of collections of parameter settings that can
> be easily applied (via the \everypar hook, say).

as if \everypar isn't overloaded enough already…
What do you ahve in mind? tolerance levels, other parameters?
automagical determination trough log-file analysis?

> Clearly some settings should change according to the width
> of the columns. Work is then needed to determine what kind
> of relationship this is, and how to impose it (maybe 'offer'
> is a better word to use) automatically.

In fact, if I would be really nasty, I'd play with the column width
until TeX comes up with a near perfect solution (including some
hyphenation points, perhaps). Then I'd ask for the other programs
to try and reproduce that. No, I'm not that nasty.

> BTW. Your Word examples were a little unfair.
> They seem to have wider columns, located further
> to the left. One of them has only 1 page, not 2
> like (all ?) the other examples.

Yes, it is not the first time I'm bitten by a typo (0.32 vs 0.4 cm 
(the latter is what I intended). The original mismatch was on the
column-width. (and I don't have Word myself, so I cannot correct it).

On 2 mrt 2005, at 10:42, Gerben Wierda wrote:

> I noticed a few things: In the second paragraph I notice a hole 
> between e
> and a in dream and i and second s in sister and more problems like 
> these.
> This disappears if I scale to 200% to 300% (I am now using Acrobat 5 on
> Win2000). Those effects are also visible with the LaTeX-produced 
> output.
> It seems to me that using PDF like this might not be a good tool.

Maybe, but what else do you have in mind? Once printed the issue should
disappear anyway. Just read you second message: I'll render them as 600 
png files for inclusion in a final document.

> Secondly, I would try this with only one paragraph. If one takes the
> second paragraph for instance, one sees that Word in the end makes a
> wide-spaced line of "what a wonderful dream it had" but TeX's finale 
> lines
> of the paragraph are as good as the first.

I didn't know which ones would be good, and which ones would be horrible
beforehand. Right now I can just pick a part I like and use that.

> TeX however produces a rather awkward long space after dream!' in the
> second line of the second paragraph which I find worse than what Word
> produces.

That may be the American typesetting tradition (double space after the 
end of a
sentence), combined with a particularly bad underful.

> Word also almost doesn't hyphenate. So you might see the combined 
> effect
> of line-by-line and how much penalty a hyphenation brings.
> This, I think, is not an easy project. But it is fundamental if you 
> want
> to be able to show people why the complex algorithm is better.
> Maybe no-hyphenation and a wider text width would be better to show the
> difference, you have one less variable in the mix.

I agree that this is indeed a problem with many parameters. Suggestions 
help are both welcome. I took this narrow width to stress the engines, 
but one
could focus on Word vs. TeX, and play with the width, the hyphenation, 
various TeX parameters.


--------------------- Info ---------------------
Mac-TeX Website: http://www.esm.psu.edu/mac-tex/
           & FAQ: http://latex.yauh.de/faq/
TeX FAQ: http://www.tex.ac.uk/faq
List Post: <mailto:MacOSX-TeX at email.esm.psu.edu>

More information about the MacOSX-TeX mailing list