[scribus] Paragraph Justification

[Ted Powell]

On Tue, Nov 15, 2011 at 08:54:13AM +0100, a.l.e wrote:
> hi ted
>> [some earlier quotes omitted]
>> Many years ago (30+) when I had access to a Triple-I phototypesetter I
>> put together an application that did simple page layout and formatting.
>> One of the user-settable parameters was a threshold on the length of
>> the last line of a paragraph, expressed as a percentage. Final lines
>> exceeding the threshold were justified, the others not. The Scribus
>> choices "Align Text Justified" and "Align Text Forced Justified" would
>> correspond to 100% and 0% respectively.
>
> nice!
>
> the more i read about this topic, the more i think that the forced  
> justified should really go to a "first/last line" tab and the threshold  
> you designed 30 years ago should be in there!
>
> if you have any spare time you're welcome to provide a patch :-)

In January 2004 I had a problem with the layers code being broken,
so I sent a patch to Franz Schmid, which was accepted. Since then,
things have generally been working for me.

When time permits, I'll have a look at this, though.

> other ideas of the same quality from your prehistorical application are  
> more than welcome!

AFAICR (as far as I can remember), if the last line of a justified
paragraph was not justified, it simply had the default spacing between
words.

If when building a non-final line a word that overflowed was always moved
(in full or in part, considering hard/soft hyphens) to the next line,
and the word spacing increased, this would mean that the average text
density of the non-final lines would generally be less than that of the
final line.

So I had three values for the spacing between words: min, default, max.
A word (or hyphen-delimited partial word) would be moved to the next
line or squeezed into the current line, depending on which had the
lesser effect on the word spacing. AFAICR, min was 0.75 or 0.8 of the
default. Max was something large; exceeding it meant that the line would
be just left-aligned even though it wasn't final.

This may have been overkill, but I had fun doing it (even though it was
all in PL/I) and did production work with the result.


-- 
Ted Powell <ted at psg.com>   http://psg.com/~ted/
The best way to get to know a system is to tinker with it. --leb