[scribus] how does scribus do justification?
louis.desjardins at gmail.com
Fri Jan 29 06:39:06 CET 2010
2010/1/28 John Culleton <john at wexfordpress.com>
> On Thursday 28 January 2010 13:37:14 Gregory Pittman wrote:
> > On 01/28/2010 12:36 PM, John Culleton wrote:
> > > Scribus, MSWord etc. only try to set one line at a time without regard
> > > what goes on in previous or following lines in the paragraph.
> > I'd like to know what your evidence is for this conjecture.
> > As I take a frame, highlight part of a line, then adjust its font size,
> > width of glyphs, kerning, it doesn't take much adjustment before the
> > line is giving away to or stealing from the following line.
> > My guess is that you may be referring to built-in automated adjustments,
> > such as has been implemented with word tracking in Scribus 1.3.5+.
> > Certainly this feature has room for improvement, but even now, it's
> > working on a paragraph level.
> > Greg
> > _______________________________________________
> > scribus mailing list
> > scribus at lists.scribus.net
> > http://lists.scribus.net/mailman/listinfo/scribus
> True enough. If you add text the line may overflow to the next, which then
> to be reset, which may cause the next line to over flow etc. till you get
> the end of the paragraph. But nowhere in this serial process is the
> as a whole evaluated to see what the optimum placement of extra space for
> justification is. If however you add material to a TeX line then the whole
> paragraph is evaluated for optimum placement of the extra space for
> justification. That is called paragraph at a time justification. TeX does
> InDesign (using the TeX method) does it, and no other DTP programs do it.
> That is why TeX programs with hyphenation allowed will have a lower count
> hyphens than the same document justified by e.g., Scribus or MSWord.
As I find this discussion very interesting, I certainly don’t want to sound
like entering it loudly. But I would like to add a couple of nuances, here.
No matter how well the programmers did their job about justification, it’s
up to the user to give the starting parameters with which the program will
have to work in order to give a justification result that will be acceptable
The first 3 parameters are:
3. Line width
Fonts have various width. It’s easy to compare 2 fonts and notice that there
can be differences of width for the same letter on those 2 fonts. When you
do this, take letter m or w and look at them at a fairly large size to see
how different they can be from one font to another. Try this with various
The Typesize (for example, 12 pts as opposed to 24 pts will be enough to
make a good comparison point) has to be in relation with the Line width. Set
a text with a 12 pts size font and set the same text with 24 pts size and
you will already see how hard the program has to struggle to give a decent
justification result, with the same Line width (or text frame width).
Already, those 3 parameters will dramatically affect the way the program
performs at justifying text, no matter how good the program is and no matter
how it justifies text, by the line of by the paragraph.
The number of hyphens the program will allow is in fact a user setting that
adds to the complexity of the justification method. It’s a 4th parameter.
The users can set this to none or zero, or to 2 or 3 or more. If set to none
or zero, this means that the program is not allowed to hyphenated words that
are in consecutive lines. The users can set the number of hyphens to 2 or 3
or even more if they wish, so that the program is allowed to hyphenate words
on 2 or 3 consecutive lines of text.
A lot of people forget about the hyphenation settings and this is one of the
reasons for poor justification results. The other reasons are related to the
first 3 parameters.
In short, keeping in hand the 4 main parameters for text justification wiill
help achieve nice results.
On top of all this, if the program allows for hanging punctuations and
letterspacing vs wordspacing parameter controls, then it’s only going to be
even easier to set. But again, this is not turning the application into a
Since version 1.2, Scribus has not failed to justify nicely, when users keep
in mind the relationship between the 4 main settings that affect
In 1.3.5, using the high-level hyphenation method, even with hyphenation
turned off we can achieve outstanding justification results. This gives the
designers more freedom to achieve nicer typography with even more dramatic
settings on the first 3 main parameters. You’ll be able to get closer to the
edge, not over it.
> In TeX of course hand kerning etc. is generally not done because the
> does it better and considerably faster. I am talking body text now, and not
> display type on covers.
> If in Scribus 1.3.5 etc if a change in line five may change the word
> between lines one and two or prior hyphenation etc. then you indeed are
> working towards full paragraph justification. But I had not heard that this
> feature was present yet.
> John Culleton
> "Create Book Covers with Scribus"
> Printable E-book 38 pages $5.95
> scribus mailing list
> scribus at lists.scribus.net
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the scribus