Gregory Pittman <gregp_ky at ...> writes: > > On 10/26/2011 10:13 AM, john Culleton wrote: > > > > Observations: > > The TeX short line sample requires one less hyphen (5) than the Scribus > > short line sample (6), and fitted two more words onto the page. The OO > > short line sample required 7 hyphens. The Masterson InDesign short line > > sample required 5 hyphens. The other InDesign sample required 7 > > hyphens. Some designers are better than others :<) > > > > My observations are that there is no inherently greater typographic > beauty when comparing TeX with the true DTP programs, including Scribus. ... > At least in my view the superiority of paragraph-wide before line > algorithms has not been demonstrated.
No need, that's a mathematical fact: Any method for linebreaking needs to choose one break point for each line. Simple line-oriented breaking only considers one candidate per line, advanced line-oriented breaking considers two candidates (at least if min. glyphextension and/or min wordspacing are less than 100%) and paragraph-orineted line breaking a la TeX considers *all* possible breaks. That means that TeX can in theory produce all breaks that another algorithm might produce, but in practice TeX chooses different breaks which are weighted better. If these better choices result in a more pleasing layout and if the difference is noticeable at all is another question. Also, TeX let's words protrude over the right edge if it's desparate where other programs just create a huge white gap. In my experience the difference between simple and advanced lineoriented breaking is higher than the diffrence between advanced line-oriented breaking and paragraph-oriented breaking, especially if the min wordspacing is 66% or less (why should you accept 500% wordspace but not 50% wordspaces?). That said, Scribus will have paragraph-oriented linebreaking in the future. I've already implemented it for a prototype, but the current text system still needs some work before it can be ready for it. /Andreas
