[Scribus] Book length projects.

[Craig Bradney]

> - The current text engine has a number of serious issues with
> performance, justification quality, etc. These are being addressed with
> the new text engine work in the unstable branch, so expect big
> improvements here.

Big improvements for sure. Its already a lot better.

> - The text engine scales very poorly for large volumes of text, and can
> become unusably slow when faced with long chains of large text frames.
> Work is also being done in this area, but I don't know how much it'll
> improve without more rendering engine changes too - I'm no longer
> keeping track of development well enough.

Not anymore

> - As mentioned, widow/orphan control is currently lacking. I think avox
> is working on widow/orphan handling as part of the new text system work.

Yes.. 

> - There's no support for automatic index creation, and there's no decent
> way to add it to current builds. However, once character styles are
> fully supported in the development series, an index tagging system could
> probably be built on top of the char styles infrastructure.

It will be relatively simple in 1.3.6+ (remember 1.3.5 is just a qt4 port,
no real code changes).

> Yeah, that'd work, but as you noted it'd be pretty tedious. I'd prefer
> to be able to use char styles to mark indexable terms myself.

No need for styles for indexing, or TOC in the future.

> > It might even be conceivable that you could specifically flag a word or 
> > group of words as "indexable" in the text frame editor and not need to 
> > tie it in with styles -- just as you could bold or underline text, you 
> > could also index it.
> 
> That'd probably be ideal, though I expect char styles or their
> underlying file format and text frame support would be used as the
> underlying mechanism for doing that.

Just the file format.

Craig