Hi damien.

1. compile location issue

- filesystem using UTF-8 with NFC(because linux)

- all book source is plain. im not editing anything. only from git repository

===========================
./pillar export --to="latex:sbabook" --outputFile=SpecBooklet SpecBooklet.pillar
CCConfigurationError: /mnt/USERDISK/onion/99_SpecUI��/BuildingUIWithSpec is not a valid directory.
===========================

error is this line. maybe pillar process then can not acceptable korea charactor in directory location.

before i look similar problem.

a few years ago, im using pharo in linux then can take input from XIM. but can not process korean charactor to pharo workspace. because input charactor process standard is NFD.

work is here. written in korean. but can read. many text is "source code"
http://workspace.onionmixer.net/mediawiki/index.php?title=Linux_cogvm_xim_howto#XIM_.EC.9C.BC.EB.A1.9C_.EB.93.A4.EC.96.B4.EC.98.A4.EB.8A.94_.EC.9E.85.EB.A0.A5.EA.B0.92.EC.97.90_.EB.8C.80.ED.95.9C_.EC.B6.94.EC.A0.81_.EB.B0.8F_.EC.A0.95.EC.83.81.EB.8F.99.EC.9E.91.EC.9D.84_.EC.9C.84.ED.95.9C_.EC.88.98.EC.A0.95

I do not know if it will be a reference.

2. sbabook issue

oh.. no.. no need more oblivoir. because oblivoir is only for CJK env. example for thai then oblivoir cannot show all charactor. oblivoir have a limit. i know already.

I think the current situation is a new one. my work need many language processing at web, mobile, and.. pdf. generate pdf using xetex with oblivoir. have a little problem now.

need study polyglossia a few days soon. dont worry. and thanks to you. I was motivated. really :D