I'll help as soon as You call. I like C++ much more than Assembler :)))
Pentium's Assembler differs from previous processors' versions in use of
registers. Now one can use general purpose registers ( ax, bx, cx, etc )
as index registers ( earlier only si and di were allowed ). This
significantly shortens code and uses stack less agressively. But
crossplatfom application should not use machine code, and Ruby should
not too.

# -----Original Message-----
# From: Yukihiro Matsumoto [mailto:matz / zetabits.com] 
# Sent: 25 ???? 2001 ?. 13:42
# To: ruby-talk ML
# Subject: [ruby-talk:16844] Re: national characters is strings
# 
# 
# Hi,
# 
# In message "[ruby-talk:16843] Re: national characters is strings"
#     on 01/06/25, "Aleksei Guzev" <aleksei.guzev / bigfoot.com> writes:
# 
# |That's good enough. But I'm afraid this could ( not would ) cause 
# |string operations to slow down. I'm not an expirienced Linux etc 
# |programer...
# 
# Ruby already has the ability to handle multibyte character 
# string (mostly Japanese), and we use it everyday.  "we" here 
# means everyone using Ruby.  We don't meet any serious 
# performance problem yet.  Do you?
# 
# |Is it a problem to use Unicode?
# 
# Well, yes and no.  Unicode is a very pragmatic solution works 
# for most of the cases.  So I want to support Unicode on Ruby 
# M17N.  But there IS problem domain where Unicode doesn't 
# work.  I want to help them too.
# 
# Plus, for some languages (e.g. Japanese), conversion cost 
# to/from Unicode can be burden.  If you can handle them 
# without any conversion, it's better.
# 
# |My yesterday effort was in adding three pairs of instance methods to
# |String: dos2koi8!, win2koi8!, win2dos! ( with their "non_bangs" ). 
# |Those methods simply convert their recievers. Those were written in 
# |Intel's Pentium Assembler (because of my preferences :)))
# 
# Since I have no knowledge about koi8, I need help.  But I 
# have no knowledge about Pentium Assembler either. ;-<
# 
# 
# 							matz.
#