International Domain Names facts and dilemmas

1
International Domain Namesfacts and dilemmas

• Elisabeth Porteneuve
• ICANN Accra, Ghana
• 9-14 March 2002

2
IDN - why we are there

• Competition by gTLD registries towards new
  customers
• Natural need of people from the worldwide
  Internet to facilitate their access to the
  Internet domain names
• General dissatisfaction of the worldwide Internet
  with ICANN and its incapability to became
  international body, triggering off a strog
  reactions from various horizons

3
Background to IDN

• ASCII subset  LDH 
• Unicode
• CJK languages and the Traditional Chinese vs
  Simplified Chinese issue
• IETF works

4
ASCII subset  LDH 

• 26 letters, a-z, upper and lowercase alike
• 10 digits, 0-9
• Hyphen-minus  - 
• Label separating period  . 
• Additional rules
• no minus at the beggining or at the end of a
  label
• no empty label,

5
1. ASCII subset  LDH  (cont)

• Necessity to have something to communicate, small
  and easy to remember, practical
• Postal address (cannot have all postmen in the
  world to speak all languages)
• Airlines, airports and landing strips indications
  (security)
• Lingua franca for international gatherings

6
2. Unicode

• The only existing table of all international
  characters, developed for printers industry in
  late 1980s, at the time when computer memory and
  processing were slow and requested for a lot of
  ingenuity to allow new features
• The origins of Unicode are rooted in works on
  unified Han

7
2. Unicode unified Han

• Subset of Chinese, Japanese and Korean (CJK)
  characters, which
• Have identical internal computer point
• Print in Chinese, Japanese or Korean design,
  according to a language context
• May have a similar meaning or not, according to a
  language context
• abc abc abc

8
3. CJK languages

• Not alphabet-based
• Ideographs
• More than 100 thousand
• Each of ideographs is a concept or a word

9
3. Unicode v.s. ISO 10646

• Unicode Consortium
• ISO Working Group responsible for ISO/IEC 10646
  is JTC1/SC2/WG2
• Unicode and ISO 10646 tables are equivalent

10
3. Simplified vs Traditional Chinese

• Peoples Republic of China works on
  simplification of characters, starting in 1950s,
  very complex and long
• Simplified Chinese used by one billion people
  nation
• Traditional still exstensively exist in the
  social life, for its long history and artistic
  value
• Development of standardized translation to Latin
  (cf. Pekin ? Beijing)

11
4. The IETF works on IDN

• Based on Unicode (there is nothing else)
• Technical scope expand today  LDH  38
  characters set into several tens of thousand of
  code points
• Discovery of many problems
• Combinatory effects (none will be able to use
  printer information without knowing which
  labguage script is used)
• Mutual incompatibility in Unicode between Unified
  Han and Chinese language including Simplified
  Chinese

12
Summary of problems and political dilemmas

• Chinese Japanese Korean (CJK)
• Latin Cyrillic Grek
• others but still unknown

13
Problems

• If the usage of mixed letters (code points) is
  allowed the IETF works on Unicode cannot exlude
  it there will be no more any unambigous printed
  URL
• Consumers confusion
• Many doubts for safe electronic commerce
• Combinatory possibilities will increase by factor
  of hundred or thousand a domain name cost to
  those companies willing to have complete
  protection

14
More problems

• The  language  for IDN is undefined
• The printed (paper or screen) information is
  undefined without knowing which script has been
  used
• A printed information does not provide for
  unilateral guessing of company
• How consumer will be able to contact a company if
  the only information he has is printed ?

15
Political problem

• Mutually inpossible to satisfy TC/SC and unified
  Han. Chinese are signatories to ISO10646 does
  it means they give advantage to unified Han over
  TC/SC ? No
• The IETF work on Unicode demonstrate there is a
  clash between Chinese language on one side and
  Korean and Japanese on another
• Accepting Unicode is equivalent to take position
  against Chinese language
• Only Chinese may solve it asking for Unicode
  changes ?

16
Not enough works on languages

• Unicode is a recent pot pourri, initially defined
  for printers industry, gathering not only
  languages, but anything which may be printed
• But there is nothing else

17
Questions

• Do we agree the Unicode is unsuitable for
  Internet Domain Names ?
• If yes, what to do now ?