Site Map

Please note:

You are viewing archival ICANN material. Links and information may be outdated or incorrect. Visit ICANN's main website for current information.

ICANN Meetings in Marrakech, Morocco

Workshop on IDN

25 June 2006

Note: The following is the output of the real-time captioning taken during the Workshop on IDN held on 25 June 2006 in Marrakech, Morocco. Although the captioning output is largely accurate, in some cases it is incomplete or inaccurate due to inaudible passages or transcription errors. It is posted as an aid to understanding the proceedings at the session, but should not be treated as an authoritative record.

>>PAUL TWOMEY: GOOD AFTERNOON. I WILL CALL THE SESSION TO ORDER, IF I MAY.
I'D LIKE TO WELCOME EVERYBODY HERE TO THE SESSION.
MY NAME IS PAUL TWOMEY, I AM THE PRESIDENT AND CEO OF ICANN, AND I WELCOME THOSE PEOPLE WHO ARE PARTICIPATING IN THIS WORKSHOP, NOT ONLY HERE IN MOROCCO BUT THOSE WHO ARE PARTICIPATING THROUGH ONLINE MECHANISMS.
THE SESSION THIS AFTERNOON IS ONE OF OUR GROWING NUMBER OF WORKSHOPS ON INTERNATIONALIZED DOMAIN NAMES. AND I THOUGHT I'D JUST MAKE A FEW SORT OF BASIC OBSERVATIONS, INTRODUCING THIS.
FIRST OF ALL, THIS IS FAR FROM A SIMPLE TOPIC. IT HAS POLICY COMPLEXITY, AS I THINK WE WILL HEAR FURTHER TODAY, AND WE CERTAINLY HAVE SEEN FROM RECENT WORK OUT OF THE IETF AND IAB, AUTHORS OF WHOM ARE IN THIS ROOM.
THIS IS NOT A SIMPLE TECHNICAL PROBLEM EITHER.
BUT, BUT, THIS IS A CHALLENGE THAT NEEDS TO BE MET TO ADDRESS THE NEEDS OF AN INTERNET COMMUNITY WHICH IS INCREASINGLY USING KEYBOARDS THAT ARE NOT IN ROMAN CHARACTERS OR IN ASCII, AND IT IS A TASK THAT WE NEED TO WORK TOGETHER TO ACHIEVE.
THE SECOND POINT I WOULD LAKE TO MAKE IS THAT WHILE THIS IS A CONVERSATION THAT'S GOING ON IN MANY PLACES, IN MANY FORA AROUND THE WORLD, INTERNATIONALIZED DOMAIN NAMES EVENTUALLY COMES DOWN TO THE ICANN COMMUNITY. IT COMES DOWN TO THE ICANN COMMUNITY BECAUSE AT THE SECOND LEVEL, IT IS PUT IN PLACE BY TOP-LEVEL DOMAINS, AND AT THE ROOT IT GETS PUT IN PLACE, IT WILL GET PUT IN PLACE THROUGH THE ICANN AND IANA PROCESSES.
SO I SUPPOSE THAT AT SOME LEVEL, IF YOU WANT TO BE IN THE IDN.IDN SPACE, YOU HAVE TO BE HERE, AND THE DISCUSSION IS IMPORTANT.
I THINK THE THIRD POINT I'D MAKE IS THAT I PARTICULARLY WELCOME, HAVING SAID THAT POINT ABOUT THE DNS, THE MANY PEOPLE WHO ARE PARTICIPATING IN THIS DIALOGUE WHO ARE COMING FROM DIFFERENT LAYERS OF THE INTERNET. I CAN SEE PEOPLE IN THE ROOM FROM THE BROWSER COMMUNITY AND OTHERS. AND I THINK THAT'S BEEN AN IMPORTANT PART OF THE PROCESS, MAKING THIS ACTUALLY WORK.
WE WILL TODAY, IN THE PRESENTATION, HAVE BOTH DISCUSSION AROUND TECHNICAL ASPECTS, BUT I THINK IMPORTANTLY, ALSO, AROUND SOME OF THE POLICY ISSUES, AND HOW DO WE THINK ABOUT SOME OF THE POLICY ISSUES WHEN IT COMES TO INTERNATIONALIZED DOMAIN NAMES AT THE TOP LEVEL. PARTICULARLY AS IT RELATES TO COMING OUT OF OUR COUNTRY CODE NAME SUPPORTING ORGANIZATION AND THE GENERIC NAME SUPPORTING ORGANIZATION.
THIS WEEK, THE PRESIDENT'S ADVISORY COMMITTEE ON IDNS WILL BE CONVENING AGAIN, TALKING THROUGH THE TIMETABLE FOR TECHNICAL TRIALS, FOR INTRODUCING -- FOR TESTING IDNS AT THE ROOT LEVEL. THERE'S A LOT OF WORK GOING ON THERE, PARTICULARLY BY TINA DAM AND OTHERS, TO ACTUALLY WORK THROUGH A TIMETABLE. THAT IS A PROCESS FOR THIS WEEK.
THERE IS ALSO DISCUSSION AROUND -- IN VARIOUS JOINT WORKING GROUPS, ON SOME OF THE POLICY AND POTENTIAL FIRST STEPS THAT COULD BE TAKEN ON THOSE LINES THIS WEEK, AND WE SHOULD HEAR MORE I THINK BY FRIDAY OR SATURDAY. BY FRIDAY.
I'D LIKE TO JUST GIVE YOU SOME IDEA OF THE AGENDA FOR TODAY. SOME OF YOU MAY HAVE A PIECE OF PAPER IN FRONT OF YOU. AND IF YOU DO, I HAVE NO EXCUSE, MY APOLOGIES, I AM GOING TO READ BACK TO YOU SOMETHING YOU HAVE IN FRONT OF YOU. BUT FIRST OF ALL, CARY KARP FROM .MUSEUM IS GOING TO PRESENT SOME INFORMATION, BASIC IDN INFORMATION, BETWEEN 3:15 AND 3:45. MAY I READ THE TITLE? I WON'T TAKE AWAY CARY'S OPENING LINE, BUT HE'S GOING TO TALK MORE. IT LOOKS VERY INTERESTING. THEN THERE WILL BE A REVIEW OF THE GNSO'S PRELIMINARY POLICY ISSUES REPORT BETWEEN 3:45 AND 4:45 AND THAT PRESENTATION WILL BE GIVEN BY BRUCE TONKIN, THE CHAIR OF THE GNSO COUNCIL. ALSO, I THINK HE WILL BE ASSISTED BY HIRO HOTTA WHO IS THE CCNSO IDN REPRESENTATIVE FROM JPRS, AND ALSO OLOF NORDLING FROM THE ICANN STAFF.
WE'LL HAVE AN AFTERNOON BREAK, A TEA BREAK, AT 4:45 TO 5:00. AND THEN AT 5:00, WE WILL HAVE A COUPLE OF SPEAKERS CONVENED BY THE GOVERNMENTAL ADVISORY COMMITTEE'S IDN WORKING GROUP, AND THEY WILL BE CONVENING SOME PEOPLE TO SPEAK IN THE GENERAL -- IN THE GENERAL ARENA.
THEN THERE WILL BE A PROCESS HAVING SOME CLOSING REMARKS AND THE REVIEW OF THE PROCESS FORWARD. AND WE SHOULD BE FINISHED HERE BY 6:00.
JUST TO DRAW YOUR ATTENTION, THERE IS ALSO A WORKSHOP ON SIMILAR TOPICS ON JUNE THE 27TH, ON TUESDAY, AT 3:30. 3:30 TO 7:30, TWO SESSIONS, ONE ON TECHNICAL IDN ACTIVITIES WHERE WE WILL HAVE PRESENTATIONS BY TINA DAM, JOHN KLENSIN, THOMAS NARTEN, MICHEL SUIGNARD AND DR. LIAN FROM TWNIC. AND WE WILL ALSO HAVE A SESSION, TOO, A PRESENTATION ON AFRICAN AND MIDDLE EASTERN IDN ACTIVITIES.
SO WE HAVE TWO PARTS, WORKING SESSIONS THIS WEEK, AND I THINK THIS MEETING WE SHOULD EXPECT TO SEE SOME CONCRETE STEPS COMING FORWARD IN PART OF THIS PROCESS.
SO I WOULD LIKE TO, FIRST OF ALL, FINISH UP BY SAYING THANK YOU TO ALL OF YOU WHO ARE INVOLVED IN THIS PROCESS. THANK YOU TO ALL OF YOU WHO ARE OBVIOUSLY KEEN TO SEE IDNS IMPLEMENTED INTO THE ROOT. I THANK YOU FOR TWO REASONS. ONE, OBVIOUSLY THERE IS A NEED FROM WHERE YOU ARE COMING FROM BUT ALSO THERE IS A CERTAIN DEGREE OF PATIENCE WHICH YOU ARE SHOWING WHICH I THINK IS IMPORTANT AS WE WORK THROUGH CAREFULLY TO ENSURE THAT THIS ACTIVITY IS DONE IN A WAY WHICH IS TECHNICALLY RESPONSIBLE, AND ALSO DONE IN A WAY WHICH ACTUALLY WORKS IN A POLICY SENSE.
AND I MIGHT JUST FINISH AT THAT AND ASK CARY TO TAKE THE NEXT SESSION.
>>CART KARP: WE'RE ASSUMING THAT THERE ARE PEOPLE IN THE AUDIENCE TODAY WHO HAVE NO PRIOR FAMILIARITY WITH IDN, SO I'M GOING TO BE STARTING SIMPLY BY GOING THROUGH SOME OF THE ABSOLUTE MOST BASIC CONCEPTS.
I CAN ASSURE YOU, THERE WILL BE SOMETHING IN THIS TO CONFUSE EVERYBODY, SO YOU WON'T BE DISAPPOINTED, EVEN IF YOU HAVE A GOOD UNDERSTANDING OF THE BASIC NOTIONS OF ALL OF THIS. BUT AGAIN, ASSUMING THAT THERE ARE PEOPLE HERE WHO DON'T, AND BECAUSE THIS SUBJECT RAPIDLY BECOMES VERY DIFFICULT TO FOLLOW, AND BECAUSE BEFORE THE SECOND WORKSHOP ON TUESDAY IS OVER, WE WILL HAVE PLUMBED THE ABSOLUTE DEPTHS OF ALL OF THIS, HERE IS WHAT, AT THE OUTSET, IS USEFUL TO KNOW FOR -- FROM THE PERSPECTIVE OF THE PROSPECTIVE -- LET'S SEE, AM I GOING TO GET THIS? OKAY. THAT ALLOWS YOU TO SEE WHAT I HAVE, BUT -- OKAY. NOW I'VE GOT IT. THERE WE GO. NOW WE BOTH SEE IT.
THIS IS FROM THE PERSPECTIVE OF SOMEBODY WHO WISHES TO REGISTER AN IDN IN THE BELIEF IT'S EXACTLY THE SAME THING AS REGISTERING ANY DOMAIN NAME WHICH IT ACTUALLY HAS TO BE IN THE PERCEPTION OF WHOEVER IS REGISTERING, BUT FOR ONE VERY IMPORTANT REASON IS NOT. AND THAT IS WHAT I INTEND TO EXPLORE HERE.
OKAY.
YOU JUST NEED TO KNOW THE NAME YOU WANT TO REGISTER, BUT AGAIN, THERE ARE THINGS BEYOND THAT THAT WILL MAKE THAT PROCESS LESS PAINFUL OR CONFUSING THAN IT MIGHT OTHERWISE BE.
COMPUTERS HAVE NO TROUBLE DEALING WITH ANY LANGUAGE YOU MIGHT WANT TO THROW AT THEM. BUT A GIVEN COMPUTER, ANY GIVEN WORKING ENVIRONMENT, IS VERY LIKELY TO BE CONFIGURED QUITE SPECIFICALLY TO DEAL WITH ONE OR A SMALL NUMBER OF LANGUAGES.
KEYBOARDS, AS PAUL MENTIONED, ARE HIGHLY LANGUAGE SPECIFIC, AND THE FONTS THAT ARE AVAILABLE, THE WAY THE CHARACTERS ARE REPRESENTED IN A DISPLAY, ARE ALSO VERY A SPECIFIC COMPONENT OF ANY LOCALE CONFIGURATION AND VERY FEW COMPUTERS ARE CAPABLE WITHOUT SOME ERUDITE PREPARATION ON THE PART OF THEIR OWNERS TO DEAL WITH ANYTHING PARTICULARLY DISTANT FROM WHAT IS NORMAL IN THE AREA IN WHICH WE ARE WORKING.
AN INTERNATIONALIZED DOMAIN NAME THAT MAKES PERFECT SENSE TO A USER IN LOCALE CAN BE OF LIMITED UTILITY TO SOMEONE IN ANOTHER AND IT CAN EVEN BE OF LIMITED UTILITY TO ONE'S OTHER THAN. WHAT HAPPENS IF YOU PRINT UP YOUR PROUDLY IDNS ON A BUSINESS CARD, GIVE THAT BUSINESS CARD TO SOMEONE WHO IS THEN INCAPABLE OF TYPING WHAT HE OR SHE SEES ON THE CARD? WHAT HAPPENS WHEN YOU YOURSELF ARE AT SOME DISTANCE FROM HOME IN AN INTERNET CAFE AND DISCOVER THAT THE GEAR IN THAT CAFE DOESN'T ALLOW YOU TO ACCESS YOUR OWN SYSTEMS?
LOCALE SENSITIVITY ISN'T SPECIFIC TO IDN. E-MAIL PREDATING IDN CONTEMPORARY WITH IT VERY OFTEN POPS UP IN OUR MAILBOXES AS A BUNCH OF ABSOLUTE GOBBLEDYGOOK. WE CAN ASSUME THIS IS SUPPOSED TO BE LEGIBLE TEXT IN SOME LANGUAGE BUT WE DON'T, ABSENT KNOW-HOW ABOUT HOW TO FIND THAT OUT, KNOW WHAT LANGUAGE THAT MIGHT BE, AND WE ASSUME IT'S SPAM AND WE JUST DISCARD IT, AND THE MOST LIKELY REASON FOR THIS IS THAT THE DISPLAY REPERTOIRE, THE THINGS THAT YOUR COMPUTER IS PREPARED TO DISPLAY, ARE NOT ADEQUATE FOR REPRESENTING WHAT IS IN THAT MESSAGE OR EQUALLY LIKELY, THAT MESSAGE ISN'T PROVIDING THE INFORMATION NECESSARY FOR YOUR COMPUTER TO BE ABLE TO DO WITH IT WHAT IT NEEDS TO DO.
THERE ARE WHOLE BUNCHES OF CHARACTER ENCODING SYSTEMS IN CURRENT USE, AND THIS IS NOTHING MORE THAN A LIST, COLUMN "A" OF ALL THE CHARACTERS THAT NEED TO BE REPRESENTED, DEALT WITH, PROCESSED, AND COLUMN "B," A NUMBER, UNIQUE IN THAT SYSTEM, FOR EACH ONE OF THESE CHARACTERS.
HOWEVER, BECAUSE THERE IS -- YOU CAN COUNT FROM ONE TO 256 IN ONE WAY AND ONE WAY ONLY, IF YOU HAVE LOTS AND LOTS AND LOTS MORE CHARACTERS THAN THAT, YOU WILL NOTE THAT SOME ENCODING SYSTEMS USE A GIVEN NUMBER TO DESIGNATE ONE CHARACTER AND OTHER ENCODING SYSTEMS USE THE SAME NUMBER TO DESIGNATE ANOTHER CHARACTER. AND THAT'S WHAT IS PROBABLY THE REASON WHY THIS E-MAIL APPEARS AT GOBBLEDYGOOK.
THERE IS, HOWEVER, ONE LARGE CHARACTER SET, THE UNIVERSAL CHARACTER SET, THAT INCLUDES HUNDREDS OF THOUSANDS OF DIFFERENT CHARACTERS. IT IS STILL GROWING. AND A TECHNICAL TERM THAT YOU DO NOW NEED TO RECOGNIZE IS THAT IN THE UNICODE FRAME OF REFERENCE, WHICH IS ONE OF THE TWO WAYS IN WHICH THIS UNIVERSAL CHARACTER SET IS PUT FORWARD, A NUMBER ASSIGNED TO A CHARACTER IS CALLED A CODE POINT.
OKAY. JARGON -- THERE ARE A COUPLE OF THINGS YOU NEED TO KNOW. RECOGNIZING THE TERM CODE POINT, IT JUST MEANS THE NUMERICAL POSITION OF SOMETHING IN UNICODE HAS TO BE KEPT IN MIND.
AND SOMEONE WHO WISHES TO REGISTER AN IDN IS LIKELY GOING TO COME WITH THE UNICODE REPRESENTATION OF THAT NAME. THIS IS WHAT IT LOOKS LIKE WHEN I WRITE IT, THIS IS WHAT IT LOOKS LIKE WHEN I READ IT, THIS IS WHAT I EXPECT IT TO BE.
OKAY?
FINE. HOWEVER, THERE ARE SEVERAL INTERNAL REPRESENTATIONS OF UNICODE. THERE IS NOT JUST -- THERE'S DISPLAY CHARACTER, WHICH IS ITSELF, NO MATTER WHAT. BUT THERE ARE SEVERAL WAYS TO REPRESENT THAT INTERNALLY. YOU DON'T NEED TO KNOW ANYTHING ABOUT THAT OTHER THAN THE TERM UTF-8 -- IS MICHAEL HERE UNICODE TRANSFORMATION FORMAT 8. IS THAT RIGHT? NOBODY GRIMACING, SO EITHER YOU KNOW THIS LESS THAN I DO OR....
IT'S ONE OF SEVERAL REPRESENTATIONS AND YOU WILL SEE UTF-8 AS A DESIGNATION WHICH YOU MAY REGARD AS SYNONYMOUS TO UNICODE IN SOME OF THE ARCANE ASPECTS OF IDN. EVERY TLD REGISTRY WILL PUBLISH A LIST OF HERMETIC CHARACTERS IF FOR NO OTHER REASON THAN TO TELL YOU SORRY, WHAT WAS GOOD IN THE EARLY DAYS WITH THE DNS IS WHAT WE ARE SUPPORTING AND WE ARE NOT SUPPORTING ANYTHING ELSE. IF THE REGISTRY IS SUPPORTING SOMETHING ELSE, THEY ARE VERY WELL ADVISED TO TELL YOU EXACTLY WHAT THAT IS.
OKAY. AND QUITE OFTEN YOU WILL SEE THE UNICODE CODE POINTS THAT CORRESPOND TO EACH OF THESE PERMITTED DISPLAYABLE CHARACTERS.
AT SOME POINT IF YOUR COMPUTER IS USING AN ENCODING SYSTEM OTHER THAN UNICODE, WHICH ACTUALLY IS QUITE A COMMON STATE OF AFFAIRS, IN WESTERN EUROPE IT CERTAINLY IS, A CONVERSION HAS TO BE MADE FROM WHATEVER SCHEME IS BEING USED BY YOUR MACHINE TO UTF-8. THAT'S FORTUNATELY NOT ONE OF THE THINGS YOU NEED TO WORRY ABOUT.
THERE IS, HOWEVER, SOMETHING THAT HAPPENS NEXT, WHICH IS THE CORE OF WHAT IT IS THAT YOU DO NEED TO WORRY ABOUT, ALTHOUGH DEALING WITH IT, HOW TO SOLVE THE PROBLEM INHERENT IN IT, ISN'T NECESSARILY YOURS. AND THAT IS THAT INTERNATIONALIZED DOMAIN NAMES ARE NOT STORED IN THE DOMAIN NAME SYSTEM IN UNICODE FORMAT.
THE DNS WAS DESIGNED TO USE AND STILL USES AND PROBABLY FOR AS LONG AS IT'S CALLED THE DOMAIN NAME SYSTEM WILL REMAIN USING ASCII, AMERICAN STANDARD CODE FOR INFORMATION INTERCHANGE. OKAY?
ASCII PERMITS THE LEGIBLE REPRESENTATION OF THE 26-LETTER BASIC LATIN ALPHABET, A TO Z, UPPER CASE, LOWER CASE. THE TEN EUROPEAN DIGITS, ZERO TO NINE, AND A FEW OTHER SYMBOLS OF WHICH THE ONLY ONES NORMALLY RECOGNIZED BY MOST REGISTRIES ARE GOING TO BE THE DASH AND THE DOT, AND DOTS DON'T REALLY COUNT AS PARTS OF DOMAIN NAME LABELS. THEY SEPARATE THEM.
OKAY. ALL OTHER UNICODE CHARACTERS NEED TO BE TRANSFORMED INTO UNICODE BEFORE THEY CAN ACTUALLY BE ENTERED INTO THE DNS. SO FOR EVERY IDN -- YOU SPEAK OF YOUR IDN AS THOUGH IT WERE A NAME. IT IS NOT. IT IS TWO NAMES, OF WHICH THE ONE APPEARS TO YOU AND HOPEFULLY WILL APPEAR TO THE USERS OF THAT NAME AS NO DIFFERENT THAN AN ASCII STRING EXCEPT IT MAY BE DECORATED OR IT WILL BE AN ENTIRELY DIFFERENT SCRIPT, BUT A NAME IS A NAME.
IN THE DNS, WHAT YOU HAVE ACTUALLY REGISTERED IS SOMETHING QUITE DIFFERENT, OFTEN UTTERLY UNINTELLIGIBLE, ILLEGIBLE TO A HUMAN AT LEAST. ANY ONE OF THESE ENCODED VERSIONS OF A UNICODE STRING IS GOING TO BE TAGGED BY AN XN DASH DASH PREFIX. THIS TELLS SOFTWARE THAT KNOWS WHAT TO DO WITH THE IDN THAT NOW IT'S TIME -- THAT IT IS NOW TIME TO DO THAT. OKAY.
A WEB BROWSER, FOR EXAMPLE, WHICH IS WHERE IDN IS USED MOSTLY, SEES THE XN DASH DASH, CONVERTS IT INTO SOMETHING -- SEES THE UNICODE STRING, CONVERTS IT INTO THIS XN DASH DASH THING, SENDS IT OFF TO DNS, GETS THE ANSWERS IT IS LOOKING FOR FROM THE DNS LOOKUP AND CONVERTS IT BACK TO WHAT YOU EXPECT TO SEE.
THERE ARE A COUPLE OF SUCH SCHEMES THAT HAVE BEEN TRIED DURING THE COURSE OF THE DEVELOPMENT OF IDN. THE ONE THAT'S CURRENTLY USED, IT IS THE ONLY STANDARDS-BASED SCHEME IS SOMETHING CALLED PUNYCODE, AND XN DASH DASH INDICATES PUNYCODE.
THE TWO DASHES, THE DASH IN THE THIRD POSITION AND THE DASH IN THE FOURTH POSITION, SAY THAT THIS IS NOT REALLY WHAT YOU THINK IT IS. THERE ARE A NUMBER OF SCHEMES THAT WE MAY BE INVOKING HERE FOR VARYING PURPOSES. XN SAYS THAT IT IS PUNYCODE.
THE INITIAL INTENTION WAS THAT NO USER, NO HUMAN BEING WOULD EVER SEE A PUNYCODE STRING, SO IT DIDN'T REALLY MATTER HOW LEGIBLE IT WAS, BUT THE SAD FACT OF THE MATTER IS THAT SUCH SITUATIONS STILL ABOUND. AND IF YOU DON'T WANT TO BE UTTERLY CONFOUNDED BY ALL OF THIS, RECOGNIZING PUNYCODE WHEN IT'S LOOKING YOU IN THE FACE IS A REASONABLE SKILL TO ACQUIRE.
YOU CAN SEE THIS QUITE EASILY IF WE'RE LOOKING AT A STRING OF ALMOST PURE ASCII BUT SAY WITH AN ACCENT OVER ONE OF THE LETTERS SOMEWHERE. SO LET'S SEE. DO I HAVE A -- NO, I DON'T HAVE ANY KIND OF A POINTER HERE. IF YOU TAKE A LOOK AT THE XN DASH DASH STRING IN THE MIDDLE OF IT, THE LINE ABOVE IT, THE WORD NASENBAR WHICH IS A GERMAN WORD, IF WE TURN THAT FROM THE UNICODE REPRESENTATION WITH THAT UMLAUTED "A" TO THE PUNYCODE REPRESENTATION WE GET THE XN DASH DASH, WE ALWAYS GET THAT, AND THEN WE GET ALL OF THE ASCII CHARACTERS, THE NASENBR, BUT THE UMLAUTED "A" IS STRIPPED OUT, AND THEN YOU HAVE A HYPHEN, AND THEN COMES THE CODE, THE STUFF THAT HUMAN BEINGS DON'T READ WHICH SIMPLY SAYS THERE IS AN UMLAUTED "A" IN THE POSITION IT IS IN THIS STRING.
IT WORKS STILL, YOU CAN KIND OF SEE WHAT YOU ARE LOOKING AT IF YOU HAVE A FEW MORE DECORATED CHARACTERS. THIS IS THE SAME WORD, DIFFERENT LANGUAGE, EXTRA DECORATED LETTER, AND YOU NOTE THAT THE THING BETWEEN THE TWO DASHES AND THE ONE DASH BECOMES MORE COMPACT AND THE THING TO THE RIGHT OF THE SECOND DASH -- OF THE THIRD DASH, THE SINGLE DASH THERE, BECOMES LONGER. AND SOMETHING THAT IS RELEVANT, CERTAINLY TO THE DEVELOPMENT OF IDN POLICIES, AND IT'S OFTEN OVERLOOKED, IS THAT IT IS THE NATURE OF PUNYCODE THAT THE SHORTEST POSSIBLE PUNYCODE SEQUENCE WILL BE EIGHT ASCII CHARACTERS LONG. THAT MEANS THAT IF SOMEBODY SAYS, "I WANT TO HAVE A SINGLE CHARACTER LABEL, EXTERNAL TO ASCII," THAT MEANS THAT THEY ARE LOOKING AT SEVEN LETTERS, SEVEN CHARACTERS IN THE DNS.
SO IF WE DECIDED THAT WE HAD PROBLEMS WHEN ASCII TLD LABELS WERE ENTERED INTO THE ROOT ZONE. DNS THAT HAD MORE THAN THREE -- THREE CHARACTERS, THEN FOUR, FIVE SIX -- SIX IS WHERE IT STOPS RIGHT NOW, THE MOMENT YOU START USING IDN FOR THIS, WE ARE AT THE SEVEN MARK AND WE HAVEN'T GOTTEN PAST A SINGLE CHARACTER, WHICH WE ARE ACTUALLY NOT SUPPOSED TO BE REGISTERING ANYWHERE. CERTAINLY NOT IN THE ROOT.
SO THESE PUNYCODE STRINGS CAN HAVE SOME POTENTIAL FOR BEING WAY LONGER THAN ANYTHING WE HAVE EVER DEALT WITH. AND ONE OF THE THINGS THAT NEEDS TO BE TESTED, THIS TECHNICAL TEST, IS WHAT DOES THE APPLICATION SPACE DO WHEN WE START THROWING THESE LONG ASCII SEQUENCES AT IT IN THE TLD SLOT.
WHAT'S CAPABLE OF BEING ABSORBED? AND BEFORE ICANN SANCTIONS A PRODUCT THAT IS GOING TO BE OFFERED, NOT LEAST TO THE GOVERNMENTS OF THE WORLD, WE NEED TO BE REASONABLY SURE THAT THAT PRODUCT IS ACTUALLY USABLE.
SO SOMETHING TO KEEP IN MIND.
WHEN WE ARE NOT -- WHEN WE LEAVE ASCII ENTIRELY, WHEN WE LEAVE THE LATIN SCRIPT ENTIRELY, THIS IS THE SAME WORD AGAIN, IN CYRILLIC SCRIPT. YOU WILL NOTICE THAT THE PUNYCODE SEQUENCE IS TRULY AND GENUINELY ILLEGIBLE. AND WE'LL NOTE HERE AGAIN ANOTHER PROBLEM. THE CYRILLIC STRING HAS WHAT LOOKS LIKE A LOWER CASE "O" IN IT AND WHAT LOOKS LIKE A LOWER CASE "A" IN IT BUT THEY ARE NOT LOWER CASE. THAT'S NOT A LATIN "O," NOT A LATIN "A," AND THE VISUAL SIMILARITIES THERE HAVE BEEN EXPLOITED FOR LESS THAN LAUDABLE PURPOSES. AND IN FACT, THAT EXPLOITATION IS LARGELY THE REASON WHY WE ARE STANDING HERE TODAY TALKING THROUGH THIS AND WHY WE'VE HAD ALL OF THESE IDN WORKSHOPS.
BECAUSE OF THESE SECURITY CONCERNS THAT ATTACH TO THIS, MANY WEB BROWSERS WANT TO BE ABSOLUTELY CERTAIN THAT YOU ARE NOT RENDERING YOURSELF UNNECESSARILY VULNERABLE TO THIS CONFUSION. SO THEY MAY ASK YOU WHAT SCRIPTS DO YOU REALLY INTEND TO USE HERE? WE'LL SHOW YOU IDN IN THOSE SCRIPTS IF YOU TELL US TO DO SO, BUT OTHERWISE, WE'RE JUST GOING TO SHOW YOU PUNYCODE BECAUSE PUNYCODE IS -- IT'S WHAT NOBODY EXPECTS TO SEE, SO YOU MAY BE -- IT WILL SIGNAL YOU, THE SAME WAY THE XN DASH DASH IS INTENDED TO TELL IDN-AWARE SOFTWARE THAT NOW IT'S TIME TO DO THE IDN TRICKS, XN DASH DASH DISPLAYED TO A USER SAYS SOMETHING IS WRONG HERE, WHICH IS ACTUALLY VERY USEFUL INFORMATION. AND THERE ARE VARYING APPROACHES IN THE SOFTWARE DEVELOPER COMMUNITY TO HOW MUCH OF THAT INFORMATION SHOULD BE DISPLAYED TO THE USER, UNDER WHAT CIRCUMSTANCES IT'S NOT TO BE DISPLAYED. SO THERE IS A TREMENDOUS HETEROGENEITY OF THE WAY IDN IS APPLIED WHICH ALSO MAKES IT VERY DIFFICULT TO FORMULATE ROBUST GENER!
ALLY APPLICABLE POLICIES ABOUT ALL OF THIS.
OKAY.
THERE'S SOMETHING ELSE THAT HAPPENS WITH THIS, WHICH IS ACTUALLY KIND OF CUTE. SOMEBODY WAS WRITING A PAPER ABOUT THE DANGERS OF ALL OF THIS, AND SAID SOONER OR LATER SOMEBODY IS GOING TO DISCOVER THAT XN DASH DASH GIBBERISH IS ACTUALLY A MEANINGFUL UNICODE SEQUENCE, AND THEN DISCOVER THAT XN DASH DASH GIBBERISH DOES, IN FACT, DECODE.
IT IS THAT ARABIC SEQUENCE WHICH IS -- IT'S A NONSENSE SEQUENCE OF ARABIC CHARACTERS, BUT NONETHELESS, THE ALGORITHMIC PROCESS WHEREBY THIS INTERCONVERSION IS CONDUCTED ALLOWS GIBBERISH TO BE CONVERTED INTO WHAT SOMEONE WHO DOESN'T UNDERSTAND ARABIC SCRIPT WILL ASSUME MAY BE A LEGITIMATE ARABIC BUSINESS ENTITY. WHAT DO I KNOW?
THIS IS A MEANINGLESS STRING, BUT SOONER OR LATER IF THERE'S GOING TO BE A SITUATION WHERE SOMEBODY DISCOVERS XN DASH DASH VALUABLE TRADEMARK, EVEN MORE FUN STARTS AT THIS POINT. I BET THERE ARE PEOPLE IN THIS ROOM WHO HAVE ALREADY DISCOVERED THEM AND THEY ARE KEEPING QUIET ABOUT IT.
BECAUSE OF ALL OF THIS THERE ARE FACILITIES ON THE NET, THEY ABOUND, ACTUALLY, THAT INTERCONVERT PUNYCODE REPRESENTATION OF AN IDN STRING WITH ITS UTF-8, REPRESENTATION, AND THE HEAVYWEIGHT OF THEM ALL -- I WON'T DEMONSTRATE THIS HERE. THIS MATERIAL WILL BE ONLINE, SO YOU CAN FIND IT -- IS THAT MIDDLE LINK THERE. AND OTHERWISE JUST USE YOUR FAVORITE SEARCH ENGINE AND TAKE A LOOK FOR UTF-8 IDN CONVERTERS. MANY REGISTRIES HAVE THEIR OWN, BUT THEY ARE ALL BASED ON A VERY SMALL NUMBER OF WHAT ARE CALLED IDN LIBRARIES, AND THE ONE I HAVE THERE WITH THE EXPLICIT LINK IS I BELIEVE THE VERY FIRST SUCH IMPLEMENTATION AND THE ONE THAT IS MOST WIDELY USED, AGAIN TO THE BEST OF MY KNOWLEDGE.
THERE ARE OTHER SOFTWARE UTILITIES THAT MAKE LIFE EASY DEALING WITH LOCALES OTHER THAN YOUR OWN. THE MOST INTERESTING ONE IS A KIND OF SOFTWARE CALLED CHARACTER PICKERS, WHICH SIMPLY ASSUMES THAT YOUR KEYBOARD ISN'T ADEQUATE FOR ENTERING THE TEXT THAT YOU NEED TO ENTER AND WILL ALLOW YOU JUST WITH YOUR MOUSE TO ESSENTIALLY BROWSE THROUGH A UNICODE CODE CHART AND CLICK ON THE CHARACTERS THAT YOU WISH TO APPEAR IN THE STRING. SOME OF THIS STUFF IS ACTUALLY QUITE SOPHISTICATED, VERY, VERY USEFUL. SO THERE'S PROBABLY A WORKSHOP TO BE HAD ON HOW TO DEAL WITH IDN, RECOGNIZING THE SEVERE LIMITATIONS OF ANY GIVEN WORKING ENVIRONMENT TOWARD THAT END.
IF YOU EXPECT YOUR IDN -- YOU HAVE THIS IDN MEANINGFUL TO YOU, YOU HAVE COME TO TERMS WITH ALL THE ARCANE ASPECTS -- WHOOPS -- ALL THE ARCANE ASPECTS OF THE SOFTWARE ENVIRONMENT, AND AGAIN, YOU PRINT THE BUSINESS CARD. YOU EXPECT TO COMMUNICATE YOUR IDN ACROSS SIGNIFICANT LOCALE BOUNDARIES, A COUPLE OF THINGS ARE GOING TO HAPPEN. BUT IF YOU REALLY TRULY WISHED FOR THAT NAME TO BE USABLE IN AS WIDE A CONTEXT AS POSSIBLE, YOU ARE GOING TO NEED TO DO SOMETHING IN ADDITION TO SIMPLY REGISTERING AN IDN NAME.
THE OBVIOUS THING, THE THING THAT ALMOST EVERYBODY DOES, AND IT ACTUALLY IS BREAKING THE DEVELOPMENT OF IDN AS A -- AS A VIABLE FACET OF THE DOMAIN NAME INDUSTRY, IS THAT YOU SIMPLY HAVE THE CLOSEST ASCII EQUIVALENT.
SO IF YOU HAVE A NAME THAT REQUIRES AN UMLAUT SOMEWHERE, YOU JUST HOLD YOUR NOSE, FORGO THE UMLAUT, IT'S NOT WHAT YOUR NAME IS SUPPOSED TO LOOK LIKE BUT IT'S ASCII AND EVERYBODY CAN USE IT. EVERYBODY CAN STILL USE IT. WE'RE STILL NOT IN A SITUATION WHERE IT CAN BE ASSUMED THAT MOST PEOPLE ARE USING IDN-AWARE SOFTWARE. NOT EVEN IDN-AWARE BROWSERS. THAT WILL CHANGE MOST LIKELY, OR AT LEAST EVERYBODY IS EXPECTING, IT'S ALMOST A MANTRA, WHEN THIS IS AVAILABLE, WHEN IE-7 IS WIDELY DEPLOYED, PROBLEMS WILL STOP. BUT YOU WILL DISCOVER THAT SOME PROBLEMS MAY STOP BUT OTHER PROBLEMS MAY MANIFEST THEMSELVES. THIS DOESN'T GET EASY.
THE LOCALE SPECIFICITY OF HARDWARE CAN'T BE TRANSCENDED BY SOME SNAZZY NEW PROTOCOLS NOR CAN THE VERY SLOPPILY ENGINEERED MULTIPLICITY OF LANGUAGES AND SCRIPTS BE RENDERED EASILY AMENABLE TO SIMPLE ALGORITHMIC PROCESSING.
ANOTHER THING YOU CAN DO IS SIMPLY PICK THE NUMBER OF TARGET AUDIENCES YOU WISH TO ADDRESS AND REGISTER AN IDN APPROPRIATE TO EACH. THAT'S PROBABLY THE MOST ELEGANT WAY TO DO IT.
AND CERTAINLY WHEN WE HAVE SOLVED THE PROBLEMS THAT ATTACH TO THE APPEARANCE OF IDN IN THE TOP LEVEL, IT WILL BECOME AN EXTREMELY IMPORTANT THING. IDN AT THIS POINT IS SEVERELY HAMSTRUNG BY THE FACT THAT IT IS SOME LEVELS BUT NOT THE TOP LEVEL, AND WE'LL BE TALKING ABOUT THAT IN SIGNIFICANT DETAIL BEFORE THIS WEEK IS OVER.
THERE'S A PRETTY GOOD LIST OF GENERAL IDN RESOURCES AT -- ON THE ICANN SITE.
YOU'LL FIND THAT THERE.
AND LOTS AND LOTS OF THE ISSUES ATTACHING TO SECURITY, CONFUSABILITY IN THE IDN SPACE IS FOUND IN A TECHNICAL REPORT PREPARED BY THE UNICODE CONSORTIUM THERE.
YOU GO TO THE UNICODE SITE, YOU'LL BE ABLE TO CLICK YOUR WAY TO IT, IT'S TR36.
YOU GO TO THE ICANN SITE, THERE'S A SHORT CLICK PATH RIGHT FROM THE WELCOME PAGE INTO THE IDN INFORMATION REPOSITORY.
DOT MUSEUM, WE'VE GOT A WHOLE BUNCH OF STUFF UP THERE AS WELL.
A WORD FROM THE SPONSOR, AS IT WERE.
OKAY, QUESTIONS?
AND THERE CANNOT BE NONE.
AH.
>>VINT CERF: HI, CARY.
IT'S VINT CERF.
JUST A COUPLE OF QUESTIONS ABOUT OUR EXPERIENCES SO FAR.
IN THE COURSE OF WHAT YOU'VE BEEN DOING, AND OTHERS, DO WE HAVE SOME EXAMPLES OF WHAT THE CURRENT SOFTWARE, WHICH HAS NOT YET BEEN MODIFIED FOR IDN, WILL EXPERIENCE WHEN IT ENCOUNTERS IDNS?
I'M THINKING THAT WE NEED SOME WARNING FOR USERS WHO WILL RUN INTO THIS ALMOST CERTAINLY AS WE INTRODUCE IDNS WHO HAVE NOT YET GOT IDN-CAPABLE SOFTWARE.
>>CARY KARP: THERE WILL BE A TO THAT FROM THE HOUSE.
I WOULD LIKE TO MENTION THAT I FEEL VERY COMFORTABLE REVIEWING SOFTWARE IMPLEMENTATIONS AND WILL RESTRICT MY RESPONSE TO NOTING THAT THE MOST RECENT VERSION OF EVERY ONE OF THE -- AT LEAST THE LARGER, BETTER-KNOWN WEB BROWSERS ARE DO THIS RIGHT OUT OF THE BOX, WITH CERTAIN EXPECTATIONS THAT ARE CLEARLY DOCUMENTED.
BUT THE ASSUMPTION THAT EVERYBODY HAS THE MOST RECENT VERSION OF THEIR FAVORITE BROWSER IS A VERY POOR ASSUMPTION.
THAT'S ONE OF THE REASONS WHY IE7 ISN'T GOING TO SOLVE THE WORLD'S PROBLEMS, BECAUSE THERE ARE STILL PLENTY OF WINDOWS 98 IMPLEMENTATIONS.
>>VINT CERF: BROWSERS ARE NOT THE ONLY OBJECTS THAT WILL ENCOUNTER IDNS, PLAINLY. SO THERE'S A LOT OF OTHER SOFTWARE THAT --
>>CARY KARP: THEN WE GET INTO THE REALLY THORNY ISSUES, AND THAT IS THAT THE UTILITY OF IDN, NASCENT THAT IT IS, IS, AT THIS POINT, BEING CONSIDERED ONLY IN THE CONTEXT OF HTTP PLATFORM.
E-MAIL, FOR REASONS THAT I SUSPECT JOHN BOTH WOULD WANT TO AND CAN FAR MORE ELEGANTLY THAN I DESCRIBE, BRINGS WITH IT A WHOLE PACKAGE OF NEW ISSUES.
TO THE RIGHT OF THE "AT" SIGN, YOU ARE AROUND WHERE THE IDN IS APPLICABLE.
TO THE LEFT SIDE, ARE YOU IN A ROUND WHERE IDN IS NOT APPLICABLE.
AND WHAT HAPPENS WHEN ONE OF THE VARIOUS CODING SCHEMES THAT ARE USED FOR UNICODE IN THE BODY OF AN E-MAIL MESSAGE COLLIDES WITH THE INTENDED CONVEYANCE OF THE IDN ENCODING MECHANISM?
SO THE MINUTE WE -- WE ARE NO LONGER TALKING ABOUT THE WEB BROWSER FACET OF THE APPLICATIONS ENVIRONMENT, WE ARE ALL IN SUCH UNPLOWED TERRITORY, IT'S ACTUALLY FRIGHTENING.
>>VINT CERF: IT SHOULD BE FRIGHTENING, BECAUSE ALL THE USERS WHO WILL HOPING THAT WE WILL PUT IDNS INTO THE DOMAIN NAME SYSTEM MAY NOT GET WHAT THEY'RE HOPING FOR AS A CONSEQUENCE UNTIL THE SOFTWARE THAT SURROUNDS THE USE OF IDNS AND THE DOMAIN NAME SYSTEM, ACTUALLY, ARE MODIFIED.
>>CARY KARP: I CAN SERIOUSLY ENVISION A SITUATION WHERE GOVERNMENTAL TRAVEL ADVISORIES WILL INCLUDE INFORMATION ABOUT WHAT YOU NEED TO DO TO SURVIVE IN AN INTERNET CAFE IN WHATEVER COUNTRY IT IS YOU'RE GOING TO.
>>JOHN KLENSIN: VINT, AS USUAL, EVERYTHING IS MORE COMPLICATED.
WE DON'T EVEN FULLY UNDERSTAND WHAT YOUR QUESTION IN TERMS OF ENCOUNTERING AN IDN MEANS.
URLS ARE DEFINED AS ALL ASCII.
SO IF ONE HAS A PIECE OF SOFTWARE WHICH IS ENFORCING TO YOUR URL SPEC, THE ONLY WAY THAT IDN CAN EVER APPEAR IN THAT URL SPEC IS IF SOMEONE TYPES IN THE PUNYCODE.
THERE IS AN IRI SPEC, WHICH IS IMPLEMENTED IN SOME PLACES AND NOT IMPLEMENTED IN OTHERS, WHICH WOULD PERMIT IDNS IN NATIVE SCRIPT TO BE TYPED IN.
IT IS IMPLEMENTED IN SOME PLACES AND NOT IMPLEMENTED IN OTHERS.
E-MAIL ADDRESSES TURN OUT TO BE FAR MORE COMPLICATED THAN THIS, AND WE'VE GOT AN IETF WORKING GROUP THAT IS MAKING GOOD PROGRESS ON IT PARTLY BECAUSE THE PEOPLE WHO WERE INVOLVED UNDERSTOOD FAIRLY EARLY HOW DIFFICULT THE PROBLEMS WERE AND DECIDED TO ADDRESS THEM SERIOUSLY, WHICH IS SOMETHING THAT, UNFORTUNATELY, DID NOT HAPPEN WITH THE IDN EFFORT.
SO THE IDN EFFORT WE ARE NOW TRYING TO CLEAN UP AFTER THE FACT, AND IT'S PROVING DIFFICULT.
WE HAVE SOFTWARE PACKAGES -- WELL, ANYTHING IS CAPABLE OF ACCEPTING PUNYCODE AS LONG AS THE PUNYCODE GETS TYPED.
BUT CONVERSIONS TO AND FROM PUNYCODE ARE TRICKY FOR USERS TO MAKE, AND THE DESIGNERS OF IDNA ASSUMED THAT USERS WOULD NEVER HAVE TO LOOK AT PUNYCODE, WHICH HAS TURNED OUT, FOR SOME OF THE REASONS CARY HAS MENTIONED, IT'S A GOOD THING THAT THEY CAN.
BUT IT'S COMPLICATED.
THE SUPPORT IS NOT GENERAL.
THERE ARE PLACES WHERE WE CAN GET THE NAMES IN, BUT WE CAN'T GET THE NAMES BACK OUT.
ERROR MESSAGES DO STRANGE THINGS.
AND IT'S A PATCHWORK.
AND THE MORE IMPLEMENTERS DECIDE THAT THEY HAVE RESPONSIBILITY TO THEIR USERS FOR NOT LETTING BAD THINGS HAPPEN LIKE SPOOFING, THE MORE COMPLICATED IT GETS.
ONE CAN QUITE EASILY DEVISE CASES WHERE THIS WORKS AND WORKS SMOOTHLY AND THEN STARTS HAND-WAVING, BUT REAL LIFE IS, UNFORTUNATELY, NOT VERY MUCH LIKE THAT.
>>CARY KARP: MICHAEL AND THEN SABINE.
>>MICHAEL EVERSON: MY POINT OR QUESTION IS JUST THE SAME AS THESE OTHERS.
WHAT IS THIS ASCII BUSINESS?
ARE WE REALLY IN THIS DAY AND AGE, 2006, STILL STUCK WITH 7-BIT CRAP AND XN DASH DASH HACKS.
CAN'T WE GROW UP AND USE REAL STUFF?
CAN WE NOT SORT THIS STUFF OUT?
IT SEEMS TO ME THAT THERE'S A WHOLE BUNCH OF SMART PEOPLE AROUND AND THAT, YOU KNOW, NEW PROTOCOLS CAN BE -- IF URL IS BROKEN, MAKE ANOTHER.
FIX IT.
BECAUSE, OTHERWISE, IT'S JUST GOING TO BE A HORROR FOREVER, ISN'T IT?
>>CARY KARP: SABINE, LET'S LET SABINE COMMENT ON THAT.
>>SABINE DOLDERER: MAYBE I WANT TO BE A LITTLE BIT PROVOCATIVE.
EVERYBODY FROM US TALKS ABOUT ASCII AND 7-BIT ASCII.
BUT THE POINTS BEHIND ASCIIS ARE 1S AND 0S AND NOBODY NO LONGER TALKS ABOUT 1S AND 0S, AND THERE ARE DAYS, GOOD OLD DAYS IN COMPUTER DESIGNS, WHERE PEOPLE ARE TALKING ABOUT 1S AND 0S.
AND NOWADAYS, EVERYBODY HAS ACCEPTED THAT 1S AND 0S ARE SOMETHING ELSE, BUT THEY'RE ASCII OR LATIN SCRIPTS.
AND I THINK WE ARE NOW HEADING FOR SOMETHING WHICH IS OBVIOUSLY, IN THE LONG RUN, IDN.
SO I THINK WE ARE IN A MIDDLE STEP, BUT I DON'T THINK WE SHOULD BE REALLY FRIGHTENED ABOUT THAT.
AND WHEN I -- I HAVE A QUESTION TO YOU.
WHEN YOU'RE TALKING ABOUT USER PERCEPTION AND USER PERCEPTION BEING AND BEING IN AN INTERNET CAFE, I THINK THAT SOMETHING WHICH HAPPENED IN THE REAL LIFE ALREADY.
WHEN I, AS A NON-NATIVE ARABIC SPEAKER AND EVEN NON-NATIVE ARABIC READER, I NEVER -- ACTUALLY, I COULDN'T UNDERSTAND THAT -- GO TO AN ARABIC COUNTRY, I HAVE TO ADMIT THAT THERE ARE PLACES WHERE I CAN'T READ ANYTHING. AND NOT ONLY INTERNET CAFES, OBVIOUSLY, BUT ALSO AT THE STREETS AND AT THE SIGNS AND WHEREVER I AM.
AND WHEN I GO IN INTERNET CAFES THERE, I ALSO HAVE TO ADMIT THAT SOMETIMES I AM UNABLE TO TYPE IN ANYTHING.
BUT THAT'S HAPPENED TO ME WHEN I GO TO AN ARABIC COUNTRY.
OBVIOUSLY.
THAT HAPPENED TO ME ALSO WHEN I GO TO A CHINESE COUNTRY.
THAT'S REAL-LIFE EXPERIENCE.
AND WE HAVE THIS EXPERIENCE BEFORE WE ACTUALLY INVENTED INTERNET CAFES AT ALL.
AND, OBVIOUSLY, I CAN TELL YOU, I THINK CHINESE PEOPLE HAVE THE SAME PROBLEM WHEN THEY COME TO GERMANY.
WHEN THEY GO TO AN INTERNET CAFE AND SUDDENLY FIND OUT THAT THEY'RE UNABLE TO TYPE IN CHINESE CHARACTERS, NOT ONLY IN THE ADDRESSING SCHEME, BUT ALSO WHEN THEY ARE WRITING AN E-MAIL OR SOMETHING LIKE THAT.
SO I THINK WE -- IT'S A VERY GOOD MOVE.
IT WAS A VERY GOOD MOVE TO MAKE THE COMPUTER SYSTEMS LOCALIZED FOR PEOPLE USING IT, USING THEIR LANGUAGE, USING THEIR SCRIPTS, TO HELP THEM IF THEY ARE ALREADY ABLE TO WRITE BY HAND, ARE ALSO ABLE TO PUT IT IN A COMPUTER FORM.
BUT I THINK WE SHOULDN'T BE REALLY ASTONISHED ABOUT THAT, THAT THERE ARE COMPUTERS WHO ONLY UNDERSTAND CHINESE SCRIPT, AND THERE WILL BE COMPUTERS UNDERSTANDING ONLY ARABIC SCRIPTS AND OTHERS WILL ONLY UNDERSTAND LATIN SCRIPTS.
SO I THINK -- AND I THINK THAT THAT'S PERFECTLY OKAY, AND I DON'T THINK WE SHOULD MIX THAT UP.
SO I THINK WE WILL SEE LOCALIZATION WHEN WE ARE USING IDN.
AND I THINK THAT'S GOOD, IT'S REALLY GOOD THAT PEOPLE ARE ABLE TO USE THEIR LOCAL CHARACTER SETS ON THE COMPUTER.
AND THAT I AM UNABLE TO USE THIS COMPUTER ALSO, BEING -- THAT'S -- UNFORTUNATELY, THAT'S THE CASE.
BUT ON THE OTHER HAND SIDE, WHY SHOULD I USE EVERY COMPUTER I AM STANDING FOR.
>>CARY KARP: I ACTUALLY FIND THAT QUITE EXCITING. AND THIS IS JUST PROOF THAT THE INTERNET IS GAINING ENTRY INTO ALL THE MORE DOMAINS OF HUMAN ACTIVITY.
AND, AGAIN, AS I SAY, IF YOU'RE GOING ELSEWHERE, THERE ARE THINGS YOU NEED TO KNOW.
AND AMONG THEM, IF AMONG THEM IS HOW TO SURVIVE IN AN INTERNET CAFE, THAT'S REALLY COOL.
THERE ARE MORE SIGNIFICANT PROBLEMS WHEN THINGS LIKE RESOURCE IDENTIFIERS APPEAR IN OFFLINE MEDIA.
REFERENCES IN THE PRESS ARE NOT SENSITIVE TO THIS PROCESS.
AND THAT'S WHERE THE NEED FOR ROBUST DEPLOYMENT OF IDN FROM LEFT-MOST TO RIGHT-MOST CHARACTER IN A DOMAIN NAME BECOMES A CRITICAL ISSUE.
JOHN.
>>JOHN KLENSIN: YEAH.
I'M -- I FIRST OF ALL WANT TO EXPRESS COMPLETE AGREEMENT WITH ALMOST EVERYTHING SABINE HAS SAID, BUT WITH ONE QUALIFICATION.
THE THING WHICH MAKES THESE THINGS HARD IS THAT SOME OF US HAVE AN ODD BELIEF THAT HAVING A SINGLE GLOBAL INTERNET IN WHICH EVERYBODY CAN COMMUNICATE WITH EVERYBODY ELSE, AT LEAST IN PRINCIPLE, IS A GOOD IDEA.
THE HARD PROBLEM HERE IS NOT GETTING INTERNATIONAL CHARACTERS ON THE NETWORK, ALTHOUGH THERE ARE HARD PROBLEMS IN THAT.
THE HARD PROBLEM IS HAVING THIS STUFF BE SUFFICIENTLY IDENTIFIED AND SUFFICIENTLY INTERCHANGEABLE AND SUFFICIENTLY INTEROPERABLE THAT WE PRESERVE A GLOBAL NETWORK RATHER THAN ENDING UP WITH A SERIES OF SEPARATED NAME SPACES IN WHICH PEOPLE CAN TALK ONLY WITH THEIR OWN COMMUNITIES AND NOT WITH OTHERS.
NOW, THE SOLUTIONS TO THAT PROBLEM ARE NOT AS SIMPLISTIC AS SOME PEOPLE BELIEVE.
THEY'RE ALSO NOT AS DIFFICULT AS SOME PEOPLE BELIEVE.
BUT, FOR EXAMPLE, IN THE E-MAIL CASE, THAT'S OUR HARD PROBLEM.
HAVING TWO PEOPLE OF THE SAME LANGUAGE AND CULTURAL GROUP COMMUNICATING WITH EACH OTHER USING ADDRESSES IN THEIR OWN LANGUAGE AND SCRIPT IS REALLY PRETTY EASY.
IT'S ONLY WHEN THEY ALSO WANT TO COMMUNICATE WITH OTHER PEOPLE USING THE SAME MAIL SYSTEM THAT WE RUN INTO DIFFICULTIES WHICH WE HAVE TO DO SOME VERY, VERY CAREFUL DESIGN AND PLANNING.
AND PART OF THAT, ALTHOUGH A VERY SMALL PART, TAKES US BACK TO THE ASCII QUESTION WHICH MICHAEL ASKED.
THE REASON FOR ASCII -- THE REASON WHY ASCII AND ITS CLONES IN INTERNATIONAL ALPHABET 5 AND ISO 636 ARE IMPORTANT IS BECAUSE THEY ARE MIND-NUMBINGLY SIMPLE FOR BOTH COMPUTERS AND HUMAN BEINGS.
ASCII HAS THE PROPERTY THAT FOR ANY GIVEN LETTER THAT YOU CAN WRITE AT ALL, THERE IS EXACTLY ONE WAY TO WRITE IT.
WHEN WE DECIDE THAT UPPER AND LOWER-CASE CHARACTERS ARE EQUIVALENT, WE USE AN ALGORITHMIC RULE TO ESTABLISH THAT EQUIVALENCY WHICH IS VERY SIMPLE, MIND-NUMBINGLY SIMPLE.
WE DON'T HAVE TO WORRY ABOUT WHETHER MAKING THE UPPER-CASE VERSION OF A CHARACTER FROM THE LOWER-CASE VERSION WILL GIVE YOU DIFFERENT RESULTS THAN IF YOU STARTED WITH THE UPPER-CASE VERSION, CONVERTED IT TO LOWER-CASE VERSION, AND CONVERT IT BACK AGAIN.
AS SOON AS WE START DEALING WITH MODERN CHARACTER SETS IN INTERNATIONAL SITUATIONS, WE IMMEDIATELY ENCOUNTER TWO PROBLEMS.
THE FIRST PROBLEM IS ONE OF LABELING.
WE HAVE TO UNDERSTAND WHAT IT IS WE'VE DONE.
IF WE'RE USING ISO 8859, WE NEED TO KNOW IN WHICH CHARACTERS IN WHICH VERSION.
IF WE'RE USING UNICODE, WE NEED TO KNOW WHETHER OR NOT WE HAVE NORMALIZED AND WHAT NORMALIZATION HAS BEEN PERFORMED AND WHAT VERSION OF UNICODE WE HAVE NORMALIZED WHICH VERSION OF UNICODE'S CHARACTERS TO.
AND IF YOU DON'T UNDERSTAND THAT SENTENCE, COME BACK ON TUESDAY AFTERNOON AND I WILL TRY TO EXPLAIN IT.
THESE DIFFICULTIES ARE WORTH DEALING WITH AND WORTH OVER COMING.
BUT IT'S NOT EASY.
AND IT GETS EASY AGAIN IF YOU IDENTIFY A SINGLE COMMUNITY WHICH AGREES TO USE EXACTLY THE SAME CONVENTIONS BUT THEIR LANGUAGES AND CHARACTERS AND DOES THAT CONSISTENTLY WITHIN THE COMMUNITY.
BUT IF WE ALSO WANT TO PRESERVE GLOBAL INTEROPERABILITY, THIS IS A TOUCHY, DIFFICULT SET OF PROBLEMS, AND GETTING SIMPLISTIC ABOUT THEM JUST GETS US INTO A LOT OF TROUBLE WE HAVE TO STRAIGHTEN OUT IN THE LONG TERM.
HAD WE NOT GOTTEN SIMPLISTIC THREE YEARS AGO, WE MIGHT NOT BE HAVING THESE DISCUSSIONS NOW.
>>CARY KARP: WE ARE ACTUALLY OUT OF TIME FOR THAT.
IF MICHAEL HAS -- I WAS JUST WONDERING IF SOMEONE WHO HAS NOT SPOKEN JUST WISHES TO SPEAK.
>> MICHAEL EVERSON: I WOULD JUST SAY THE THING THAT WE OUGHT NOT TO BE DOING IS TO FIX THE XN DASH DASH HACK, BUT TO FIX OTHER THINGS SO WE CAN GO FORWARD TO REAL CHARACTER SET.
THE UNICODE CHARACTER SET, THE UNIVERSAL CHARACTER SET, CONTAINS ASCII CHARACTERS.
THEY ARE A SUBSET OF IT.
>>CARY KARP: OKAY.
ANYBODY ELSE?
BRIEF?
NO SPEECHES, HUH?
TEN SECONDS, OKAY.
>>KHALED FATTAL: (SPEAKING ARABIC) I HOPE YOU UNDERSTAND THAT, YOU ANSWER THE QUESTION.
THE POINT I WAS TRYING TO MAKE IS, THESE ARE EXCELLENT DEBATES THAT I THINK ARE VERY VALUABLE, BUT FOR THE PURPOSE OF BEING IN MOROCCO, IN AN ARAB COUNTRY, DEBATING ALL THESE ISSUES, I THINK PERHAPS THE -- WE COULD HAVE BEEN FAR MORE HELPFUL TO THE LOCAL COMMUNITY IF WE COULD EXPLAIN IT TO THEM IN AS SIMPLE AS POSSIBLE.
BECAUSE I'M SURE SOME OF THE POINTS, AS VALID AS THEY ARE, COULD HAVE BEEN LOST AS TO WHY WHAT THEY WANT IS NOT AVAILABLE SO THAT THEY CAN BE PART OF THIS GLOBAL INTERNET.
PERHAPS IT'S JUST --
>>CARY KARP: I'M AVAILABLE.
WE CAN GLADLY PROTRACT THIS DISCUSSION IN THE CORRIDOR AFTERWARDS OR ANYTIME DURING THE REST OF THE WEEK.
SO IF YOU WOULD LIKE TO GATHER PEOPLE FROM THE LOCAL COMMUNITY TO SPEAK TO THIS, I'D BE DELIGHTED TO PARTICIPATE IN THAT DISCUSSION.
EDMON, LAST SPEAKER, LAST TEN-SECONDER.
>> EDMON CHUNG: I JUST WANT TO POINT OUT, ONE OF THE THINGS THAT YOU MENTIONED IN THE PRESENTATION ABOUT THE PUNYCODE FORM WHICH MAY HAVE I.P. CONSIDERATIONS.
I THINK THAT'S A VERY GOOD POINT AND OFTEN MISSED POINT.
FOR EXAMPLE, XN DASH DASH, IN FACT, I'VE DONE SOME RESEARCH, A LITTLE BIT OF RESEARCH, AND XN--COCA-COLA, XN--MCDONALD'S, AND XN--IETF ACTUALLY CONVERTS TO VALID DOMAINS.
SO IT IS AN ISSUE THAT I THINK THE I.P. CONSTITUENCY -- I MEAN, I.P. COMMUNITIES SHOULD REALLY TAKE --
>>CARY KARP: NO.
THE INTERNET COMMUNITY.
>> EDMON CHUNG: THE INTERNET COMMUNITY.
>>CARY KARP: OKAY.
AND NOW A MAN WHO NEEDS NO INTRODUCTION.
>>KURT PRITZ: THANK YOU VERY MUCH, CARY.
THAT WAS EXCELLENT.
GOOD AFTERNOON, EVERYONE.
I'D LIKE TO INTRODUCE THE NEXT ROUND THE SPEAKERS.
THEY ARE MR. BRUCE TONKIN, THE GNSO CHAIR, AND FROM MELBOURNE I.T; HIRO HOTTA, THE CCSO, NSO IDN REPRESENTATIVE AND FROM JPRS.
AND OLOF NORDLING, ICANN STAFF POLICY OFFICER.
SO IF YOU WOULD COME UP TO THE STAGE, PLEASE.
THEY'RE GOING TO DISCUSS THE ICANN STAFF PRELIMINARY POLICY ISSUES REPORT ON IDNS TO SCRIBE THE WORK OF THE JOINT CCNSO, GNSO, IDN WORKING GROUP, AND THE NEXT STEPS THAT GROUP'S GOING TO TAKE.
>>BRUCE TONKIN: OKAY.
I GUESS THE SORT OF THINGS THAT WE'RE LOOKING AT FROM A POLICY PERSPECTIVE, AT A VERY HIGH LEVEL, WE'RE RESPONDING TO SOME OF THE WISHES IN THE GNSO AND ALSO IN THE CCNSO COMMUNITY, WHICH IS ESSENTIALLY AN EXISTING GTLD OPERATOR MAY WISH TO INTRODUCE AN IDN BY STRING THAT RELATES TO THAT EXISTING GTLD, AND PRESUMABLY THEY'D WANT TO DO THAT BECAUSE THEY BELIEVE THAT THEIR CUSTOMERS WOULD PREFER TO SEE THEIR TLD OPERATED IN A SCRIPT OR LANGUAGE THAT'S APPROPRIATE FOR A PARTICULAR MARKET.
AN EXISTING CCTLD OPERATOR MIGHT ALSO WISH TO INTRODUCE AN IDN-BASED STRING THAT RELATES TO THAT EXISTING CCTLD, PARTICULARLY, PERHAPS, RELATING TO THE SCRIPTS THAT ARE USED WITHIN THAT PARTICULAR COUNTRY.
THIRDLY, A PARTY MIGHT WISH TO INTRODUCE A NEW IDN STRING THAT HAS NO RELATIONSHIP TO AN EXISTING TLD, AND, REALLY, FROM A POLICY POINT OF VIEW, THAT'S ESSENTIALLY THE SAME AS THE WORK THAT WE'RE DOING ON NEW GTLDS.
BUT, IN ANY CASE, ALL OF THESE ARE TRULY NEW TOP-LEVEL DOMAINS, AND THEY DO REQUIRE ADDING NEW MATERIAL INTO THE ROOT ZONE.
THE APPROACH THAT WE USE AS A POLICY DEVELOPMENT PROCESS -- AND THIS IS JUST A REMINDER, BECAUSE IT PUTS SOME CONTEXT WHERE WE ARE IN OUR PROCESS SO FAR -- BUT, FIRST, WE REQUEST AN ISSUES REPORT.
AND OLOF NORDLING HAS MANAGED THE PRODUCTION OF AN ISSUES REPORT.
IT'S STILL IN DRAFT FORM, THOUGH.
AND WE'RE REALLY TRYING TO SEPARATE A LOT OF THE DISCUSSION TO DATE AT ICANN HAS BEEN ABOUT THE MANY TECHNICAL ISSUES ASSOCIATED WITH IDNS.
BUT WE ALSO NEED TO START THINKING ABOUT WHAT SOME OF THE POLICY ISSUES ARE.
AND WE'VE HAD A FIRST CUT AT CAPTURING THOSE POLICY ISSUES, AND WE HOPE TO CONCLUDE THAT ISSUES REPORT AT THE END OF THIS MEETING AFTER WE TAKE MORE INPUT, AND SO EARLY JULY, WE SHOULD HAVE A FINAL ISSUES REPORT ON IDNS.
THAT ISSUES REPORT WILL ALSO HAVE SOME SUGGESTIONS AS TO HOW TO STUDY THOSE ISSUES WITHIN THE ICANN POLICY DEVELOPMENT FRAMEWORK.
WITHIN THE GNSO, IF WE THEN DECIDE TO GO AHEAD AND DEVELOP SOME POLICIES IN THIS AREA, WE THEN BASICALLY FORMALLY ANNOUNCE THAT, AND WE GO OUT TO A PUBLIC COMMENT PERIOD.
WE THEN CREATE A SET OF INITIAL RECOMMENDATIONS, WHICH CAN BE IN THE FORM OF A TASK FORCE REPORT.
AND, AGAIN, WE GO OUT FOR COMMENT ON THOSE INITIAL RECOMMENDATIONS.
THEN WE HAVE A FINAL REPORT FOR THE COUNCIL.
AND THEN A -- THE COUNCIL PREPARES A REPORT FOR THE BOARD, AND THEN THE BOARD VOTES ON THAT WITH RESPECT TO NEW POLICY.
SO WHERE WE ARE SO FAR IN THAT POLICY PROCESS, WHERE WE HAVE THE DRAFT ISSUES REPORT, THE GNSO FORMED A WORKING GROUP TO CONSIDER THAT DRAFT AND IDENTIFY WHAT ISSUES COULD BE CONSIDERED WITHIN THE GNSO POLICY DEVELOPMENT PROCESS.
AND THE GNSO IS ALSO LIAISING WITH THE CCNSO AND WITH THE GAC.
AND WE BRIEFLY MET WITH THE GAC JUST BEFORE THIS MEETING.
AND WE, AS I SAID, HOPE TO HAVE AN ISSUES REPORT BY NEXT MONTH.
THE WORK FROM A GNSO POINT OF VIEW IS THAT WE SUGGEST WE WOULD STUDY IDN GTLDS USING THE SAME FRAMEWORK AS WE'RE CURRENTLY STUDYING FOR NEW GTLDS.
SO, REALLY, IDNS, FOR US, ARE JUST A SPECIAL CASE OF A NEW GTLD.
AND MANY OF THE ISSUES ARE COMMON BETWEEN WHETHER YOU'RE JUST ADDING AN ASCII TLD OR WHETHER YOU'RE GOING TO ADD AN IDN-BASED TLD, THEY'RE SIMILAR ISSUES.
BUT IN SOME CASES, THERE ARE SOME ISSUES THAT ARE, I GUESS, HIGHLIGHTED BY THE EXISTENCE OF IDNS OR THE POTENTIAL EXISTENCE OF IDNS IN THE ROOT, WHICH ADD A LOT MORE COMPLEXITY.
AND WE NEED TO STUDY SOME OF THOSE COMPLEXITIES, I GUESS, AS EARLY AS POSSIBLE.
SO THIS, THEN, IS JUST TO GIVE A FEEL FOR SOME OF THE FRAMEWORKS THAT WE'RE USING.
IF WE LOOK AT, FIRSTLY, SELECTION THAT CAN BE BROKEN DOWN, I GUESS, INTO TWO ISSUES.
ONE IS SELECTING THE APPROPRIATE IDN STRING TO ADD TO THE ROOT.
AND THEN THE SELECTION OF AN APPROPRIATE ORGANIZATION TO OPERATE A TLD THAT USES THAT STRING.
SO THAT'S REALLY AT A HIGH LEVEL THE TWO ISSUES.
AND THEY APPLY EQUALLY TO COUNTRY CODE-BASED TLDS AS THEY DO TO GENERIC TLDS.
SO, THEN, WHEN YOU BREAK THAT DOWN FURTHER, YOU WOULD BE SAYING, WELL, WHAT ARE THE REQUIREMENTS THAT AN ORGANIZATION WOULD NEED TO MEET TO BE ALLOCATED AN IDN TLD?
SOME OF THOSE REQUIREMENTS WOULD BE THE SAME AS ALLOCATING ANY TLD.
SO WE WOULD EXPECT THERE'S MINIMUM TECHNICAL CRITERIA, THAT THE ORGANIZATION HAS THE RESOURCES TO OPERATE A TLD TO THE CORRECT TECHNICAL STANDARDS AND THEY HAVE SOME EXPERIENCE AND TECHNICAL STAFF, ET CETERA.
THE OTHER THING THAT'S COMING OUT OF THE NEW GTLD PROCESS IS THAT AS WE INTRODUCE NEW STRINGS, THEY SHOULD BE DIFFERENTIATED WITH RESPECT TO PURPOSE.
IN OTHER WORDS, IF WE'RE GOING TO ADD SOMETHING NEW, IT SHOULD HAVE A CLEARLY DIFFERENTIATED PURPOSE FROM WHAT'S ALREADY IN THE ROOT ZONE.
BUT, OF COURSE, WHEN WE START ADDING THE COMPLEXITY OF IDNS IT MIGHT BE HARD TO TELL WHETHER A STRING IS, IN FACT, DIFFERENT IN PURPOSE, BECAUSE WE COULD BE DEALING WITH DIFFERENT SCRIPTS.
AND ALSO DIFFERENT LANGUAGES SO A PARTICULAR STRING MIGHT HAVE A DIFFERENT MEANING IN DIFFERENT LANGUAGES.
AND THEN IF WE'RE SAYING, WELL, WE'RE WANTING TO INTRODUCE SOMETHING THAT'S DIFFERENT FROM SOMETHING ELSE, IT'S OKAY IF WE'VE ONLY GOT THREE EXISTING GTLDS, BUT AS WE ADD MORE GTLDS AND WE START GETTING UP INTO, YOU KNOW, MAYBE A HUNDRED TLDS IN TOTAL, OR NEW TLDS IN TOTAL, THE INTERACTIONS BETWEEN THE MEANINGS AND THE PURPOSE BECOME A LOT MORE COMPLEX TO MANAGE.
ALSO, HOW MANY STRINGS ARE APPROPRIATE FOR A PARTICULAR TLD PURPOSE?
SO AN EXISTING OPERATOR THAT SAYS, I WANT TEN STRINGS TO MEET THE PURPOSE OF THE COMMUNITY OF USERS THAT RELATE TO THAT PURPOSE, IS THAT APPROPRIATE?
WHAT IF THEY SAID THEY WANTED A HUNDRED OR 200 OR A THOUSAND?
YOU KNOW, WHAT SORT OF LIMITS WOULD BE PLACED ON AN EXISTING ORGANIZATION ADDING IDN-BASED STRING EQUIVALENTS OF THEIR CURRENT TLD?
IF WE LOOK AT CCTLDS, TODAY, CCTLDS ARE RESTRICTED TO THE STRINGS THAT ARE IN THE ISO3166 TABLE.
SO HOW WOULD YOU DECIDE WHAT IDN STRINGS-ALLOWED FOR CCTLDS?
AND MY COLLEAGUE, HIRO HOTTA, WILL TALK ABOUT THAT SHORTLY WITH SOME INITIAL THOUGHTS ON THAT TOPIC.
GTLD OPERATORS MAY BE PREVENTED FROM USING AN IDN STRING THAT COULD BE CONFUSED WITH A COUNTRY NAME IN A PARTICULAR SCRIPT AND LANGUAGE.
SO THIS IS HOW WE TRY AND CONTINUE TO MAINTAIN SOME FORM OF DIFFERENTIATION BETWEEN STRINGS THAT ARE INTENDED FOR USE BY COUNTRIES AND STRINGS WHICH ARE INTENDED FOR OTHER USES.
WE LOOK AT ALLOCATION PROBLEMS.
OBVIOUSLY, TWO PARTIES MAY WISH TO HAVE THE SAME STRING.
THEY MIGHT EACH DO THEIR MARKET RESEARCH AND FIND THAT THIS PARTICULAR STRING SEEMS TO BE A STRING THAT WOULD BE WIDELY RECOGNIZED BY THEIR USER BASE.
SO THEY BOTH WANT IT.
AND SO YOU NEED TO HAVE A LOOK AT, WELL, WHAT ARE THE ALLOCATION METHODS AND SOME OF THIS WORK IS HAPPENING WITHIN THE NEW GTLD PROCESS.
BUT WHEN WE ADD IDNS TO THIS, WE SUDDENLY HAVE A LOT MORE COMPLEXITY.
SO THE SAME STRING MAY HAVE DIFFERENT MEANINGS FOR PEOPLE IN DIFFERENT PARTS OF THE WORLD.
TWO STRINGS THAT ARE TECHNICALLY DIFFERENT, AND BY THAT I MEAN THEY HAVE A DIFFERENT ASCII EQUIVALENT IN THE ROOT, MAY VISUALLY LOOK SIMILAR WHEN WE'RE TRANSLATING IT INTO THE IDN REPRESENTATION.
AND AS I MENTIONED BEFORE, IT STARTS TO GET HARD TO TELL, ARE THESE STRINGS REALLY DIFFERENTIATING OR NOT?
SO HOW MANY STRINGS ARE APPROPRIATE FOR A PARTICULAR PURPOSE, AGAIN, COMES UP.
AND SO FORTH.
CONTRACTUAL CONDITIONS.
IN TERMS OF GTLDS, WE HAVE AN ENVIRONMENT WHERE A GTLD OPERATOR ENTERS A CONTRACT WITH THE -- WITH ICANN, AND THE QUESTION IS, WELL, WHAT ADDITIONAL REQUIREMENTS WOULD THERE BE ON THAT OPERATOR IF THEY WERE OPERATING AN IDN STRING?
AND PROBABLY THE MAIN ONE IS THAT THEY WOULD COMPLY WITH IETF STANDARDS THAT RELATE TO IDNS, AS WELL AS ANY ICANN IDN GUIDELINES.
NOW, THESE ARE SOME OTHER ISSUES THAT DON'T NEATLY FIT INTO ANY OF OUR EXISTING CATEGORIES.
AND SOME OF THESE HAVE BEEN DISCUSSED BEFORE.
BUT IF YOU HAVE TWO TLD STRINGS FOR THE SAME PURPOSE, SHOULD THE SECOND-LEVEL STRINGS RESOLVE TO THE SAME INTERNET LOCATION?
SO, IN OTHER WORDS, IF I HAD THE DOMAIN NAME BRUCE.NAME, AND I HAD DOT NAME IN A CHINESE CHARACTER EQUIVALENT, SHOULD BRUCE DOT CHINESE CHARACTER EQUIVALENT ACTUALLY GO TO THE SAME PLACE OR NOT?
OR SHOULD THEY BE, EFFECTIVELY, COMPLETELY INDEPENDENT?
ALSO, SHOULD THE SCRIPT USED IN A STRING AT THE SECOND LEVEL MATCH THE SCRIPT USED IN THE STRING AT THE TOP LEVEL?
SO THAT MAY BE I HAVE BRUCE IN LATIN CHARACTERS, DOT NAME IN LATIN CHARACTERS.
IT MAY BE THAT IF I'M GOING TO HAVE DOT NAME IN AN EQUIVALENT CHINESE CHARACTERS, I SHOULD ONLY BE ALLOWED TO HAVE BRUCE IN WHATEVER THE CHINESE CHARACTER'S EQUIVALENT OF THE NAME "BRUCE" IS.
SO THESE ARE SORT OF DIFFERENT OPTIONS.
AND THESE THINGS MAY BE CONSIDERED TO BE SOMETHING THAT'S A CHOICE OF THE TLD OPERATOR TO DECIDE, OR IT MAY BE SOMETHING THAT WE MAY WISH TO MAKE SOME SORT OF POLICY, PARTICULARLY IF IT HELPS USER ACCEPTANCE AND DECREASES SOME OF THE POTENTIAL CONFUSION IN -- AT THE USER LEVEL.
ANOTHER THING, SHOULD THERE BE COMPETITION BETWEEN TWO OPERATORS OF STRINGS FOR THE SAME PURPOSE?
SAY A PARTICULAR COUNTRY MAY DECIDE AT SOME POINT THAT IT WANTS TO HAVE A SEPARATE ORGANIZATION RUN THE STRING IN A DIFFERENT SCRIPT.
DOES THAT ADD ANY BENEFIT?
AGAIN, IS THAT SOMETHING THAT SHOULD JUST BE DEALT WITH AT A LOCAL LEVEL RATHER THAN AT A GLOBAL LEVEL?
WE HAVE SOME EXISTING POLICIES THAT RELATE TO GTLDS, ONE OF WHICH IS A UNIVERSAL DISPUTE RESOLUTION POLICY.
THAT IS LIKELY TO BECOME A LOT MORE COMPLEX AS WE START ADDING ADDITIONAL SCRIPTS BOTH TO THE LEFT AND TO THE RIGHT OF THE DOT.
AND NATURALLY, AS CARY AND OTHERS HAVE MENTIONED, THE SCOPE FOR CONFUSING HUMANS INCREASES, WHICH LEADS TO ADDITIONAL OPPORTUNITIES FOR PHISHING AND SPAM. SO I DON'T THINK THESE ARE THINGS WE WILL BE ABLE TO SOLVE IN THE SHORT TERM, BUT WE NEED TO BE AWARE OF THEM AND PERHAPS TAKE SMALL STEPS FORWARD TO TRY TO MINIMIZE THE SCOPE FOR PROBLEMS IN THE EARLY DAYS SO WE DON'T ADD TOO MANY STRINGS AT ONCE FOR A PARTICULAR PURPOSE AND WE LEARN WITH JUST MAKING SOME SMALL STEPS FORWARD.
WITH THAT, I WILL TURN IT OVER TO MY COLLEAGUE HOTTA TO GIVE A PERSPECTIVE ON COUNTRY CODES.
>>HIRO HOTTA: THANK YOU, BRUCE. IS IT ON? YES.
LET ME GIVE SOME COMMENTS, ESPECIALLY FROM OUR CCTLD VIEWPOINT.
PLEASE TAKE WHAT I WILL SAY AS A VIEW FROM JUST ONE CCTLD MANAGER, NOT AS A CONSENSUS VIEW FROM ALL THE CCTLDS.
FROM THE VIEWPOINT OF CCTLDS, I THINK THE ISSUES THAT SHOULD COLLECT FOCUS ARE THE FOLLOWING THREE.
ONE, SELECTION OF IDN CCTLD STRINGS. WHO SELECTS WHAT THE APPROVAL CRITERIA ARE, HOW MANY. HOW COLLISIONS ARE RESOLVED IF MULTIPLE CCS SHOULD WANT THE SAME STRING.
AND TWO, SELECTION OF IDN CCTLD MANAGER. AUTOMATIC ASSIGNMENT TO THE EXISTING ASCII CCTLD MANAGER, OR NEW SELECTION.
THREE, HARMONIOUS USER EXPERIENCE AT THE GLOBAL LEVEL. NO NEED TO SUCH HARMONY OR DO WE NEED SOME -- TO SOME EXTENT SUCH A HARMONY.
SELECTION OF MANAGER IS MAINLY SOLVED AT THE LOCAL LEVEL. BUT A SELECTION OF IDN STRING SHOULD BE SOLVED THROUGH COOPERATION AMONG LOCAL AND GLOBAL EFFORTS. AND ACHIEVING HARMONIOUS USER EXPERIENCE MUST BE SOLVED THROUGH BEST PRACTICE OR GUIDELINES ELABORATED GLOBALLY ONCE SUCH EXPERIENCE IS CONSIDERED NECESSARY.
TO MAKE ONE STEP FURTHER IN SELECTING IDN CCTLD STRING, ONE OF THE IDEAS IS TO COME UP WITH A PARALLEL TABLE WHICH SHOWS THE CORRESPONDENCE BETWEEN ISO 3166 TWO-CHARACTER ASCII COUNTRY CODE AND A CORRESPONDING IDN TLD STRING BEING RELATED TO THE NAME OF THE COUNTRY OR REGION.
OF COURSE THERE ARE HURDLES EVEN IN MAKING THIS SIMPLE TABLE, SUCH AS THE SITUATION WHERE TWO OR MORE OFFICIAL LANGUAGES EXIST IN ONE CC, OR THE HARD DECISION ON WHO DECIDES THE PROPOSED STRING.
HOWEVER, IT MAY BE WORTH TRYING, AS IT SEEMS NECESSARY ANYWAY ON THE WAY TO FULL DEPLOYMENT OF IDN TLDS.
AS TO THE COOPERATION BETWEEN GNSO AND CCNSO PROPOSED IN THE GNSO REPORT, COOPERATION IS ABSOLUTELY NECESSARY. AS TO WHETHER CCNSO PDP IS NEEDED OR NOT WILL BE DISCUSSED LATER, SO I WILL STOP HERE. THANK YOU.
>>OLOF NORDLING: WELL, IF I JUST MAY SAY ONE LITTLE WORD ABOUT THE PRELIMINARY ISSUES REPORT. OF COURSE OUR LEADING LIGHT IN ALL THIS WORK, AND I THINK THAT HAS BEEN ESTABLISHED ALREADY IN VANCOUVER, THAT WOULD BE THE GOOD USER EXPERIENCE. AND IT APPEARS TO BE PERFECTLY CLEAR THAT IT'S NOT JUST A MATTER OF THE TECHNICAL MATTERS BUT MOST CERTAINLY THERE ARE VERY DIFFICULT POLICY ISSUES AND WE TRIED TO ADDRESS THEM IN PARALLEL WITH THE TECHNICAL TESTS STARTING UP.
SO THAT WAS THE STARTING POINT FOR THE PRELIMINARY ISSUES REPORT.
WE TRIED TO, FROM THE STAFF SIDE, COVER AS MUCH TERRITORY AS WE COULD WITHOUT GOING TOO FAR OUT AND WITHOUT DEALING WITH THE TECHNICAL ISSUES IN THIS ISSUES REPORT. SO -- WELL, I GUESS THAT IS WHAT WE DO IN FIRST STEP, IS REALLY TO ASK FOR YOUR INPUT ON WHETHER THE ISSUES REPORT, PRELIMINARY ISSUES REPORT, IS COMPLETE FROM THAT PERSPECTIVE.
AND WE'VE POSTED IT FOR PUBLIC COMMENTS, AND I MUST SAY, IF THAT'S ANYTHING TO JUDGE FROM, WELL, IT'S REALLY COMPLETE. WE GOT FOUR RESPONSES AND THEY WERE ALL SPAM. SO IT MAY NOT BE THE FULL TRUTH, BUT THAT'S WHERE WE STAND RIGHT NOW.
SO WE'RE VERY KEEN TO RECEIVE YOUR INPUT ON THE COMPLETENESS OF THE REPORT.
THANK YOU.
>>BRUCE TONKIN: THANKS, OLOF. SO AT THAT POINT, I GUESS WE'RE OPEN FOR ANY PUBLIC COMMENTS ON ANY ASPECTS OF WHAT WE'RE TRYING TO DO AT A POLICY LEVEL. ANY SUGGESTIONS ANYONE MAY HAVE IN TERMS OF WHAT ISSUES WE SHOULD BE FOCUSING ON WOULD BE USEFUL. AND I GUESS ANY COMMENTS ABOUT THE IDEA OF CREATING A NEW TABLE -- NOT NECESSARILY A NEW TABLE BUT PERHAPS AN EXTRA COLUMN TO THE EXISTING ISO 3166 TABLE THAT INCLUDES STRINGS RELATING TO COUNTRY CODES.
>>KURT PRITZ: HELLO. I JUST WANT TO MAKE ONE COMMENT FIRST. AS OPPOSED TO TRADITIONAL ICANN MEETINGS, OUR EXCELLENT TRANSCRIBERS ARE SITTING TOWARD THE BACK OF THE ROOM, AND CANNOT TAKE DOWN WRITTEN NAMES. SO IF -- SO IF EVERYONE FOR THE REMAINDER OF THE WORKSHOP COULD STATE THEIR NAME CLEARLY FOR THE TRANSCRIPTION, IT WOULD BE GREAT FOR THE FUTURE RECORD OF THE MEETING.
>>JOHN KLENSIN: JOHN KLENSIN. JUST ONE OBSERVATION, BRUCE. THE ISO SUBCOMMITTEE RESPONSIBLE FOR 3166 HAS BEEN ASKED THE QUESTION ABOUT NON-ISO 646 BASIC VERSION CODES IN THAT TABLE SEVERAL TIMES BEFORE, FOR VARIOUS AND SUNDRY REASONS.
THEY HAVE, TO PUT IT POLITELY, REJECTED THE IDEA. THE INTERNATIONAL POSTAL UNION DOESN'T LIKE THE IDEA. THE INTERNATIONAL CURRENCY UNION DOESN'T LIKE THE IDEA. THE RADIO BUREAU OF THE ITU DOESN'T LIKE THE IDEA.
EACH OF THESE ORGANIZATIONS AND SEVERAL OF THE NATIONAL MEMBER BODIES SEE AN ADDITIONAL SET OF EQUIVALENTS TO THE TWO OR THREE CHARACTER ALPHA CODE LISTS TO BE A SERIOUS THREAT TO INTEROPERABILITY IN THEIR WORLDS.
SO THE RESPONSE SO FAR HAS BEEN SOMEWHERE IN THE VICINITY OF NOT ON YOUR LIFE. I WON'T USE STRONGER LANGUAGE.
ONE COULD, OF COURSE, GO BACK AND ASK THEM AGAIN IF ONE WANTED TO BE AMUSED BY GETTING DIFFERENT ANSWERS, BUT I WOULDN'T COUNT ON IT.
>>BRUCE TONKIN: I MIGHT JUST RESPOND TO THAT, JOHN. I WOULDN'T NECESSARILY SEE IT BECOMING AN ISO STANDARD, BUT I THINK IT'S BASICALLY USING THE, IF YOU LIKE, THE LEFTMOST COLUMN WHICH IS THE LIST OF COUNTRIES, AND THEN USING THAT AS A REFERENCE POINT.
THE TABLE IS PROBABLY GOING TO EFFECTIVELY BE AN ICANN TABLE.
>>JOHN KLENSIN: YEAH, THE DIFFICULTY IS THAT STUFF CHANGES. AND IF WE DON'T HAVE EITHER THEM DOING IT OR A CLOSE COORDINATION POINT WITH THEM DOING IT, WE END UP WITH CHAOS VERY, VERY QUICKLY.
>>MICHEL SUIGNARD: MICHEL SUIGNARD. IN FACT, I WOULD LIKE TO MENTION IN FACT THE LIST IS NOT THE TECHNICAL ISSUE. IN FACT, I HAVE DONE SUCH A LIST FOR THE USAGE OF THE IDN PACK. SO THE LIST CAN BE DONE. THE ISSUE REALLY IS MORE DO WE WANT TO OWN THE LIST, BECAUSE THE LIST I CAN PROPOSE, I CAN SHOW. ALL THE CCS, YOU CAN GO TO THE NAMES OR DO RESEARCH ON THE WEB SITES, TO THE LIBRARIES TO FIND OUT WHAT IS THE NATIVE NAME, THE SHORTEST NAME, IF YOU WANT, AND YOU CAN FIND IT.
SO IF ICANN WANTS TO CREATE SUCH A LIST, BUT THEN YOU HAVE TO OWN IT BECAUSE NOBODY ELSE WANTS TO OWN THE LIST, UNLIKE THE CC THAT CAME FROM THE ISO STANDARD.
SO IT'S NOT REALLY A TECHNICAL ISSUE. I MEAN, I HAVE ON MY OWN MACHINE HERE, I HAVE THE LIST OF THE COUNTRY NAMES. I'LL BE GLAD TO SHARE THOSE NAMES WITH ANYBODY THAT WANTS.
>>VINT CERF: IT'S VINT CERF AGAIN.
I THINK THE FUNDAMENTAL QUESTION REALLY IS WHO WANTS TO OWN THAT LIST. THE PROBLEM IS WHAT HAPPENS WHEN IT HAS TO CHANGE. HOW DO YOU DECIDE WHETHER SOMETHING SHOULD BE ADDED, HOW DO YOU DECIDE WHETHER SOMETHING SHOULD BE REMOVED?
AND UP UNTIL NOW, THE INTERNET COMMUNITY HAS TRIED HARD NOT TO OWN THE QUESTION WHO IS A COUNTRY OR WHAT IS A COUNTRY, WHAT IS THE RECOGNIZED GOVERNMENT OF A COUNTRY, AND MANY OF THE OTHER ISSUES THAT ARISE IN CONNECTION WITH THESE LISTS.
PERSONALLY, AND I GUESS SPEAKING AS SOMEONE WITH SOME EXPERIENCE WITH THIS, I WOULD PREFER NOT TO OWN THAT PROBLEM. IT'S HARD ENOUGH AS IT IS TO CONDUCT THE BUSINESS OF ICANN WITHOUT ALSO HAVING TO FIGURE OUT WHETHER DIPLOMATIC RELATIONS SHOULD BE ESTABLISHED WITH A NEW ENTITY.
SO I AM VERY SYMPATHETIC TO WHAT BRUCE IS SAYING IN THE SENSE OF HAVING A PARALLEL THAT WOULD ALLOW US TO RELATE IDN SYMBOLS TO AN EXISTING LIST OF COUNTRY CODES. BUT IF WE'RE GOING TO DO IT, I THINK JOHN KLENSIN IS RIGHT, THAT IT HAS TO BE CLOSELY COORDINATED WITH THE 3166 LIST TO THE EXTENT THAT WE CHOOSE TO CONTINUE TO RELY ON THAT FOR GUIDANCE AS TO WHAT CODES WE SHOULD BE ATTENDING TO.
>>BRUCE TONKIN: I THINK, VINT, TO TAKE THE SPECIFIC, WE'RE REALLY TALKING ABOUT CONCEPTS THAT HAVE ONLY BEEN DISCUSSED IN THE LAST FEW DAYS HERE SO I AM TALKING ON THE FLY HERE, BUT I THINK THE -- WHEN WE DISCUSS THE IDEA OF WHAT HAPPENS WHEN A NEW COUNTRY WANTS TO BE ADDED, THEY GO TO THE ISO AND GET ADDED THERE. AND AT THAT POINT, ONLY WHEN THEY ARE ON THE ISO LIST DO WE THEN CONSIDER CREATING A CHARACTER EQUIVALENT.
SO WE'RE ANCHORING OUR LIST, IF YOU LIKE, AGAINST THEIR LIST OF COUNTRIES. SO IT'S A PARALLEL THING, NOT ANOTHER COMPLETELY INDEPENDENT TABLE. SO I FULLY APPRECIATE THE COMMENT BOTH YOU AND JOHN MADE. WE STILL NEED TO COORDINATE THAT, BUT I DON'T SEE ICANN CREATING THE LEFTMOST COLUMN BEING THE COUNTRY NAME OR THE COUNTRY. THAT WOULD BE DEALT WITH EXTERNALLY. WE WOULD JUST BE LOOKING AT THE SECOND COLUMN, IF YOU LIKE, OR THE THIRD COLUMN.
>>MOUHAMET DIOP: THANK YOU, BRUCE. MOUHAMET DIOP FROM SENEGAL.
I THINK THAT THIS PART OF THE IDN DISCUSSION IS VERY INTERESTING BECAUSE WE'RE DEALING WITH POLICY. WE KNOW THAT THE EASIEST PART IS THE TECHNICAL PART. AND THE HARDEST PART OF THE DISCUSSION IS THIS PART RELATED TO THE POLICY.
AND I WANT TO REMIND THAT ICANN IS AN ORGANIZATION THAT'S NOT DEALING SPECIFICALLY WITH COUNTRIES. I THINK THAT THE CORE OF OUR MISSION IS TO DEAL WITH COMMUNITIES.
WE KNOW THAT WE HAVE NO KNOWLEDGE ABOUT HOW TO DEAL WITH COUNTRIES. THAT'S WHY WE TAKE REFERENCE THAT IS NOT OURS, AND WE SAY THAT IDEALLY, WE'RE NOT GOING TO ATTACH ANYTHING ON IT BECAUSE THAT'S NOT OUR MISSION TO TRY TO THINK ABOUT IT OR JUST TOUCH IT OR MODIFY IT BECAUSE WE HAVE NO COMPETENCE TO DEAL WITH THIS DATABASE REFERENCE.
SO I THINK THAT WAS REALLY WISE NOT TO TOUCH IT AND NOT TO BIND IT IN ANY OF OUR POLICY IN TERMS OF EVOLUTION BECAUSE WE HAVE NO RIGHT TO TOUCH IT AND WE HAVE NO RIGHT TO MAKE IT INVOLVED.
SO MY CONCERN IS WHEN WE'RE TALKING ABOUT IDN, WE FIND AN EASIEST WAY TO THINK ABOUT HOW WE'RE GOING TO MATCH, ALLOWING AN IMPLEMENTATION OF SOME COUNTRY CODE. BUT I THINK THAT IT JUST HAVE TO BE CONFINED INTO AN IDEA, NOT AN IMPLEMENTATION. BECAUSE WHAT I'M TRYING TO SAY IS, COUNTRIES LIKE CHINA HAVE A VERY SIMPLE IMPLEMENTATION OF IDN, AND EVERYONE AGREES IT'S ONE COUNTRY, ONE LANGUAGE AND IT'S EASY TO DEAL WITH IT BUT MANY OTHER COUNTRIES ARE NOT IN THE SAME SITUATION. SO WE'RE TALKING ABOUT POLICY. WE HAVE TO DISCUSS WHERE THAT IS GOING TO ALLOW US TO IMPLEMENT THINGS VERY EASILY IN A WAY THAT CAN BE IMPLEMENTED IN DIFFERENT PARTS OF THE WORLD. SO WE ARE DEALING WITH COMMUNITY HERE. AND THE FACT THAT WE TRIED TO MAKE GNSO AND CCNSO WORK CLOSELY TOGETHER IN ORDER TO GET SOMETHING DONE IS SOMETHING THAT ICANN WILL OWN AS AN OWN APPROACH OF THE ICANN COMMUNITY.
SO IF WE BIND IT TO THE ISO 3166, WE'RE STUCK AND WE CANNOT MOVE BECAUSE IT IS SOMETHING THAT HAS BEEN MANAGED BY OTHER INTERNATIONAL ORGANIZATION THAT ARE NOT ICANN, AND WE CANNOT TELL THEM WHAT TO DO. SO WE HAVE TO COME BACK BECAUSE WE HAVE OUR OWN LIST. THE GTLD LIST IS LIST THAT WE HAVE CREATED, AND WE HAVE DECIDE TO MAKE THIS DOMAIN NAME EXIST IN THE ROOT SERVER.
SO I THINK ON THE IDN APPROACH, WE CANNOT AVOID GETTING THE SAME APPROACH THAT WE HAVE WHEN IT IS TO EXPAND THE DOMAIN NAME SPACE. IT IS OUR RESPONSIBILITY TO ALLOW THE EXPANSION OF THE DOMAIN NAME SPACE.
HOW ARE WE GOING TO DO IT? HOW ARE WE GOING TO FACILITATE IT? IT WILL BE OUR INTERNAL PROCEDURE. WE CAN'T ALLOW GETTING AS A PROPERTY OF ICANN AND THE RESPONSIBILITY OF ICANN TO DECIDE HOW WE'RE GOING TO MAKE IT HAPPEN.
AND I THINK THAT IF YOU ARE YOU ARE TRYING TO MATCH IT OR BIND IT TO ANY OTHER EXISTING ORGANIZATION, WE WON'T GO NOWHERE BECAUSE WE KNOW THAT WE'LL HAVE NO ROOM TO GO ON OUR SPEED OR TO MAKE IT HAPPEN ON OUR SPEED.
SO THE IMPLEMENTATION OF THE CCTLD IS SOMETHING SPECIFIC, AND I THINK THAT THE RESPONSIBILITY FROM JAPAN HAVE ALREADY ASK THE QUESTION THAT WE FACE IS REALLY IF WE ALLOW TO DO AN EXPERIMENTATION UNDER ANY CC, WE HAVE TO ANSWER THE QUESTION REGARDING UNDER WHICH RESPONSIBILITY WE ARE GOING TO MAKE IT.
AND MY OWN OPINION IS WE NEED TO DECORRELATE THE ACCEPTATION OF THE CCTLD WITH THE EXISTING RESPONSIBILITY OF THE COUNTRY CODE DOMAIN NAME MANAGEMENT UNDER THE 3166 ISO CODE. THANK YOU.
>>MICHAEL EVERSON: MICHAEL EVERSON. THE TOKEN LINGUIST.
I HAVE A NUMBER OF HATS. I AM THE EDITOR OF ISO 15924. I WAS THE EDITOR. IT'S A STANDARD NOW. THAT IS THE CODE FOR THE REPRESENTATION OF NAMES OF SCRIPTS. I AM THE REGISTRAR OF THE UNICODE CONSORTIUM AS THE REGISTRATION AUTHORITY.
AND I AM ALSO THE LANGUAGE TAG REVIEWER FOR EXTENSIONS TO ISO 639 FOR THE PURPOSE OF TAGGING TEXT IN HTML AND SO ON AND SO FORTH.
THERE IS AN RFC 3066-BEECE IT'S CALLED, AND IN THAT -- IN THAT WORK WE HAVE DEALT WITH THE QUESTION OF WHAT HAPPENS, OF HOW TO DEAL WITH THE ISO 3166 MAINTENANCE AGENCY. QUITE RECENTLY, SERBIA AND MONTENEGRO SPLIT. THIS HAS -- OBVIOUSLY HAS -- IF CCTLD, IF THE CC, .CS WHICH USED TO BE CZECHOSLOVAKIA BUT THE SERBIA/MONTENEGRO INSISTED ON GETTING THAT TWO LETTER CODE. BUT THAT NO LONGER EXISTS. WE ARE GOING TO HAVE TO GET TWO NEW ONES. I DON'T KNOW IF ICANN TRACKS THAT BUT PRESUMABLY YOU DO.
THE STANDARDS NORMALLY, AND CERTAINLY THE LANGUAGE CODE STANDARDS, MAKE PROVISION FOR TRANSLATION OF THE CODE ELEMENTS INTO OTHER SCRIPTS. NOW, FOR SOME OF THOSE THINGS IT'S REALLY QUITE STRAIGHTFORWARD. FOR AN ALPHABETIC SCRIPT LIKE GREEK, CYRILLIC, GEORGIAN OR ARMENIAN, YOU COULD VERY EASILY MAKE A SET OF ABBREVIATIONS AND RUN IT BY SOME GREEK EXPERTS OR AUTHORITIES AND MAKE YOUR LIST. SOMEBODY HAS TO DO THAT BUT THERE'S NO REASON IN PRINCIPLE WHY SOME OF THE ALPHABETIC SCRIPTS COULDN'T ACTUALLY HAVE REASONABLE LISTS THAT YOU DON'T HAVE TO WORRY ALL THE TIME ABOUT WHETHER OR NOT THIS IS LANGUAGE OKAY. FOR SOMETHING LIKE ARABIC WHERE YOU GENERALLY HAVE TO WRITE OUT A WHOLE WORLD FOR A COUNTRY, BECAUSE IT DOESN'T MAKE SENSE TO ABBREVIATE, THAT'S FINE. BUT I THINK ON A SCRIPT-BY-SCRIPT BASIS, SOME OF THIS ACTIVITY WOULD NOT BE ALL THAT DIFFICULT. IT WOULD NOT TAKE MORE THAN THREE HOURS TO DRAW UP A LIST OF POSSIBLE CYRILLIC DOT CC THINGS. YOU JUST SAY IF YOU USE CYRIL!
LIC, THIS IS WHAT WE ARE USING AND THAT'S THAT.
NO ONE COMPLAINS IN GENERAL THAT DOT IE AS OPPOSED TO DOT IR IS IRELAND BECAUSE IR WAS GIVEN TO IRAN. NO ONE WHO USES THE LATIN SCRIPT CARES ABOUT THAT. WE JUST USE IT. SO THE QUESTION OF THE FUNCTIONALITY FOR SOMEONE WHO WANTS TO USE THEIR OWN SCRIPT, FOR SOME SCRIPTS THAT'S NOT A PROBLEM.
I THINK I MENTIONED THIS AT ONE OF THE OTHER MEETINGS, MAYBE LUXEMBOURG, MAYBE VANCOUVER, I DON'T KNOW. ICANN NEEDS TO TACKLE THIS. THEY NEED TO SET UP A GROUP OF PEOPLE THAT SAYS OKAY, SOMEBODY SORT THIS OUT. AND WHETHER THAT GOES THROUGH AN RFC PROCESS OR IT'S AN INTERNAL DOCUMENT OR WHATEVER, I DON'T KNOW. IT WOULD CERTAINLY BE OF GENERAL USE TO HAVE A LIST OF THESE THINGS AVAILABLE FOR ANYBODY WHO WANTS TO MAKE USE OF THESE STANDARDS. SO PUT IT ON YOUR AGENDA.
>>KELLY KANG: HI, I AM KELLY FROM NIDA. I HAVE A BRIEF COMMENT ABOUT THE IDN.
THE IDN GTLD POLICY NEEDS THE INPUT FROM NOT ONLY THE GNSO BUT ALSO THE CCNSO AND INDIVIDUAL COUNTRIES. ALSO IT NEEDS TO TAKE INTO CONSIDERATION SPECIAL CONDITIONS OF EACH CCTLD.
FOR EXAMPLE, IN CASE OF KOREA AS A NATION OF A SINGLE LANGUAGE, A SINGLE LANGUAGE CHARACTER SET, ONE CANNOT IGNORE THE FACT THAT ALL KOREAN TLDS WOULD AFFECT THE INTERESTS OF THE KOREAN SPEAKING COMMUNITY. THUS, IT IS ESSENTIAL THAT WE HAVE ENOUGH TIME TO GATHER THE OPINIONS OF GNSO AND CCNSO AND GAC, FORM COMMON GROUND. THANK YOU.
>>BRUCE TONKIN: WHAT'S THAT? I HAVE JUST BEEN INFORMED THERE IS AN AFTERNOON BREAK AND WE RECONVENE WHEN? AT 5:00 P.M.
OKAY.
SO ARE THERE ANY FURTHER COMMENTS ON THE POLICY WORK?
OKAY. AT THAT POINT, THEN, WE'LL RECONVENE AT 5:00 P.M. AND TAKE ADVANTAGE OF THE OPPORTUNITY TO HAVE A BREAK.
(BREAK UNTIL 5:00 P.M.)
>>PAUL TWOMEY: THIS NEXT SESSION IS DEDICATED TO SOME OF THE DISCUSSIONS SOME OF THE GAC MEMBERS HAVE BEEN HAVING AND WANTING TO HAVE.
IT'S AN ONGOING WORK IN THE GAC.
THE TWO-PART PRESENTATION, ANDRZEJ BARTOSIEWICZ IS GOING TO SHARE A PRESENTATION THAT HE'S BEEN SHARING WITH THE GAC MEMBERS ON SOME ASPECTS AND RISK ASPECTS ABOUT IMPLEMENTATIONS.
AND PANKAJ AGRAWALA FROM THE GOVERNMENT AND REPUBLIC OF INDIA WILL THEN SPEAK AFTER ANDRZEJ.
SO, ANDRZEJ, YOU GOT CONNECTED?
>>ANDRZEJ BARTOSIEWICZ: OKAY, MAYBE I WILL USE -- OKAY.
LET'S....
OKAY, SO THE TECHNICAL PROBLEMS ARE SOLVED.
MY NAME IS ANDRZEJ BARTOSIEWICZ, AND I WILL HAVE VERY BRIEF PRESENTATION ON EXISTING IDN.IDN RISKS.
AND THIS IS THE ISSUE I'VE BEEN -- I HAD A PRESENTATION FOR EUROPEAN GAC MEMBERS.
I WOULD LIKE TO TALK ABOUT THREE ISSUES WHICH HAVE THREE DIFFERENT PROBLEMS THAT ALREADY AROSE WITHIN THE IDN AT THE ROOT LEVEL.
THIS IS -- THE PRESENTATIONS AT THE PREVIOUS SESSION WERE ADDRESSING THESE ISSUES AS WELL AND WERE PREPARING -- THE DISCUSSION WAS HOW TO AVOID THE PROBLEMS AND HOW TO IMPLEMENT IDNS AT THE ROOT LEVEL.
I'M JUST TALKING ABOUT SOMETHING THAT ALREADY HAPPENED.
SO SOME IDN IMPLEMENTATIONS AT THE ROOT LEVEL, ACTUALLY, NOT ROOT LEVEL, BUT THE IMPLEMENTATIONS WHICH IN SOME CIRCUMSTANCES, LIKE BY SOME ISPS ALLOW USERS TO HAVE ACCESS TO SO-CALLED IDNS AT THE ROOT LEVEL.
BUT THIS IS NOT THE RIGHT SOLUTION, AND AS I UNDERSTAND, ICANN IS WORKING ON THE IMPLEMENTATION OF IDNS AT THE ROOT LEVEL.
THE FIRST ISSUE IS, WE HAVE SOME EXISTING CCTLDS THAT ALREADY OFFER IDN.IDN REGISTRATION.
WE HAVE SOME COMPANIES THAT THEY OFFER POSSIBILITY TO REGISTER DOMAIN NAMES IN -- UNDER -- LOOKING LIKE CCTLDS.
BUT THIS IS NOT THE REGISTRY, THE OFFICIAL REGISTRY, WHICH RUNS THE CCTLD.
AND THE THIRD POSSIBILITY ARE LOCAL GTLD IMPLEMENTATIONS.
OKAY.
THE FIRST SITUATION WE HAVE WHEN THE EXISTING CCTLD IS OFFERING TO THE CUSTOMERS POSSIBILITY OF REGISTRATION OF THEIR DOMAIN NAME UNDER SO-CALLED GLOBAL IDN TLD.
BUT THIS IS NOT WHAT EXISTS IN THE ROOT SERVERS.
THE EXAMPLES ARE CHINA, THERE ARE SOME ARAB STATES.
MY COLLEAGUE FROM IRAN WILL HAVE A PRESENTATION I THINK ON TUESDAY ON THEIR IMPLEMENTATION.
BUT, GENERALLY, THIS IS, I WILL SAY, SOME KIND OF TEST BED.
SO IF THE USER IS CONNECTED TO THE ISP, INTERNET SERVICE PROVIDER THAT OFFER RESOLUTION OF THOSE IDN CCTLDS, IT'S OKAY.
BUT IF THE USER IS, LET'S SAY, CHANGING THE ISP OR GOING ABROAD, HAS NO ACCESS TO SUCH DOMAIN NAME.
FROM MY PERSPECTIVE, IT IS NOT -- IT'S NOT THE RISK TO DNS, BECAUSE WHEN WE HAVE THIS PROCESS OF IDN IMPLEMENTATION AT THE ROOT LEVEL ALREADY FINISHED, THIS IS JUST A DELEGATION OF SUCH DOT IDN CCTLD TO THE EXISTING CCTLD MANAGER.
SO, FROM MY PERSPECTIVE, THIS IS RATHER I'LL SAY EVEN IT'S FACILITATING ROLE, IT'S NOT A THREAT TO DNS.
BUT OF COURSE IT CAN MAKE SOME CONFUSION.
THE SECOND SITUATION IS THAT THE COMPANY, THE THIRD PARTY, NOT THE OFFICIAL COUNTRY-CODE TOP-LEVEL DOMAIN MANAGER, IS OFFERING TO CUSTOMERS THE POSSIBILITY OF REGISTRATION THEIR DOMAIN NAMES UNDER IDN CCTLD.
VERY OFTEN, THE CUSTOMERS ARE NOT AWARE THAT THIS IS NOT -- FIRST, THAT IT'S NOT THE OFFICIAL CCTLD FOR THE COUNTRY; AND, SECONDLY, THAT THIS IS ONLY -- SUCH DOMAIN NAMES ONLY RESOLVES IN SOME NETWORKS WHEN THEY ARE CONNECTED TO SOME ISPS.
AND WHEN WE HAVE THE PROCESS OF IDN.IDN IMPLEMENTATION WITH ICANN, WE WILL END UP IN THE SITUATION WHO IS THE MANAGER OF THE CCTLD.
AND THE THIRD POSSIBILITY, WHICH MAKES CURRENTLY CONFUSION TO THE INTERNET USERS, AND THIS IS, I THINK, VERY VISIBLE, AND MY COLLEAGUES FROM VERISIGN, AFILIAS, NEUSTAR CAN SHARE THEIR VIEWS PROBABLY, WHEN THERE IS A THIRD PARTY WHICH OFFERS REGISTRATION UNDER LOCAL DOT COM, DOT NET, DOT ORG, WE HAVE SEVERAL EXAMPLES.
WE HAVE ONE EXAMPLE REGISTERING DOMAIN NAMES IN CYRILLIC IN CYRILLIC DOT COM, DOT NET, DOT ORG.
WE HAVE COMPANY OFFERING DOMAIN NAMES IN KOREAN, IN HEBREW, ET CETERA.
AND WHEN THE USER WHO IS USING SUCH A DOMAIN NAME IS CHANGING THE ISP, OR ESPECIALLY FLYING ABROAD AND MOVING TO U.S. AND TRIES TO RESOLVE SUCH DOMAIN NAME, THERE'S A PROBLEM.
SO THEY ARE CALLING VERISIGN, WHY THE DOMAIN NAME IS NOT RESOLVING.
AND, OF COURSE, IT'S NOT THE VERISIGN OR WHOEVER'S FAULT.
IT IS -- THIS DOMAIN NAME DOES NOT EXIST.
AND HOW IT'S POSSIBLE?
VERY BRIEFLY, THE TECHNICAL ISSUES, WE CAN HAVE ONE POSSIBILITY IS THE PATCHING DNS.
IT'S VERY BAD SOLUTION.
WHEN THE ISP ALLOWS RESOLVING DOT IDNS DOMAIN NAMES NOT IN THE ROOT SERVER, BUT IN SOME OTHER NATIVE ROOT.
AND THE SECOND POTENTIAL SOLUTION, THEY ARE PLUG-INS TO THE INTERNET BROWSERS.
BUT THIS SOLUTION, IT'S NOT VERY POPULAR, BECAUSE, GENERALLY, USERS ARE NOT VERY KEEN ON DOWNLOADING NEW VERSIONS.
THIS ISP-BASED SOLUTION, ISP HAS TO MODIFY THE RESOLUTION, DOMAIN NAMES RESOLUTION PROCESS.
THEY HAVE TO RESOLVE SOME DOMAIN NAMES NOT IN THE GOLDEN DNS TREE, BUT IN SOME -- IN SOME OTHER ALTERNATIVE ROOT SERVERS.
AND, OF COURSE, WHEN USER IS NOT USING THAT ISP ANYMORE, WE END UP IN THE SITUATION THAT THIS DOMAIN NAME DOESN'T RESOLVE, AND USER IS CONFUSED.
IN THIS CASE, THE KEY ROLE OF THE INTERNET SERVICE PROVIDERS, AS I'VE BEEN TOLD, THERE IS MORE THAN 100 ISPS ALREADY OFFERING SUCH SERVICES.
STATISTICS, THERE IS MORE THAN 20,000 IDN.IDN DOMAIN NAMES ALREADY REGISTERED, IF WE CAN SAY "REGISTERED."
IT'S NOT AN IMPRESSIVE NUMBER, BUT, ANYWAY, THIS CONFUSES INTERNET USERS.
IF YOU ARE INTERESTED IN SOME RECENT NEWS, HERE ARE THE THREE LINKS TO PRESS RELEASES WHICH ADDRESS THIS ISSUE.
THANK YOU.
SHUKRAN.
WE HAVE THE Q AND A AFTER THE PRESENTATION?
ANY QUESTIONS?
>>MICHEL SUIGNARD:.
IT'S MICHEL SUIGNARD.
YEAH, IT WAS MY IMPRESSION THAT MOST OF THE COUNTRIES THAT DID IDN.IDN WERE IN FACT ADDING, YOU KNOW, THE ROOT LEVEL NAME AT THE END, SO THEY'RE REALLY STILL RESOLVING AS -- FOR EXAMPLE, CHINA WILL ADD, IN FACT, AFTER THE CHINESE NAME, THEY WILL ADD THE DOT CN KIND OF HIDDEN FROM THE USER, BUT STILL WHEN YOU ARE GOING ON THE RESOLVER, IN FACT, YOU ARE APPENDING THE RIGHT WAY.
SO YOU WOULD JUST RESOLVE ON THE NORMAL NAME SPACE OF DNS.
COULD YOU CONFIRM OR --
>>ANDRZEJ BARTOSIEWICZ: I CANNOT SPEAK ON CNNIC.
AS I KNOW, THIS SITUATION IS IN .IR.
BUT THERE IS, I WOULD SAY, NO CONFUSION BECAUSE THE SAME -- THE ASCII EQUIVALENT IS ALREADY REGISTERED.
BUT --
>>MICHEL SUIGNARD: MY POINT WAS REALLY THAT YOU'RE MAKING SOME, YOU KNOW, VERY, YOU KNOW, BLUNT, I WOULD SAY, AN ASSERTION ABOUT SOME OF THE BEHAVIOR OF SOME OF THE --
SO YOU HAVE TO BE -- BECAUSE THAT WAS SAID, YOU KNOW, THERE WAS A LOT OF RUMORS, YOU KNOW, ON THE INTERNET AT LARGE ABOUT THIS HAPPENING.
IN FACT, WHEN WE WENT TO THE ROOT OF IT, WE FOUND, IN FACT, THEY WERE JUST RESOLVING.
I DON'T DISAGREE THAT THERE ARE SOME SITUATIONS OF SOME ISPS DOING EXACTLY WHAT YOU SAID.
BUT SO FAR, MY UNDERSTANDING IS THAT NO CCTLDS ARE DOING THAT.
YOU KNOW, THEY'RE STILL ADDING THE CORRECT SUFFIX AT THE END.
AND I WOULD LIKE, OBVIOUSLY, TO KNOW WHAT'S GOING ON IF IT'S NOT -- THAT'S THE CASE.
>>VINT CERF: IT'S VINT CERF.
I DON'T KNOW IF PROFESSOR QIAN IS IN THE ROOM, BUT IF HE ISN'T, I'M GOING TO -- ARE YOU HERE?
OH, OKAY.
YOU TELL ME IF I SAY THIS WRONG.
MY UNDERSTANDING IS THAT IN CHINA, IN FACT, THE DOT CN IS ADDED TO THE CHINESE CHARACTERS THAT LOOK TO THE USERS AS IF THEY ARE TOP-LEVEL DOMAINS.
BUT THE .CN IS APPENDED AFTER THE LOOKUP TAKES PLACE.
THE GOOD PART IS THAT THEY DON'T SEE THE ROMAN CHARACTERS AND THAT MAKES A MORE COMFORTABLE INTERACTION.
THE BAD PART, OF COURSE, IS THAT IF THEY SHOULD PASS THE THINGS THAT LOOK LIKE DOMAIN NAMES TO THEM WITH THE TOP-LEVEL DOMAINS IN CHINESE TO ANYONE ELSE IN THE WORLD, THEY ATTEMPT TO LOOK THEM UP, IT WON'T WORK BECAUSE THEY RESOLVERS DON'T HAVE THE SPECIAL FEATURE THAT .CN HAS ADDED, BECAUSE IT RECOGNIZES THOSE SPECIAL CASES.
AND IT'S EXAMPLES LIKE THAT, I THINK, THAT ANDRZEJ, I HOPE YOU WOULD AGREE, ARE REALLY PROBLEMATIC, BECAUSE THE FREEDOM TO EXCHANGE POINTERS AROUND THE WORLD SO PEOPLE CAN FIND THINGS BECOMES VERY COMPLEX IF WE DON'T HAVE A SINGLE AND UNIFORM LOOKUP IN THE DOMAIN NAME SYSTEM.
DID I SAY IT RIGHT?
OKAY.
>>RAM MOHAN: HI, RAM MOHAN.
I'M FROM AFILIAS.
ANDRZEJ, I WANTED TO FURTHER EMPHASIZE ONE OF THE THREATS THAT YOU HAD TALKED ABOUT, WHICH WAS WHAT WOULD -- I THINK IN ONE OF YOUR SLIDES, YOU HAD SOMETHING THAT SAID, IF YOU'RE TRAVELING ABROAD, A DOMAIN NAME IN A DOT IDN MIGHT WORK IN THE COUNTRY, BUT IF YOU TRAVEL ABROAD, IT MAY STOP WORKING.
IT'S ACTUALLY -- IT CAN GET EVEN WORSE THAN THAT.
IN THE COUNTRIES THAT WE KNOW OF, MANY OF THEM HAVE SOME COORDINATION BETWEEN ISPS THAT ACTUALLY DO SOME COHERENCE IN HOW THEY ARE RESPONDING.
BUT IT IS CERTAINLY POSSIBLE TODAY THAT MULTIPLE ISPS MAY IMPLEMENT MULTIPLE INSTANCES, IF YOU WILL, OF A QUOTE, UNQUOTE, IDN TLD.
SO, LITERALLY, YOU CAN BE IN THE SAME COUNTRY, SWITCH INTERNET SERVICE PROVIDER, AND TYPE THE SAME STRING AND GO TO A COMPLETELY DIFFERENT PLACE.
SO THAT THREAT IS PROBABLY EVEN WORSE THAN -- YOU DON'T HAVE TO LEAVE THE COUNTRY, YOU DON'T EVEN HAVE TO GET UP FROM YOUR SEAT TO HAVE THIS HAPPEN.
>>ANDRZEJ BARTOSIEWICZ: OF COURSE, I CAN ONLY CONFIRM THAT I SIMPLIFY THIS WITH TRAVELING ABROAD, BECAUSE THIS IS VERY VISIBLE.
BUT, YES, EVEN IF YOU ARE USING, I DON'T KNOW, WITH YOUR MOBILE PHONE AND YOU TRY TO OPEN THE WEB PAGE TYPING A DOMAIN NAME IDN.IDN, THE GTLD EQUIVALENT OF DOT COM, NET, ORG, AND SO ON, AND DEPENDING IN WHICH VLAN, OR DEPENDING TO WHICH (INAUDIBLE) TS OPERATOR YOU ARE CONNECTED, YOU CAN GET THE DIFFERENT RESULTS.
SO IT'S REALLY -- IT'S REALLY, I THINK, DANGEROUS AND INTERNET USERS MAY LOSE THE TRUST IN IDNS WHEN SUCH SITUATIONS WE ARE FACING NOW.
>> YES.
JAE CHUL FROM KOREA.
I SEE YOUR EXAMPLE IN KOREA.
BUT IN KOREA, THERE IS A KEY WORD SYSTEM, IT'S NOT DOMAIN NAMES.
SO IN KOREA, THERE IS TOO MANY DOMAIN NAMES -- NO, KEY WORD NAMES, NETPR, ANY OTHER NAMES.
SO BUT YOUR CASE IN YOUR PRESENTATION IS NOT CORRECT.
SO NEXT TIME, PLEASE CONTACT WITH ME.
I WILL CORRECT YOU.
THANK YOU.
>>ANDRZEJ BARTOSIEWICZ: ACTUALLY, I KNOW ABOUT THE KEY WORDS.
AND THE KEY WORDS IS AN OKAY SOLUTION, BECAUSE IT'S NOT MIXING WITH IDNS, WITH DOMAIN NAMES.
BUT I CAN SEND YOU THE LINK TO THE PRESENTATION FROM THE COMPANY THAT OFFER DOMAIN NAMES.
AND I AGREE WITH THE KEY WORDS.
I KNOW ABOUT THE KEY WORDS.
BUT THERE IS DOMAIN NAMES DOT COM AND DOT NET IN -- OFFERED BY ONE COMPANY IN KOREA.
IF WE HAVE TIME, I CAN OPEN THE PRESENTATION, MAYBE LATER IN THIS SESSION.
ACTUALLY, YOU CAN FIND THIS PRESENTATION IF YOU OPEN THE SECOND LINK.
AND ONE OF THE PRESENTATIONS FROM KOREA IS ON THAT ISSUE.
OKAY.
THANK YOU VERY MUCH.
[ APPLAUSE ]
>>PAUL TWOMEY: OKAY.
THE NEXT SPEAKER IS PANKAJ AGRAWALA FROM THE GOVERNMENT OF INDIA.
>>PANKAJ AGRAWALA: THANK YOU, PAUL.
THIS IS MORE IN THE FORM OF A REVIEW OF THE IDN WORKING GROUP'S ACTIVITIES IN THE GAC AND HOW FROM THE TIME THAT THIS WORKING GROUP WAS INSTITUTED.
THE WORKING GROUP HAS BEEN PERFORMING TWO BASIC FUNCTIONS.
ONE OF THEM IS TO FAMILIARIZE THE MEMBERS OF THE GAC WITH THE TECHNICAL AND PUBLIC POLICY ISSUES THAT ARE INVOLVED.
AND THEN, OF COURSE, TAKE A POSITION AND ADVISE THE BOARD ON THE PUBLIC-POLICY MATTERS PERTAINING TO IDNS.
IN THE BEGINNING, THE ISSUES THAT MAINLY WERE ADDRESSED IN THE WORKING GROUP WERE ON THE TECHNICALITIES AND THE COMPLEXITY OF THE WHOLE ISSUE.
BUT GRADUALLY, AS ICANN ITSELF STARTED -- INSTITUTED THE PRESIDENTIAL ADVISORY COMMITTEE, AND THEN WE HAD THE TWO TESTS, TECHNICAL TESTS, THAT WERE ANNOUNCED.
AND THAT GOT A TURN IN THE SYSTEM.
AND THE GNSO HAS COME OUT WITH ITS 5TH DECEMBER PAPER.
AND THE CCNSO, BOTH OF WHICH WERE PRESENTED IN THE SESSION BEFORE US.
AND BEFORE -- I NEED NOT REPEAT, BUT I NEED TO ENDORSE THE ISSUES THAT HAVE BEEN IDENTIFIED BY BOTH THE SPEAKERS BRUCE TONKIN AND HIRO HOTTA AS LIVE ISSUES.
AND THE GAC ALSO FEELS THAT AT EVERY LEVEL, FROM THE TECHNICAL SIDE TO THE PUBLIC-POLICY SIDE, THERE ARE COMPLEXITIES.
THE COMPLEXITY OF THE LANGUAGES SPOKEN IN DIFFERENT PARTS OF THE WORLD. THERE ARE SCRIPT VARIATIONS, GEOGRAPHICAL DIVERSITIES, SOCIAL AND CULTURAL DIMENSIONS, AND OF COURSE WE HAVE PHONETIC AND VISUAL VARIATIONS.
SO THEREFORE, IT'S NOT -- IT'S PERMUTATION, COMBINATION OF SO MANY VARIATIONS THAT WE HAVE A CHALLENGE BEFORE US.
HOWEVER, WHAT WE FELT THAT THERE HAS TO BE SOME KIND OF A SIMPLIFICATION OF THE PROCESSES.
AND THE IDEA IS TO MAKE IT AS SIMPLE AS POSSIBLE. AND THEREFORE, WE HAVE -- I HAVE DESIGNED A SMALL NOTE BASED ON THE EXPERIENCES OF ALL THE SPEAKERS. AND WHAT WE FEEL IS THAT IN THE FIRST STEP, WE CAN TAKE THE DEVELOPMENT OF THE BASIC LANGUAGE TABLE, SO THE TASKS IN IT WOULD BE IDENTIFY THE COLLECTION OF CHARACTERS FOR A GIVEN LANGUAGE TO BE LAUNCHED. IDENTIFY THE POTENTIAL VARIANTS WITHIN THE LANGUAGE, LIST OF CHARACTERS. IDENTIFY POTENTIAL VARIANTS WITH OTHER UNICODE CHARACTERS. DETERMINE LIST OF VALID CHARACTERS FOR REGISTRATION.
THEN WE CAN START TO ADDRESS CERTAIN PUBLIC-POLICY ISSUES, EITHER WHICH EMANATE FROM IPR AND CONTEXTUAL CONSIDERATIONS.
SO THERE THE TASK WOULD BE DETERMINE ADDITIONAL LIST OF RESERVE NAMES IN THE LANGUAGE TO BE LAUNCHED. DETERMINE IF SUNRISE PERIOD IS NECESSARY, AND IF BROADER VARIANT RULES OR FURTHER IP ISSUES, WE HAVE TO MAKE THE RULES. DETERMINE ADDITIONAL CONTEXTUAL RULES, MAXIMUM/MINIMUM LENGTH, INVALID PREFIXES, SUFFIXES. DRP, THE DISPUTE RESOLUTION PROCESSES.
ADDITIONAL CONSIDERATIONS ARE THE PRICING AND THE LAUNCH PROCEDURES.
I WOULD END MY -- THIS RECOMMENDATION WITH A SMALL, MAYBE A PHILOSOPHICAL STATEMENT THAT THE WRIGHT BROTHERS PROBABLY WOULD NOT HAVE BEEN ABLE TO FLY THEIR AIRCRAFT IF THEY HAD WAITED FOR THE AIR TRAFFIC CONTROLLER AND ITS REGIME TO BE SET UP BEFORE THEY TOOK OFF.
SO THE ISSUE IS MANY TIMES TECHNOLOGY MOVES AHEAD WITH A LITTLE DISREGARD FOR THE REGIME THAT EXISTS. IT MAY NOT BE FULLY CONDUCIVE, BUT IF THE EQUIVALENT OF A GAC WERE THERE TO CONSTANTLY KEEP LOOKING AT ALL THE PUBLIC-POLICY ISSUES, THEN THE MOTOR CAR MANUFACTURER WOULD NOT HAVE BEEN ABLE TO MANUFACTURER A CAR WHICH WOULD HAVE TRAVELED AT 300 KILOMETERS PER HOUR, BECAUSE EVERY COUNTRY WOULD HAVE TAKEN THE POSITION THAT OUR MAXIMUM SPEEDS ARE ABOUT 85 MILES, 110, 130, 140 KILOMETERS, OR 200.
SO THE ISSUE I AM TRYING TO BRING FORTH IS THAT IT IS TIME WE IDENTIFIED SOME BASIC PUBLIC-POLICY ISSUES AND REDUCED THE COMPLEXITIES TO A POINT WHERE WE CAN GO AHEAD WITH THE IDNS. BECAUSE THE MAJORITY OF THE WORLD TODAY DOES NOT SPEAK OR WRITE ENGLISH.
THANK YOU.
I CAN TAKE QUESTIONS.
THERE WON'T BE ANY.
>>PAUL TWOMEY: YOU HAVE OBVIOUSLY SPOKEN WITH WISDOM THERE, PANKAJ. I THINK YOU HAVE INSTILLED SILENCE IN THE AUDIENCE.
OKAY. WELL, PERHAPS -- THANKS FOR THAT, AND FOR THOSE GOOD CLOSING WORDS.
PERHAPS WE'LL TAKE THAT AS THE OPPORTUNITY TO CLOSE THIS SESSION, THANK THE SPEAKERS FOR THEIR CONTRIBUTIONS. LET'S PLEASE ENSURE WE HAVE COPIES OF PRESENTATIONS FOR THE TECHNICAL STAFF, STEVE CONTE'S TEAM, OR IF STEVE IS IN THE ROOM HERE, AND THIS AGENDA ITEM WILL RECONVENE, AT LEAST IN A PUBLIC FORUM, ON TUESDAY AFTERNOON, MOROCCO TIME. ALTHOUGH I KNOW IT'S BEING DISCUSSED IN VARIOUS WORKING GROUPS IN THE INTERIM PERIOD.
THANK YOU VERY MUCH, EVERYBODY. THANK YOU FOR PARTICIPATING. WE'LL SEE YOU AGAIN ON TUESDAY, EXCEPT AS IS NOT UNCOMMON IN MY LIFE EXPERIENCE, THE CHAIRMAN WISHES TO....
THANK YOU FOR THAT.
THANK YOU FOR THAT, VINT.
JUST A GENERAL ANNOUNCEMENT TO REMIND THE AUDIENCE THAT IN ROOM REDA, I HAVE 4 HERE. REDA 4 THAT THERE IS GOING TO BE A BRIEFING ON CONSUMER PROTECTION ISSUES, WHICH WILL INVOLVE SOME KEY CONSUMER PROTECTION AGENCIES FROM EUROPE, NORTH AMERICA AND ASIA. SO I THINK PEOPLE WILL FIND THAT A VERY INTERESTING AND I THINK A VALUABLE BRIEFING.
I WOULD EXHORT AS MANY PEOPLE AS POSSIBLE TO ATTEND.
THANKS VERY MUCH.
[ APPLAUSE ]

© Internet Corporation for Assigned Names and Numbers