Total IDN count update: 550k (com + net)

[sramuk]

Total IDN count update: 550k (com + net), July 2006
Four months after owen of Domainsite quantified total number of com and net IDNs to be ~425k (http://www.idnforums.com/forums/1358-possibly-interesting-idn-stats.html), I was able to recalculate it to have grown to 550k. Thats about 30% growth or an annualised 90% growth. We can surely attribute a significant part of it to our Olney's IDN Forum and the members here (thats without counting ccTLDs like jp and cn). Excluding the flat growth in the large Hangul(4%) and Latin(1%) script domains, the growth is an impressive 75%. Even though domains in most other languages have increased dramatically, there are 90k more domain names in Han script alone and only 35k more domain names all in other languages put together. Interest in Thai domains seem to have increased dramaticaly though from a smaller base.
I hope others might be able to spot additional trends from this data.



Unicode Script July March Growth
-------------- --------- --------- ------
Han 208063 114704 81%
Hangul 163270 157612 4%
Latin 113136 111660 1%
Katakana 26852 18198 48%
Cyrillic 11397 6490 76%
Arabic 7021 4327 62%
Hiragana 6505 4457 46%
Hebrew 3402 2105 62%
Greek 2652 1854 43%
Common symbols
/and Other 2587 1823 42%
Thai 2147 669 221%
Devanagari 1966 1137 73%
Tamil 391 139 181%
Gurmukhi 193 46 320%
Gujarati 179 76 136%
Bengali 223 174 28%
Telugu 142 94 51%
Kannada 120 78 54%
Georgian 68 80 -15%
Ethiopic 68 77 -12%
Armenian 46 37 24%
Malayalam 34 12 183%
Oriya 6 12 -50%
Tibetan 6 50 -88%
Khmer 9 2 350%
Lao 8 2 300%
Myanmar 3 8 -63%
Runic 3 2 50%
Sinhala 2 0 0%
Canadian Aborg 3 16 -81%
Cherokee 1 3 -67%
Yi 1 1 0%
-------------- --------- --------- -------
Total 550504 425945 29.24%

p.s.
Instead of averaging the code points like owen did, I tried to identify the first non-latin unicode character in the domain to determine the script. This might have caused the dips in some langauges (or is it drops?). I am not sure if I have to separately count the still registered testbet bq race code domains and if that has impacted the results.



Edit: Formatting.

[touchring]

Amazing, RD's portfolio is like 1% of the total. I bet the lot of us here own like 15% of all IDNs.

[alex]

Thanks for the interesting data.

How are Japanese Kanji counted or which category do they fall under? How are mixed Kana and Kanji or mixed Latin and Kana/Kanji counted?

EDIT: Reading your notes, I see that you used the first non-latin character to determine the category.

[sramuk]

How are Japanese Kanji counted or which category do they fall under?


From what I understand, Japanese Kanji (http://en.wikipedia.org/wiki/Kanji) characters are part of Unicode Han script (http://en.wikipedia.org/wiki/Han_unification)along with Chinese Hanzi (http://en.wikipedia.org/wiki/Hanzi)and Korean Hanja (http://en.wikipedia.org/wiki/Hanja) characters.

[blastfromthepast]

From what I understand, Japanese Kanji (http://en.wikipedia.org/wiki/Kanji) characters are part of Unicode Han script (http://en.wikipedia.org/wiki/Han_unification)along with Chinese Hanzi (http://en.wikipedia.org/wiki/Hanzi)and Korean Hanja (http://en.wikipedia.org/wiki/Hanja) characters.

Correct. It is possible only to discern scripts, not languages using Unicode.

[bwhhisc]

Is there any effect about the many registrations done accidently in "Africcaans" by forgetting to tag the language code.

[blastfromthepast]

Is there any effect about the many registrations done accidently in "Africcaans" by forgetting to tag the language code.

Language code is used only during the registration process. It is not retained and does not tag the domain in any way.

[alex]

From what I understand, Japanese Kanji (http://en.wikipedia.org/wiki/Kanji) characters are part of Unicode Han script (http://en.wikipedia.org/wiki/Han_unification)along with Chinese Hanzi (http://en.wikipedia.org/wiki/Hanzi)and Korean Hanja (http://en.wikipedia.org/wiki/Hanja) characters.

Thanks for the clarification. I wonder how the percentages work out for each language.

[Rubber Duck]

Language code is used only during the registration process. It is not retained and does not tag the domain in any way.

Where did you get that information? That is one that has been worrying quite a lot of people. I have always been relaxed about this because I felt it was one that Verisign and ICANN would have to sort out, but it would be nice to get some clarifiication.

OK, own up you lot. Who has been hitting Sikh or Gurmuki?

[Olney]

Yeah I know that some of the original programmers for IDNs were or are on the site. Are you part of the developing team?
Also thank you for giving credit to the site for expanding the increase of IDNs. Hopefully soon I can do the same for the Japanese Internet Market.

[thefabfive]

:o I seem to own 10% of the Hebrew domains.

Great info, sramuk. Thanks.

[sramuk]

Yeah I know that some of the original programmers for IDNs were or are on the site. Are you part of the developing team?


If the question is to me whether I am associated with Domainsite, No, I am not. Owen is.

[Olney]

No not for DomainSite
Some of the actual programmers for the actual IDN system were on here. That was my question.


[QUOTE=Olney]
Yeah I know that some of the original programmers for IDNs were or are on the site. Are you part of the developing team?
[QUOTE]

If the question is to me whether I am associated with Domainsite, No, I am not. Owen is.

[sramuk]

No not for DomainSite
Some of the actual programmers for the actual IDN system were on here. That was my question.


Wow, we could get a lot of things clarified by them.
Not me.

[mulligan]

Yeah I know that some of the original programmers for IDNs were or are on the site. Are you part of the developing team?
Also thank you for giving credit to the site for expanding the increase of IDNs. Hopefully soon I can do the same for the Japanese Internet Market.

Where was this credit given?

[Olney]

In his first post... not on another site..
He just acknowledged that the members & forum increased the registrations of IDNs...

Where was this credit given?

[Drewbert]

Language code is used only during the registration process. It is not retained and does not tag the domain in any way.

I can confirm this.

The language tag is used during the nameprep stage of the registration process. Once
the "all clear" is given to the string, it's registered. The language tag is NOT stored at Verisign in the registration database.

[touchring]

I can confirm this.

The language tag is used during the nameprep stage of the registration process. Once
the "all clear" is given to the string, it's registered. The language tag is NOT stored at Verisign in the registration database.


Phew! I can sleep tonite, knowing that my names are not Afrikaan. :p

[wdsbg]

GO IDN's GO .... ;)

[Rubber Duck]

Phew! I can sleep tonite, knowing that my names are not Afrikaan. :p

One more stupid thing that many have worried themselves silly about now just evaporates in a cloud of smoke.

Common you lot, there must be another boggieman story out there somewhere to frighten the Newbies?

[Drewbert]

>Common you lot, there must be another boggieman story out there somewhere to frighten the Newbies?

Did you hear the one where you must make an annual sacrifice of a "top level premium" virgin IDN to Drewbert in order to keep him happy?

[Rubber Duck]

>Common you lot, there must be another boggieman story out there somewhere to frighten the Newbies?

Did you hear the one where you must make an annual sacrifice of a "top level premium" virgin IDN to Drewbert in order to keep him happy?

No, we've been trying to keep the whole Drewbert thing under wraps. Way too scary!

[touchring]

One more stupid thing that many have worried themselves silly about now just evaporates in a cloud of smoke.

Common you lot, there must be another boggieman story out there somewhere to frighten the Newbies?


Jokes aside, there are always risks involved in domaining. A group of investors just lost 74,000 EU domains over nite. See details: http://www.idnforums.com/forums/5353-eurid-suspends-74-000-eu-domain-names-and-has-sued-400-registrars.html

[Drewbert]

There certainly are risks when you try and do an end-run around the rules.

ICANN might not act on rule breakers (because it suits them not to) but the Europeans obviously do.