FYI
http://www.circleid.com/posts/20140519_universal_acceptance_of_all_tlds_now/
A piece by Stéphane Van Gelder on Universal Acceptance of TLDs.
Universal acceptance of top level domains hasn't really meant much to
most Internet users up until now. As long as .COM was basically the default TLD, there wasn't much of an issue.
No longer. With 263 delegated strings (according to ICANN's May 12, 2014
statistics) adding to the existing 22 gTLDs that were already live on the net after the 2004 round of Internet namespace expansion, the problem of universal acceptance gets very real.
What is the issue? Being able to use any TLD with any browser (Safari,
Chrome, Firefox...) or email client (outlook, Apple mail...), no matter what the environment (Mac, Windows...). Non-ASCII Internationalized Domain Names (IDNs) have made the issue even more crucial because universal acceptance is needed to fulfil their promise: giving millions of people who use different keyboards and scripts access to an "own language" Internet use they've never been able to experience up until now.
Non-Latin script TLDs are considered one of the most important Internet
naming innovations of the last decade because they are key to reducing the digital divide between those who are comfortable typing ASCII web addresses and those who are not. Those whose native alphabets are Chinese, Arabic, Cyrillic or others have jumped at the chance to type their own scripts into the address bars. The high registration volumes reached by IDN ccTLDs like .рф (Romanised as .RF for Russian Federation) or more recently the Chinese character new gTLDs launched by TLD Registry .在线 (.ONLINE) and .中文网 (.WEBSITE) are clear indications of the pent-up demand for non-Latin script web addresses.
Ensuring new web addresses can be used everywhere, by everyone
Because they are that important, ICANN's new gTLD program has given
deliberate priority to IDNs. The 104 applications for non-Latin script TLDs were allowed a faster track through the evaluation and delegation process if they wanted it.
Despite that, most new gTLDs that have gone live so far are of the
traditional ASCII kind. But as TLD Registry's suffixes show, the Internet is not going to stay all-Latin for long. Others also have ambitious plans to bring Internet navigation to new audiences using different alphabets. Stable Tone is one such registry operator. It will soon be launching 世界, (.WORLD in Chinese) with an innovative plan to bring international brands to Chinese audiences.
Without universal acceptance, those plans will suffer. The products will
be out there, but their intended audiences won't actually be able to use them. Why? Because in many ways, the Internet still works as if there had never been an expansion of its namespace. There are 3 main issues.
Internet browsers tend to carry out summary checks when a destination is
typed into their address bars. Some browsers will actually filter out anything that doesn't resemble the "classic" TLDs. Type www.example.guru, one of the new TLDs that is already proving very popular, and browsers expecting .COM or something similar will either tack that suffix on the end and search for www.example.guru.com, or take the user to a search engine to try and "help".
The rate of expansion is also an issue. In 2004, only a very limited
number of TLDs were added to the Internet root. So as suffixes like .BIZ and .INFO came online, browser developers had time to adjust. But the new gTLD program means an increase of several orders of magnitude for the number of TLDs becoming valid Internet namespaces. How can developers of browser, email and other Internet services keep up?
IDNs have brought a third universal acceptance issue. Users must be able
to type, and see, their URLs in the intended native script. If it's a Chinese address, you should be able to type that in as easily as if it was ASCII. Having to resort to some kind of encoder to get the Chinese script complicates the user experience to such an extent that there's a risk users simply won't bother. And that could ultimately mean the failure of IDN TLDs that, ironically, Internet users actually want.
There's a fourth, slightly separate but equally important issue: mobile
device use. Browsing and emailing environments tend to be even more technically limited on smartphones and tablets than they are on desktop computers. Yet those are the devices which the next billion Internet users, those that stand to benefit from new TLDs and especially IDNs, are most likely to use.
All that leads to a situation where currently, there is a lack of
universal acceptance for these new TLDs. Take those mentioned earlier. A web address like 房地产.在线 (RealEstate.online) simply isn't reachable using the majority of web browsers running on PCs and mobile devices in China. A part-ASCII address like nic.在线 has the same issues.
Chinese data shows that around 50% of the browsers used in the country
are not compatible with Chinese IDN gTLDs and that on mobile devices, tha vast majority of the browsers available aren't compatible.
Single focus
So what to do? Awareness of the problem isn't new. ICANN, for instance,
has been working on it since 2004, when it opened a discussion forum on universal acceptance, and even before. Technically there are no major hurdles to universal acceptance. For example, the technology needed to correctly render IDN web addresses was developed even before the new gTLD program, when work was being done on the IDN ccTLD program designed to give a path to non-ASCII country codes to countries such as Russia, where .рф is considered the second national suffix after the ASCII .RU.
The major stumbling block for universal acceptance today is probably
coordination. And let's face it, that's an area where organisations like ICANN have traditionally not been very strong. ICANN, the technical community's IETF, and others, are very good at taking on a variety of problems at once. They are not so good at focusing their collective energies on a single problem.
Yet that's exactly what's needed to guarantee universal acceptance.
Across-the-board efforts targeting all aspects of this one problem. For example, if technical operators such as browser developers are not made sufficiently aware of new gTLDs by the governance community, then how can they be expected to adapt their products to suit?
It's incumbent on the technical and governance communities to ensure that
demand is met. Disappoint these non-ASCII hopefuls now, and new gTLDs as a whole will suffer the consequences.