Skip to content

Instantly share code, notes, and snippets.

@mrazzari

mrazzari/IDN not ready for prime time.markdown

Last active December 19, 2015 11:49
Show Gist options
  • Star 0 You must be signed in to star a gist
  • Fork 0 You must be signed in to fork a gist
  • Save mrazzari/5950011 to your computer and use it in GitHub Desktop.
Save mrazzari/5950011 to your computer and use it in GitHub Desktop.
Download ZIP
My experience setting up a WP site using an Internationalized domain name. TL;DR: it can work, but requiring extra developer effort, and being noticeably unreliable. OK for blogs / early adopters.
Raw
IDN not ready for prime time.markdown

A couple of years ago I helped organize this usability conference in Buenos Aires: http://www.diseñoinclusivo.org.ar/

Notice the "ñ" for diseño (design) in the URL above.

This URL is there "just for fun". We ended up using the anglicized version, http://www.disenoinclusivo.org.ar for a bunch of reasons.

WordPress issues:

  • The WP admin login breaks when using IDN.

  • So does the Flash file uploader.

  • There are a bunch of plugins "fixing" IDN issues: Random sample: 1, 2, 3.

  • Googling for WP and IDN will show a lot of issues.

  • I didn't even research WP multi-site issues with IDN, or issues with specific plugins.

The WP issues can be fixed with some research effort, plugins and workarounds. But it seems really fragile, even today. Error prone, glitches just waiting to happen.

These fixes are way more difficult when the domain is replicated by third parties we don't control. An URL is something that gets passed through a lot of places automatically, including RSS, social media posts, content aggregators, all kinds of parsers.

It's very likely it will be processed, and break or look ugly. Examples below:

  • Some browsers, with certain security settings, will show punycode (the domain's ASCII version): xn--diseoinclusivo-tnb.org.ar. This is the default in IE7 and above. It also shows a yellow information bar and alert box. (IDNs aren't even supported by IE6, pages don't load at all)

  • A parser that can't handle Unicode will break, and show this: diseÃñoinclusivo.org.ar.

  • It could be URL encoded, looking like this: dise%F1oinclusivo.org.ar.

  • Many (most?) filters to validate URLs and emails only allow these characters in a domain name: A-Z a-z 0-9 - _ .. So our domain will be rejected, or worse, might silently fail, losing our ñ character: diseoinclusivo.org.ar.

  • It can present an access barrier, as typing special characters often requires a combination or succession of keystrokes. Less tech-savvy people may not know how to do that, or may have a wrong keyboard configuration that makes it difficult, such as mobile or internet cafe users.

Bottom line, it could work, but requiring extra developer effort, and being noticeably unreliable.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment