On Mar 17, "D. J. Bernstein" <djb@cr.yp.to> wrote: > Internationalized domain names are a failure if non-ASCII glyphs don't > appear on the screen. Have you completely lost sight of the goal here? I'm just going to whack my 2p in here... Would I be right in understanding that what you're saying is that if a given application does not display non-ASCII glyphs correctly, when attempting to display IDNs, then you would categorise IDNs as having been a failure? If that is the case, you had best prepare for failure, as what you're asking for is for every OS vendor/contributor (*BSD, Linux as well as the MS, Sun, Cisco(!!), and VMS crowds) to start supporting IDNs before they can be deployed. In fact, we should be altering it at an internal level within every OS, and ensure old OSes are destroyed, so as to make sure application developers don't have to worry about it. But of course, they will, and so all software ever written needs to be modified if it ever had the ability to understand domain names. What you're suggesting is IMHO ludicrous. If that is what you are suggesting... > The interoperability problems begin when IDNA gobbledygook is converted > to the local character encoding for display. If the result is copied to > another program---through pipes, copy-and-paste, whatever---then that > program's lookups will fail. Mail will bounce, web links will fail, etc. That's not actually completely true. Badly written software will break. In particular throwing up chars to the shell might break things, but that's particularly specific to a platform. Pipes won't break. Copy and Paste may depending on OS, and 'from where and to where'. Generally, it won't. If it does, it's because somebody has written some *really* crusty code that can't handle ASCII outside 0-127. The IETF shouldn't be in the business of making sure everythign works with badly written code. > The conversions I'm talking about are specified in the IDNA documents. > They will produce interoperability problems visible on the wire. You are > incorrect in saying that this is outside the IETF's scope. You are kind of right, but your premise is all over the shop. The example you've given is outside the scope of IETF (IMHO), but the broader issues obviously are worth a look at. What I'm having problems understanding, and I suspect others are feeling the same, is what exactly you are proposing should be done about it apart from telling the WG to drop everything and listen to you? -- Paul Robinson