Hi, On Tue, Nov 25, 2025 at 01:40:46AM -0500, Petr Menšík wrote:
Python3 says ZWJ is not printable.
Mmmmm. And yet, if you don't render it when you are dealing with a stream of octets that represent UTF-8 encoded Unicode characters, when the string in question requires a ZWJ, then the join is elided and the resulting string doesn't render correctly.
These definitions are corner cases.
The ZWJ is a corner case to you because your particular writing system doesn't use it. It seems to me elsewhere in this thread you have accused people of not caring about those whose first language isn't US English as the reason why anyone could be opposed to your suggestion to modify, without any kind of versioning control, the semantics of zone file entries. But if you can't even be bothered to understand _why_ the question I asked was one that has some consequences for your position, I submit that it is you who is ignoring the issues related to at least some non-ASCII-using writing systems.
It is up to higher layers to render text. That is job of GUI toolkits, terminal emulators and similar.
Great. So what problem is it you're trying to solve by insisting on an interpretation of the octets that come through from a zone? Surely, whatever else is the case, we can agree that DNS zone files are some distance off from the "higher layers", GUI toolkits, and so on?
I think in native code iswprint() should be used to guess
So now you want to make all of these decisions subject to whatever biases are implicit in some programming language's design?
PRECIS or RFC 9839 is an implementation detail.
Perhaps, but the fact that you need one of these things to specify your subset is _not_ a detail, but a fundamental issue.
utf-8 encoded text as a normal text. If a domain name can contain socks icon, why not a TXT record? Why are no emoticons rendered?
Emojis are not good identifiers because there is no conceptual character behind them. The UTC is _usually_ crystal clear about this.
URI can have normalized form of only A-label input with percent encoded path characters in URI.
The above suggests to me that you don't know what an A-label is, either. Frankly, I don't know why I am bothering to continue to pursue this discussion, since you seem completely immune to the suggestion that perhaps the topic is rather more complicated than you appear to think, and that knowing even a little bit about it before pontificating about the obvious right solution might help make your position a little bit more defensible. It doesn't especially matter to me anyway: I'm not really an active participant in the IETF these days, and I've no reason to suppose that will change. But your proposal is, to be blunt, uninformed codswallop, and I urge you to go learn some things about the topic of writing systems before proposing again something with a bunch of hand-waving about how trivial it is. Best regards, A -- Andrew Sullivan [email protected] _______________________________________________ DNSOP mailing list -- [email protected] To unsubscribe send an email to [email protected]
