> how about to add this ? > > "or most preferrably, it MAY use both forms displayed side-by-side".
While we might not want to go into what is preserved (since different folks probably have different preferences), it wouldn't hurt to point out that both forms could be used i.e. that the MAYs are not mutually exclusive. Note thatIn the case of an actual display it is likely to be possible to use both, but there are cases of use (such as the output from a textual application running under xterm) where the output format might be constrained to only one field which contains the address. But I don't think the document needs to go into that detail. > BTW, if IDN display issue deserves this specification, why doesn't IDN > copy&paste recommendation have its place in the document ? > I think copy&paste is a kind of higher-layer communication protocol between > processes which may reside on the same machine or on the two different > machines. That requires proper negotiations between applications/OSes to > work correctly. I think this would be a slippery slope. If we think copy&paste needs to be specified in the standard, then the case can be made the other implementation specific inter-application communication mechanisms (such as Unix pipes etc) warrant the same level of specification. Having said that, I wouldn't mind seeing a well-informed and well-written informational document which discusses implementation issues around copy&paste if such a document provides useful information to the community. Erik
