Re: Followup on the ccharset idea

[Michael Edenfield <kutulu@kutulu.org>]

* David Oftedal <david@start.no> [031021 13:01]:
> Michael Edenfield wrote:
> >As you said, regardless of what the current character set in X-chat is,
> >domain names need to be converted to IDN format to get resolved, so the
> >two issues would need to be tackled seperately.
> >
> >--Mike 
> 
> Hm, yes. After mentioning the (now) wcharset idea in the channel, the 
> general consensus seemed to be that this would be hard to code, since it 
> would require parsing every message to look for clues as to what window 
> it was headed for. So that probably won't be implemented.

Don't we already do that?  We need to know which window to send data to
at some point.  I'm also working up an increasingly-larger patch to make
clearer distinctions between queryies and DCC chats.  (It fixes the
olders but in the bug tracker -- trying to /msg someone with the same
nickname as a /dcc chat session sends to the wrong place).  One of the
requirements of the patch is to figure out early on what type of window
it's heading for.  I suppose it would depend on how early in the
message's life cycle we need to know what character set to use.

> And with regards to IDN - That's what I'd heard too. It uses some 
> "clever" solution to convert Unicode domain names into something else 
> for backwards compatibility. Sort of like a Unicode emulation layer on 
> top of the extra DNS layer that's already on top of IP. (The word 
> "stupid" comes to mind, if you'll excuse the trolling.) I think two 
> questions arise here... 1) Is it patented? and 2) Will it ever catch on? 

It's possibly patented, I didn't check the exact implementation details.
However, IETF requires a binding agreement with all patent holders
*before* they will approve a patented process as a draft standard.  The
agreement must explicitly give a non-exusive lisence to any "normal" use
of the patent as part of implementing the standard.  There's a
half-dozen of these agreements on file with IETF on their web site,
including ones from NuNames and DNS II.

As for catching on, it's been in use for ~2 years but most people
haven't heard of it.  However, it's apparently a BIG issue in the Far
East, moreso than Europe, where the character sets aren't even remotely
similar to US Latin.  And there's an awful lot of people in China. :)

> But as for the default charset command - I actually hadn't taken Windows 
> into account here. Perhaps it's a good idea after all? It'd be cool if 
> X-Chat could move away from the whole Unix paradigm - I'm not really a 
> big fan of using bizarre commandlines for doing everything either.

As you maybe can tell I'm a *huge* fan of moving away from a
UNIX-centric approach.  In some cases I prefer to move Windows into the
UNIXy realm, or at least give me a choice.  For example, I love the
easier highlight-and-middle-click cut/paste option that XChat brings
over from XWindows.  But in cases like this, where there's just *no*
alternative (see also: /dns and /exec) in Windows that functions like
UNIX, I think it's well worth the effort to accomodate Windows users.
If possible, and if it won't break the working platforms, of course :)

--Mike

PGP signature