Re: [tlug] The 11th Commandment

[Curt Sampson <cjs@example.com>]

On 2008-04-09 17:10 +0900 (Wed), Josh Glover wrote:

> But part of grokking Unicode is knowing to store your data in UTF-8
> internally, and encode to some other crazy charset as late as possible
> in the output process. Which we do. :)

Right. In other words, wait until as late as possible before destroying
the data.

On 2008-04-10 02:00 +0900 (Thu), Edward Middleton wrote:

> For all [of Unicode's] goodness it isn't supported as universally at
> the client end as the local encodings, in any given language. Which
> means you end up being forced to choose between good standards or good
> client support.

Not really. You give Unicode to those that know it, and, for Japan,
Shift_JIS for those that don't.

You'd think that if Starling's web framework could do this, Amazon's
could. (Yeah, yeah, I know, that's not nice.)

> ...by converting to their charset right before you output it.
> This approach also confers the bonus of allowing you to disregard
> completely the details of the target encoding; you can just consider
> it a black box and be done with it.

Not really, no. You still have to deal with code points that can't be
displayed in the target charset and encoding, though just changing them
to '?' is probably not an entirely unreasonable option.

cjs
-- 
Curt Sampson       <cjs@example.com>        +81 90 7737 2974   
Mobile sites and software consulting: http://www.starling-software.com