Re: (random) musings on I/O disciplines

At 04:49 PM 9/27/00 -0400, Andy Dougherty wrote:
>On Wed, 27 Sep 2000, Dan Sugalski wrote:
>
> > Perl will, presumably, know how to go from A to B, and deal with that. If
> > you do something that wants UTF8 and your scalar's got EBCDIC data in it,
> > then perl will do the conversion on the fly.
>
>I'm more worried that perl might mistakenly *think* the scalar's got
>EBCDIC when, in fact, it's just random binary data.  I want to be sure
>there's a way to tell perl "hands off".

My gut feel is that there'll be a "default encoding" for filehandles and we 
*won't* try and get clever. And binmode() will actually do something 
meaningful on Unix and force a filehandle to no translation. (So even if we 
do get clever, you can explicitly turn it off)

Having a default encoding would make it easier to not have to bother 
thinking about things unless you're dealing with non-native files. (Unicode 
on WinXX, EBCDIC on IBM Big Iron, and ASCII pretty much everywhwere else, 
with an option to choose the default at build time for folks that need it)


					Dan

--------------------------------------"it's like this"-------------------
Dan Sugalski                          even samurai
[email protected]                         have teddy bears and even
                                      teddy bears get drunk