Re: several fixes

On 18.04.01 at 06:57, Bjoern Hoehrmann <derhoermi@gmx.net> wrote:

>* Terje Bless wrote:
>>Is Iconv a showstopper for you?
>
>I'm on Windows NT, I have to get libiconv or something similar for the
>iconv() implementation and get that running

Didn't Brian get ActiveState to PPM it? <checks archives> Yup. See
<URL:http://lists.w3.org/Archives/Public/www-validator/2001JanMar/0221.html>
for details.


>I'm still fighting against other packages like XML::Parser 2.30...

I'm willing to bet ActiveState has a PPM of that too. In fact, I'm pretty
sure it's in the standard distribution!


>>Would it be worthwhile to make a simple way to disable Iconv? At the
>>expense of charset transliteration or by falling back to nkf or somesuch?
>
>Just using the (far superior, if you ask me ;-) Unicode::*-modules would
>be the best option here.

It's an option, and what Liam did IIRC, but it requires a bit of work to
get right. Iconv is UNIX specific, but it's a standard. The Unicode modules
are Perl specific. Investigating Unicode::* (and how Liam did it ;D) is on
my TODO.


>>>  * use $q->redirect for redirects
>
>>CGI.pm doesn't allow for a body object (required by RFC 2616). I'll bug
>>lstein about at some point.
>
>Why is just printing it out no option here?

Doh! Ignore the funny little man in the corner; he doesn't know what he's
talking about. CGI::redirect doesn't work how I thought it worked.

Yes, we should use $q->redirect for redirects. Why? Did you hear me say
anything different? You must be on drugs man! :-)


>It's only 'suggested' by RFC 2616, i.e. it's a SHOULD not a MUST. 

In RFCese "SHOULD" means "must, unless there is a compelling reason to
diverge and only if you are _sure_ you know what you are doing". IOW, both
MUST and SHOULD denote requirements; MAY indicates an optional behaviour.

Received on Wednesday, 18 April 2001 02:05:17 UTC