[Seaside-dev] Re: what happened to WASqueakUtf8Codec?

[Philippe Marschall]

2009/2/15, Lukas Renggli <renggli at gmail.com>:
>> Oh and something else:
>> TextConverter newForEncoding: 'ISO-8859-1'
>> returns nil. That's why the old code had a special case for that one.
>
> Why would you want to instantiate a codec by name?

Because I know the name of the encoding I want.

> Normally one of the
> instances from the list returned by "WACodex allCodecs" is chosen.
> Of course that can be fixed, but is it worth it? I wanted to avoid
> forcing platforms to understand all variations of codec names like
> 'UTF8', 'UTF-8', 'utf-8', 'utf8', #utf8, #'utf-8', #UTF8, #UTF8', ...

Understandably. It would help if we had a clear policy of what we
support and what we don't support (case, hyphens) and make that
consistent over all platforms

ISO-8859-1 is the preferred MIME name so I think it makes sense to support it:

http://www.iana.org/assignments/character-sets

> but instead just deal with a list of possible names. The new
> implementation does not requires a platform to do encoding/decoding.
> This makes porting much simpler.
>
>> You see there was a reason the old code was like this. It was tweaked
>> and tested over a long period. If you don't understand why something
>> is there please ask. At the very least just test whether it works.
>
> I honestly don't know any project that uses anything but the
> WANullCodec in Squeak. The encoded Squeak streams are so incredibly
> slow, its not believable.

Then we can just remove it. That would simply a lot of code.

Cheers
Philippe