Re: lynx-dev 2.8.4 progress

[Vlad Harchev]

On Mon, 18 Jun 2001, Vlad Harchev wrote:

> On Mon, 18 Jun 2001, Henry Nelson wrote:
> 
>  Hello Henry, 
> 
> > > "feature", and that W3C doesn't permit browsers to do what lynx is doing.
> > 
> > Vlad, did you find some rfc or other reference about that stripping of
> > blank space?  Is that supposed to be left up to the server side or cgi?
> > Interested, but just haven't had the time to look it up.  Thanks for
> > references.
> 
>  HTML 4.0 spec says nothing about stripping trailing white spaces. It just
> says (in interact/forms.html#edef-INPUT):
> ----------
>    The  INPUT  element's  type attribute determines which control will be
>    created.
> 
>    text
>           This  type  creates a single-line text box. The value submitted
>           by a text control is the input text.
> ----------
>  Nothing is said about whether browser should or can "beatify" or "normalize"
> the text that was input. I think it should be treated as "no normalization
> required" or "any modification to text that was input is prohibitted". And as
> I said, all browser I know of (IE, Mozilla, Netscape, Konqueror, Opera, links)
> trim trailing whitespaces, so it seems to be a de-facto standartized.

 Just to add - HTML spec is the document that specifies what part of user's
input to submit exactly. Some RFCs out there related to HTTP specify how to
encode raw bytes, etc, but they don't extend HTML spec in the subject we are
talking about. So HTML spec is the only document that applies here.

 Yes, the trailing white spaces can be trimmed by CGIs for those input
fields for which it's known that trailing white spaces are not meaningful.
Only CGIs know which are those fields, and they have to do trimming if
any other browser but lynx has to be supported. So it's at least safe (but
more correctly, it's more correct) not to trim trailing whitespaces.

 Best regards,
  -Vlad


; To UNSUBSCRIBE: Send "unsubscribe lynx-dev" to address@hidden