From: "Robert Comer" 

> Your issue isn't with Microsoft though. What alternative do you propose to
> maintain backwards compatibility, not to mention cross browser
compatibilty?
> IIRC this isn't exactly a new thing with HTML.

Wow, me agreeing with Rich, Microsoft, a standards body, and you, all at
the same time, must be a record. 

> George needs to write a more intelligent spam filter. Why not parse
> displayable text out of the HTML and do the normal word thing with that?

I would think that stripping out all comments and then looking for keywords
would be a decent way to do it.  Or have an on the fly comment stripper /
string scanner is probably how I'd do it. But it might be tough depending
on just what's available to him though...

- Bob Comer


"Paul Ranson"  wrote in message
news:3f8e8223$1{at}w3.nls.net...
> Your issue isn't with Microsoft though. What alternative do you propose to
> maintain backwards compatibility, not to mention cross browser
compatibilty?
> IIRC this isn't exactly a new thing with HTML.
>
> George needs to write a more intelligent spam filter. Why not parse
> displayable text out of the HTML and do the normal word thing with that?
>
> Paul
>
> "Adam Flinton"  wrote in message
> news:3f8e5dd6{at}w3.nls.net...
> > Robert Comer wrote:
> >
> > >>Am I the only person who thinks that comments in code are
for humans
not
> > >
> > > for
> > >
> > >>the machine? I mean isn't the whole reason you mark something as a
> comment
> > >>so that the machine knows to just skip it?
> > >
> > >
> > > It's a sneaky way to get in higher level commands without breaking
> backwards
> > > compatibility -- not a bad way to do it at all.
> > >
> >
> > It's not sneaky, it's lame & it's a loathsome way of shoving in
> > processing instructions. Comments should mean that. i.e. for humans.
> > Using them to form part of / create the documentation for a human to
> > read (e.g. javadoc) OK but putting processing instructions in there is
> > simple lazy bodging.
> >
> > Adam
> >
>

--- BBBS/NT v4.01 Flag-5