[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Finishing the XML-tagging discussion

> > One alternative - not necessarily one I am advocating, but just listing for
> > completeness - is to use a similar approach but swapped round:
> >
> > image/svg;ContentLanguage=xml

> I favor something like this.  it seems like it will interact more
> cleanly with content negotiation mechanisms (as opposed to those
> mechanisms having to acquire new pattern-matching facilities)

> it's true that as MIME is architected the parameter space is on a
> per-content-type basis.  but I don't see any great harm if we define
> some global parameter name space (say, parameter names that begin
> with "$" are global).  Then we could do something like:

> content-type: image/svg; $superclass="application/xml"

First of all, this approach was discussed when MIME was first designed and was
roundly rejected at that time. At the time I was actually in favor of doing
this, but I've since decided that the people who were opposed to it were right.

Second, the definition of a global parameter namespace is far from the worst
thing about this. The biggest problem this approach has is that now you have
what amounts to a mandatory parameter, one which has to appear with every
appearance of a given media type and must have a specific value in those
cases. This is completely contrary to the entire intent of the parameter
space: Parameters were intended to convey information that isn't invariant
with the media type.

Third, this introduces the possibility of silly states in a major way.

Fourth, the infrastructure upgrades to handle this are far nastier
than you seem to think, and have to happen at both the sending and receiving

Fifth, as I said before, this doesn't work with places where only media
types and not parameters are carried. Such places do exist and fixing
them all is going to be nearly impossible.

In summary, what this amounts to is nothing less that a complete redesign of
how  MIME works. And in case my position on this wasn't clear before, this is a
100% ABSOLUTE TOTAL SHOWSTOPPER for me. I will fight this in every way i
possibly can -- I really do think it is that bad of an idea. Frankly, I haven't
seen anything I believe is so destructive and confusing proposed since the use
of line or character counts were considered as an alternative to
boundary markers back in 1991.