[OU] "Content-Transfer-Encoding" HTTP Request Header "Wild" ???

[Psmith]

	Following is dialog from Chandler Tech help after my failure to
get signed up for Chandler Hub at

		<https://hub.chandlerproject.org/login?signup=true>

------------------------------------------------------------------------
-------------------------------------------------------------

> I realize you have your hands full with getting the program fleshed 
> out, but this is not an Opera problem.
> They closely follow the
> 3W standards.

With respect, following standards does not instantly make everything on
the web interoperate.  Interoperability is hard work, and having built
software systems for many years, it is extremely rare for a standard to
define everything and its context so precisely that nothing can be left
to interpretation.  Those standards that do nail everything down so
precisely are too big and unworkable to ever get correctly implemented.
In addition, W3C standards are full of "may", "might", "should", and
other optional language such that two implementations can both implement
a standard correctly and still not interoperate.

> Most likely it is one of Microsoft's proprietary but illegal 
> instructions intended to advance its browser monopoly.

It is not.  It is the use of the "Content-Transfer-Encoding" HTTP
request header.  This header is defined by specifications (MIME
standards, not HTTP per-se), but it's essentially unheard of in the
wild, so a variety of server implementations, including Cosmo (Chandler
Hub) do not know how to process it.  It's so rare in HTTP that I had to
look it up when I first saw the problem and filed the original bug
tracking the issue:

https://bugzilla.osafoundation.org/show_bug.cgi?id=10996

> When and if they get a chance to fix, I would appreciate it.  In any 
> event, thanks for your response.

We appreciate your comments and hope someday as well for Chandler Server
to work properly with Opera browsers.