[
Lists Home |
Date Index |
Thread Index
]
At 01:50 PM 6/12/2002 -0400, Mike Champion wrote:
>I honestly think that this underestimates the fine work that Paul Prescod
>and others have done to evangelize REST, the importance of specifying
>services as URIs, integrating with the rest of the Web, etc. The actual
>humans in the W3C working groups have generally come quite a long way in
>the last two months. What might look like a wishy-washy compromise
>really does level the playing field, and gives vendors a transition
>plan from "RPC everywhere" to "REST whereever it really works better."
If the envelopes weren't glued to the messages and intermixed with them
throughout, I might have more patience for it.
> >We've tried this at the Web Standards Project. My continual frustration at
> >the uselessness of such tactics and the W3C's refusal to "demand" anything
> >led me to wander away to places where I could more bluntly state my
> >frustrations.
>
>Fair enough ... but so what if the W3C did "demand" something? How much more
>clearly could it "demand" valid HTML 4 and/or XHTML than it does? How much
>good has that demand done? Suppose it drummed vendors who support legacy
>HTML fixups (or mindless SOAP-RPC) out of the organization?
>(That is almost literally unthinkable, since the W3C is after all
>a vendor consortium ....). What, realistically, could it do?
It could do what it's done quite successfully with XML syntax - require
error reporting and halts to processing when programs encounter information
that just ain't right. While it still requires outsiders to value those
restrictions and shame companies into supporting them, that seems to have
worked well. (This notion received mixed responses on www-tag, but I have
a tiny bit of hope for it.)
For SOAP-RPC, a similar answer might be to drop the section of the document
that blesses them. I can imagine how popular that would be (not).
> >it's worth laying out the trenches
> >and settling in for a very long guerilla battle to come.
>
>I sympathize ... but who is the "enemy" ?
"Web Services" in their current form seems an adequate description.
> Anyway, victory will come from
>SHOWing people how much cheaper and cleaner and better a RESTful application
>of XML (with or without SOAP) can be, not getting some authority to
>demand RESTfulness.
I wish it was that simple. SHOWing people examples isn't exactly a winning
hand when a small army of vendors large and small has spent a year crowing
that their magical toolkits will erase interoperability problems.
At this point, I think it makes sense to keep talking about alternative
approaches to SOAP, and wait for the fallacy of the Web Services claims to
come home. If everything SOAPy turns out clean, it's just a waste of
energy. If not, then we'll be better prepared for the next round of fixing
things.
In the meantime, I guess we're all stuck talking about SOAP on a regular
basis as part of our daily XML lives. Between that and W3C XML Schema...
yeesh.
Simon St.Laurent
"Every day in every way I'm getting better and better." - Emile Coue
|
