Have I succumbed?
One of my friends, who for now may remain anonymous, recently sent me to visit Tim Bray's piece on
It is really funny.
And it got me thinking.
I guess when I first started looking at the WS specs I thought there were “a lot of them” too. And now I really don't. So I could see I might have symptoms of that terrible disease that stalks all of us who work with protocols – Protocol Poisoning.
I am one who has seen first-hand the terrifying effects of the disease – the result of breathing too much of your own protocol. The terrible numbness, blindness and deafness it causes – for which in many cases there is no possible cure…
How do you find out when you are slipping into intoxication and ultimately will succumb? You really need some kind of objective measure and protection – a modern day canary if you will.
Basically, as Tim points out, we are talking in the case of the Web Services standards about 783 pages for a set of standards covering distributed systems protocols for “Security, Reliable Messaging, Transactions, Metadata, Messaging, Management, and Business Processes”.
Hey, wait. Actually, does that sound so bad?
The only way to know is to compare it with things that have been done before. How about a successful thing – lke LDAP? After all, it was really simple, right?
So I did a bit of research (see previous post) to find out how many pages are in the LDAP spec collection. The answer –
550 pages! So are we looking at a case where the specs for all of distributed computing are of the same order of magnitude as the “lightweight” directory access protocol?
But Protocol Poisoning is dangerous stuff – we need at least one other objective measure. Maybe we should compare it with DCE? I once ordered the documents and had to move into a larger office. The book on RPC alone was about 700 pages…
So my short term take on this is that the protocol poisoning hasn't gotten to me yet. By the objective measures I could come up with, WS is pretty svelt.