[Ex-plain] ZiNG etc (SRW)

Mike Taylor mike at tecc.co.uk
Thu Apr 25 18:03:04 CEST 2002

> Date: Wed, 24 Apr 2002 09:08:25 -0400
> From: "LeVan,Ralph" <levan at oclc.org>
> > So the question for me now (and I admit this is WAY off topic for the
> > ex-plain list and probably the for ZNG list too) is this: why did you
> > and many other consider SOAP ("toolkits will probably be available
> > soon") more attractive to the implementor than Z39.50 ("many real,
> > live, debugged, combat-tested toolkits available right this instant")?
> Here's why I want to abandon the current infrastructure:
> <snip>

Thanks for these thought-provoking reasons.  It's useful to see them
nailed down, and something of a relief to find that the SOAP bubble
(har!) is not _just_ a matter of "Everyone else is doing it so why
don't we?"

Just one point --

> First, writing robust servers is hard.  We've been in the business
> for 10 years now and we still haven't got it perfect.  About half of
> our serious support problems end up being caused because funny
> things happen inside sockets; especially when machines are under
> loads.  There are lots of robust http servers; I'd prefer an
> architecture that lets them take care of my communications problems.

Have you used Yaz's back-end server infrastructure?  I've never had
the slightest difficulty with it.  It runs standalone, under inetd, as
a single multiplexing process, fork()ing for each connection, spawning
threads, really however you want it.  Not run into any resource leaks,
socket funnies or anything else of that ilk.

Moral: there are good Z39.50 server tools!  (And no, before it occurs
to anyone, Sebastian didn't put me up to saying this ... I can imagine
him blushing as he reads this :-)

 _/|_	 _______________________________________________________________
/o ) \/  Mike Taylor   <mike at miketaylor.org.uk>   www.miketaylor.org.uk
)_v__/\  "There are many elements to story-telling: style, characters,
	 plot, ideas, etc.  Too many stories try to get by with only
	 one of these" -- Rheal Nadeau.

More information about the Ex-plain mailing list