SPAM: Re: [OPEN-ILS-DEV] Full OpenSRF over HTTP

[Mike Rylander]

On 10/18/07, Ben Ostrowsky <benostrowsky at gmail.com> wrote:
> On 10/18/07, Mike Rylander <mrylander at gmail.com> wrote:
> > (to be a pedant about precision) clients aren't mobile,
> > nor forced or guaranteed to migrate (implied by "migratory") ... I
> > guess I'd like something softer -- "client might move, but that's OK."
>
> Would "portable" express that sense?  I use Firefox Portable on my
> desktop simply because it's easier to have multiple single-purpose
> installations of a portable web browser.
>

I don't think so, unless I'm misunderstanding what you mean.  Since
it's not about the client it's a little harder to explain and grok, I
think.

The point of view we're considering here is that of the back-end
server process.  Right now, it is an error for more than one client ID
to talk to a specific server process within a stateful session.  The
server will report to the unrecognized client that the session it is
trying to get involved with is broken (as far as that client is
concerned) and refuse to honor that client's requests.  It's an
anti-hijacking mechanism that the server uses to protect itself and
the data that will be going to the client.

This new setting would tell the server process that it should ignore
that error condition, and use the new client ID as the endpoint at
which it should spew results.

I'm still not sure what to call it, but I don't think portable hit's
the nail on the head.  However, thanks, and keep the suggestions
coming!

-- 
Mike Rylander
 | VP, Research and Design
 | Equinox Software, Inc. / The Evergreen Experts
 | phone:  1-877-OPEN-ILS (673-6457)
 | email:  miker at esilibrary.com
 | web:  http://www.esilibrary.com