Re: 504

[Pat Barber <mboceanside@xxxxxxxxxxxxxxxx>]

I made the comment about "auto detect" but that routine will only
work with sync devices(typically). It will not "detect" a cable as
such, since it is NOT a device but a channel/path to a device. Error
detection code is probably very scarce on "Ops Console" since it is
a async device and they(system programmers) never even considered all
the bad things that can happen if that connection is broken.

Error correction and error determination is a fairly complex piece of
code, but it appears nobody went back when "Ops Console" came along
and redid that part.(part of the "do away with twinax conspiracy")

Just make sure your system value QSCPFCONS is set correctly...



Leif Svalgaard wrote:
>
> From: Steve Richter <srichter@AutoCoder.com>
> > > the line should the cable be disconnected. Simply assume that it is
> always
> > > there and that it always works (and as now: hang with a Reference code
> > > if not found during IPL). So, I see no excuses. It is still terrible.
> > >
> >
> > The console connection has to work even if os400 is not loaded on the
> > system.  So the protocol code must make a minimum of system calls.
> > Probably no system calls.  If that is the case, adding any requirement
> other
> > then "connect at ipl" probably does more harm in the form of possible bugs
> > and problems then it is worth.
> >
>
> still terrible.
> Someone mentioned that you can plug in a new cable and the system
> will "autodetect" the new cable. Why not autodetect the console cable?
> This will not complicate the IPL protocol. No valid excuses as far as
> I can see. But the thread has run its course and I learned something
> and it is not good.