On Monday, 11/24/2003 at 01:19 EST, "Post, Mark K" <[EMAIL PROTECTED]>
wrote:

> I understand that they are synonyms _now_.  The problem is, they didn't
> start out that way, and the lack of appropriate documentation to that
effect
> has caused problems in the field.  (Perhaps only to me, but still....)
>
> I'm not saying that code changes need to be made (at this point), just
> documentation changes.  The long-term solution may be to remove the
escon
> device name entirely, but that's something that can happen down the
road.

I understand and I was agreeing with you.  I was just taking issue with
your implication that the ctc0/escon3 example from the book is wrong.  The
example is fine.  It's just grossly misleading.  Removing "escon" from the
configuration documentation (not the code) would help in the short term,
IMO.

Footnote #7 on p.96 states "When using the channel device layer only, all
CTC network devices are named ctc n , regardless of whether they are
virtually defined or a real ESCON" but I find that a bit *too* obscure.

In general, I would have to say that the relationship of device names and
channel numbers (the 'n' suffix), and their relationship to *interface*
names is something that could be better documented.

FWIW, there is a Reader's Comment Form in the back of the book (but no
e-mail address, maybe use [EMAIL PROTECTED]).

Alan Altmark
Sr. Software Engineer
IBM z/VM Development

Reply via email to