RE: [iSCSI:] Logout request -- reason? Correction

[Black_David@emc.com]

Luben,

> In other words you have no problem with those changes
> taking place, you simply oppose to them, lest the
> PDU format become more object oriented friendly.

I think that's somewhat overdone and out of line.  The formats
have been this way for a while now, and we are well past the
point where this sort of change is appropriate for abstract
beautification reasons such as "object oriented friendly".
We could spend the next 6 months rearranging the iSCSI header
structures to suit some ideals of beauty and architectural
cleanliness, and even though there are doubtless some
people who would like to "help" with this, I'm fairly certain
that the rough consensus of the WG is not to spend the next
6 months on this activity.

> You should've at least recognised that this would
> be a _good_ architectural change, and then give your reasons
> for _not_ implementing it, rather than calling them
> ``gratuitous changes''.

There's quite a bit of "running code" out there that deserves
some degree of respect - the fact that these changes break it
suggests that something other than an opinion that the result
will be more "object oriented friendly" is needed to motivate
them.  At the moment, both "rough consensus" and "running code"
appear to suggest that they are not appropriate.

Thanks,
--David
---------------------------------------------------
David L. Black, Senior Technologist
EMC Corporation, 42 South St., Hopkinton, MA  01748
+1 (508) 249-6449 *NEW*      FAX: +1 (508) 497-8500
black_david@emc.com         Cell: +1 (978) 394-7754
---------------------------------------------------

> -----Original Message-----
> From: Luben Tuikov [mailto:luben@splentec.com]
> Sent: Tuesday, May 28, 2002 11:53 AM
> To: Paul Koning
> Cc: ips@ece.cmu.edu; John Hufferd
> Subject: Re: [iSCSI:] Logout request -- reason? Correction
> 
> 
> John Hufferd wrote:
> > 
> > There is nothing broken here, this is preference.  I see no 
> reason to move
> > it again.  Julian, did announce it and we accepted it.
> > We should be moving towards closure.
> > 
> 
> Paul Koning wrote:
> > 
> > I objected before and I'll object again.
> > 
> > It would be very nice if we could stop making gratuitous changes to
> > packet formats and get the spec finished instead.
> 
> In other words you have no problem with those changes
> taking place, you simply oppose to them, lest the
> PDU format become more object oriented friendly.
> 
> The people in C wouldn't mind that much since they are used to
> dealing with bit shifting and what not, but the people in
> C++ where this inconsistency would mean an ugly base class,
> and an unbecoming implementation would cringe at the look of it.
> 
> You should've at least recognised that this would
> be a _good_ architectural change, and then give your reasons
> for _not_ implementing it, rather than calling them
> ``gratuitous changes''.
> 
> I'm not posting here out of whim -- I really do care.
> 
> -- 
> Luben