Re: the latest draft, LPR-LPD

[leo j mclaughlin iii <[e-mail]>]

>
>The reason for the new send* commands was to allow interoperability
>with old agents.  With 1179, there was no way for an agent to
>communicate with a daemon and determine whether it was safe to use the
>"extensions" specified in 1179....
>

Well...if you have an only new client and an only old server the failure
mode in both 1179 and the new RFC is the same -- you can't print.

The advantage of the draft mechanism is that unlike 1179 clients, the
new specification allows for draft clients to dynamically adapt to old
or new servers.

>
>1179 is completely broken, as far as interoperability goes.  The new
>RFC does not allow the 0-length encoding or any other changes to the
>semantics of the existing commands.
>

Then the new RFC is even more broken as far as interoperability goes.  If
the number of email messages I have recieved asking questions about 1179
is any indication, a lot of folk out there have implemented 1179 compliant
lpr clients.  For the new RFC to break these existing implementations
is a rather large departure from normal protocol development.

enjoy,
leo j mclaughlin iii
[e-mail]