*BSD News Article 12787


Return to BSD News archive

Xref: sserve comp.org.usenix:3297 comp.unix.bsd:11674 comp.org.sug:678 comp.os.386bsd.misc:102
Newsgroups: comp.org.usenix,comp.unix.bsd,comp.org.sug,comp.os.386bsd.misc
Path: sserve!newshost.anu.edu.au!munnari.oz.au!spool.mu.edu!agate!ames!sgi!rhyolite!vjs
From: vjs@rhyolite.wpd.sgi.com (Vernon Schryver)
Subject: Re: POSIX Printing
Message-ID: <gl9j1jq@rhyolite.wpd.sgi.com>
Organization: Silicon Graphics, Inc.  Mountain View, CA
References: <C3tqy6.8Jv@sugar.neosoft.com> <C3ut15.KrE@ra.nrl.navy.mil> <C3wJ2r.Aup@ra.nrl.navy.mil>
Date: Mon, 15 Mar 1993 00:51:17 GMT
Lines: 50

In article <C3wJ2r.Aup@ra.nrl.navy.mil>, atkinson@itd.nrl.navy.mil (Randall Atkinson) writes:
> In article <C3w5M3.8sG@sugar.neosoft.com> peter@NeoSoft.com (Peter da Silva) writes:
> 
> >I don't want to see changes in the LPD protocol. I want to see a new protocol
> >that corrects LPD's historical dregs. I don't think LPD is fixable in any
> >downward-compatible manner.
> 
> So write one and then suggest it to the IETF.  
> 
>   No need to become a standards droid or go through political
> gyrations to get the IETF to consider a proposal for a printing
> protocol.  Replacing the lpd protocol should be fairly easily if you
> have specific thoughts on how to design such a networked printing
> protocol and your proposal is technically better.  The other nice
> thing about doing this via the IETF is that folks really will
> implement it if the IETF standardises it.


That is backwards, at least for now.  And as long as IETF standards are
not as bad as the rubish from ANSI, IEEE, CCITT, and the rest.

The IETF claims it will standardize it if there are at least 2
independent implementations and if it doesn't look like a completely
stupid idea.  That last is where the inevitiable old-boys network and
politics comes in.  The "implement first and then standardize" rule is
the check against political nonsense that is rampant in the "real"
standards committees.

The proper steps are:

    1. write a note describing the protocol, and publish it as an
	"internet draft".

    2. if you need special ports or other assigned numbers such as
	multicast addresses, cast the description as an RFC so that Jon
	Postel &co. will assign the numbers you need.

    3. write an implementation, and give away the source.

    4. convince someone else to write an implementation

    5. take it to the IETF.


You don't really have to give away the source in step (3), but if you
don't, you are unlikely to interest enough people to help it become an
official standard.


Vernon Schryver,  vjs@sgi.com