Path: utzoo!attcan!uunet!samsung!usc!snorkelwacker!bloom-beacon!EXPO.LCS.MIT.EDU!rws
From: rws@EXPO.LCS.MIT.EDU (Bob Scheifler)
Newsgroups: comp.windows.x
Subject: Re: xdm, x-terminals, and SELECTING hosts
Message-ID: <9009221448.AA26379@expo.lcs.mit.edu>
Date: 22 Sep 90 14:48:27 GMT
References: <1990Sep21.201834.17389@csvax.cs.ukans.edu>
Sender: root@athena.mit.edu (Wizard A. Root)
Organization: The Internet
Lines: 22


    The obvious solution is xdm(1), but I am having a hard time seeing how the
    XDMCP supports my needs.  What I want to do is run a sort of meta-xdm that
    puts a screen on the xterminal advertising a collection of hosts
    that would be happy to manage the display, then let the user select
    the desired host.  At that point the standard indirect protocol would
    work fine, except that only one host would be in the list of acceptable
    hosts.  Unfortunately, I can't see any way to do this within the protocol.

It is possible to do within the current XDMCP definition, although it
isn't perfect.  If you put the "chooser" in the X terminal, then it's
pretty easy: the X terminal does an indirect or broadcast query, gets
a bunch of responses, displays them, and the user selects one.  That
was pretty much the model we had in mind when XDMCP was designed.
Now, if you want to put the chooser in xdm (which in retrospect is
really a pretty good place to put it), then there is a bit of a problem,
as there is no direct way for xdm to do a "proxy" query on behalf of
the terminal.  This can be hacked around by using a two-stage session,
with some state retained in xdm.  Extending XDMCP with a proxy mechanism
would be another solution.

This is an area that we are actively working on for R5.