Re: interworking issues

pays@faugeres.inria.fr Tue, 01 November 1994 17:44 UTC

Received: from ietf.nri.reston.va.us by IETF.CNRI.Reston.VA.US id aa07134; 1 Nov 94 12:44 EST
Received: from CNRI.Reston.VA.US by IETF.CNRI.Reston.VA.US id aa07130; 1 Nov 94 12:44 EST
Received: from haig.cs.ucl.ac.uk by CNRI.Reston.VA.US id aa21923; 1 Nov 94 12:44 EST
Received: from bells.cs.ucl.ac.uk by haig.cs.ucl.ac.uk with local SMTP id <g.02646-0@haig.cs.ucl.ac.uk>; Tue, 1 Nov 1994 15:24:27 +0000
Via: uk.ac.nsfnet-relay; Tue, 1 Nov 1994 15:24:19 +0000
Received: from faugeres.inria.fr by sun2.nsfnet-relay.ac.uk with Internet SMTP id <sg.11590-0@sun2.nsfnet-relay.ac.uk>; Tue, 1 Nov 1994 15:23:39 +0000
X400-Received: by /PRMD=inria/ADMD=atlas/C=fr/; Relayed; 01 Nov 94 16:23:14+0100
Date: Tue, 01 Nov 1994 16:23:14 +0100
Sender: ietf-archive-request@IETF.CNRI.Reston.VA.US
From: pays@faugeres.inria.fr
To: c.robbins@nexor.co.uk, pays@faugeres.inria.fr
Subject: Re: interworking issues
cc: Paul-Andre.Pays@faugeres.inria.fr, osi-ds@cs.ucl.ac.uk, punters@nameflow.dante.mailnet.fi
Message-ID: <783703394.27940.0-faugeres.inria.fr*@MHS>

> 
>    >	and I will not forget :-) PAradise OIFP (VALUE X.500) which
>    >	enabled to identify and often fix several key problems
>    >	thanks Bruno Koechlin
> 
> I will second that, it was, despite my inital sceptisim, very useful
> for us in identifing, resolving and testing many of the interworking
> issues between our products. 

Thanks a lot Colin.
I will just add that it inderectly also improved a lot the interop
with other products (even not participating to the platform
or tests)

> 
>    >7. some clients have strange behaviours preventing interop
>    >   this may come from the search filter see above
>    >   this also come from non-conformance (or Quipu proprietary features)
>    >   -- this seems to be the case of "de" (at least the publicly
>    >   available DUA at ULCC)
> 
> Could you please elaborate upon this?


I have not all the desirable information to come to conclusion

  often clients tend to include "non searchable" attributes into their
	search filters  -> rejected -> no result
	pb: how to know what will be accepted???? RFC...

  I suspect and in some case nearly proofs that some clients
     algorithm are heavily dependant on the "leaf nonleaf"
     indicators
       - this is clearly something needed and a LACK in the std
       - this is however QUIPU proprietary

  finaly I have observed "strange" behaviours (which must be
  understood as : I have not been able to understand identify
  because of time and because of the debug level of DSAs
  serving many 1000 queries a day)
    if you want more, then use the "de" and observe your own logs
      or ask UCL/ULCC to send the "de"logs... 
   for other clients.... I think it is up to the designers
	to test their products againt several other products
	and they have the french master at hand to check what happens
	  -- but please dont use OPAXdsa as a "crash test system" :-)
    sorry I cann't help more

> 
>    >  eg. there will be no real interworking as long as products
>    >  will at the same time
>    >	set the "DontUseCopy" flag
>    >	while not setting correctly Operationprogress and Nameresolution
>    >   -- this may be acceptable with the QUIPU knowledge model, but
>    >      is not acceptable for any conforming implementations
> 
> I believe this is one of the issues identified in the OIFP testing.  
> As you are aware, we adapted the knowledge model used within our
> version of QUIPU to behave correctly if the remote DSA is not known to
> be a QUIPU system.  The behaviour under 'dontusecopy' was one of the
> last components we worked on before the OIFP funding ran out so may not
> have been fully tested within the platform, but I am confident the
> mechanism now works correctly (as installed on the DANTE central
> machines).

Indeed the interop seems much better, though we have not tested
this entirely.... And I was not attacking your product...
The trouble is that there are still many other products
(including old Quipus) around and not all :-) queries are
chained by Giant Tortois... Moreover we have no information
about which software is sending which query...
> 
> 
> Colin
> 

-- PAP
>