<html>
I think it is important that we address the VRVS claims up
front. Address the encryption stuff. AES is not
breakable the way they claim. I think the point needs to be made on
the quality and usability differences of the audio and video. I'd
like to see this addressed very seriously. VRVS and AG are very
different in their goals. We should not try to argue that VRVS is
bad, but just that desktop configurations have limitations (severe
limitations in my mind for group oriented activities). Get
the technical details right. In these kinds of reports is better to
go overboard in spelling out what they other guys have done right and
then focus on what is different. Be articulate and a bit
verbose.<br><br>
At 06:34 AM 9/17/2002 -0500, Ivan R. Judson wrote:<br>
<blockquote type=cite class=cite cite> <br>
<font face="arial" size=2 color="#0000FF">Here's the feedback. I'll be
working on this in the background, but input is encouraged. Clearly
there is a disconnect and we have to get our story straight so we can
answer these questions. The nuts and bolts of the issues have to support
our claims, or we're wasting our and Tony's time.</font><br>
<br>
<font face="arial" size=2 color="#0000FF">--Ivan</font><br>
<br>
<br>
<dl><font face="tahoma" size=2>
<dd>-----Original Message-----
<dd>From:</b> Michael Daw
[<a href="mailto:mike.daw@man.ac.uk" eudora="autourl">mailto:mike.daw@man.ac.uk</a>]
<dd>Sent:</b> Tuesday, September 17, 2002 3:24 AM
<dd>To:</b> Sue Rogers; Stewart Macneill; Stephen Booth; Richard Ansorge;
Paul Jeffreys; Mike Giles; Mark Hayes; Mark Cavanagh; Malcolm Atkinson;
Liz Carver; Kate Caldwell; John Gordon; John Brooke; Jeremy Sharp; Ivan
R. Judson; Henry Hughes; David Salmon; David Hutchison; David De Roure;
David Boyd; Chris Osland; Brian Gilmore; Alan J. Flavell; Peter Clarke
<dd>Cc:</b> galvez@hep.caltech.edu
<dd>Subject:</b> The return of the report<br><br>
</font><font face="arial" size=2 color="#0000FF">
<dd>Dear team,</font>
<dd> <font face="arial" size=2 color="#0000FF">
<dd>It's not over yet!</font>
<dd> <font face="arial" size=2 color="#0000FF">
<dd>Below is Tony's second e-mail on the subject of the report forwarding
a response from the VRVS team at Caltech. (Just so you know, his first
e-mail said "I liked your report very much and hope we can take
something forward from your recommendations".) </font>
<dd> <font face="arial" size=2 color="#0000FF">
<dd>So - Tony wants the report revised in light of these comments. While
we're at it, we should also incorporate some concerns expressed by a
couple of other people. Then the report can be disseminated widely, as
Tony says.</font>
<dd> <font face="arial" size=2 color="#0000FF">
<dd>What I suggest is that those of you who are interested in helping
with the re-draft let me know. I'll then co-ordinate some kind of meeting
to discuss how to take this forward. I'd like to keep the current
contributors on this distribution list so that they can comment on
revisions and remain on the contributor list. I've cc-ed Phillippe too so
that he - or someone at Caltech - can be a full member of the re-draft
team.</font>
<dd> <font face="arial" size=2 color="#0000FF">
<dd>Please tell me if you can think of better ways forward, or have
questions or comments on anything related (or unrelated, for that
matter).</font>
<dd> <font face="arial" size=2 color="#0000FF">
<dd>--Mike</font>
<dd>
<dd> <font face="tahoma" size=2>
<dd> -----Original Message-----
<dd>From:</b> Tony Hey
[<a href="mailto:Tony.Hey@epsrc.ac.uk" eudora="autourl">mailto:Tony.Hey@epsrc.ac.uk</a>]
<dd>Sent:</b> 16 September 2002 13:44
<dd>To:</b> 'mike.daw@man.ac.uk'
<dd>Cc:</b> Tony Hey; Ray Browne (Ray.Browne@dti.gsi.gov.uk); James
Fleming; Anne Trefethen
<dd>Subject:</b> FW: FW: Background Document for JCSR<br><br>
</font>
<dd>Mike <br><br>
<font size=2>
<dd>Thanks for your note - I am forwarding some comments from the VRVS
team at Caltech which seem to me to have some merit! And this answers my
strategy as to publication - I would like your team to consider these
comments and produce any revisions (if necessary - but I will require an
answer from you to their points!) to your paper. Then we would like it
circulated widely. There is also every possibility that JCSR will be
willing to fund some of your recommendations ...</font><font size=2>
<dd>Best wishes</font> <br><br>
<font size=2>
<dd>Tony</font> <br><br>
<font size=2>
<dd>Professor Tony Hey</font> <font size=2>
<dd>Director e-Science Core Programme</font> <font size=2>
<dd>Engineering and Physical Sciences Research Council (EPSRC)</font>
<font size=2>
<dd>Polaris House</font> <font size=2>
<dd>North Star Avenue</font> <font size=2>
<dd>SWINDON</font> <font size=2>
<dd>Wiltshire SN2 1ET</font> <font size=2>
<dd>ENGLAND</font> <font size=2>
<dd>Tel: +44(0)1793 444022 Fax: +44(0)1793 444505</font> <font size=2>
<dd><a href="http://www.epsrc.ac.uk">http://www.epsrc.ac.uk</a> </font><font size=2>
<dd><a href="http://www.research-councils.ac.uk/%A0" eudora="autourl">http://www.research-councils.ac.uk/ </a> </font><font size=2>
<dd><a href="http://www.escience-grid.org.uk">http://www.escience-grid.org.uk</a> </font><font size=2>
<dd><a href="http://umbriel.dcs.gla.ac.uk/NeSC/general/%A0%A0%A0" eudora="autourl">http://umbriel.dcs.gla.ac.uk/NeSC/general/ </a> </font><font size=2>
<dd>-----Original Message-----</font> <font size=2>
<dd>From: Philippe Galvez [<a href="mailto:galvez@hep.caltech.edu">mailto:galvez@hep.caltech.edu</a>] </font><font size=2>
<dd>Sent: 15 September 2002 21:48</font> <font size=2>
<dd>To: Harvey Newman</font> <font size=2>
<dd>Cc: Tony Hey; Philippe Galvez; Alan J. Flavell</font> <font size=2>
<dd>Subject: Re: FW: Background Document for JCSR</font> <br><br>
<font size=2>
<dd>Dear Tony,</font> <br><br>
<font size=2>
<dd>Harvey and I went over the document and made additional comments. I hope</font> <font size=2>
<dd>you will find it useful and constructive. feel to contact us for more</font> <font size=2>
<dd>detail. I will be glad to discuss any topics that you will find</font> <font size=2>
<dd>appropriated.</font> <br><br>
<font size=2>
<dd> Regards,</font> <font size=2>
<dd> Philipps</font> <br><br>
<font size=2>
<dd>cc: Harvey Newman, Alan Flavell</font> <br><br>
<font size=2>
<dd>------------------------------------</font> <br><br>
<font size=2>
<dd> o My first comment concerns the list of contributors. I see on the</font> <font size=2>
<dd>list Ivan Judson (from ANL/AG, AG's architect) and Jim Miller from</font> <font size=2>
<dd>InSORS. In this regard, I am a bit surprised that the people preparing</font> <font size=2>
<dd>the report didn't contact us directly if they wanted a balanced</font> <font size=2>
<dd>evaluation, while avoiding technical errors and misconceptions.</font> <font size=2>
<dd>Hopefully, Alan Flavell contributed to the VRVS parts as he is one</font> <font size=2>
<dd>person who understands for the most part what VRVS is.</font> <br><br>
<font size=2>
<dd> o The executive Summary is good.</font> <br><br>
<font size=2>
<dd> Comments by Section follow:</font> <br><br>
<font size=2>
<dd> o 2.1: Technically speaking, I don't understand the planned usage of</font> <font size=2>
<dd>openMCU to enable interoperability between H.323 and AG. First of all,</font> <font size=2>
<dd>VRVS has done it -- completely. Second, it is definitely not the right</font> <font size=2>
<dd>technical solution. I already informed the AG guys about it but looks</font> <font size=2>
<dd>like they want to insist on using it without first understanding the</font> <font size=2>
<dd>openMCU architecture and its limitations..</font> <br><br>
<font size=2>
<dd> o 2.4: The interoperability between AG and VRVS is already done. The</font> <font size=2>
<dd>further deployment of AG/VRVS in UK and other countries could be done</font> <font size=2>
<dd>easily, as well as full support for more AG virtual venues.</font> <font size=2>
<dd>Something that people didn't get is that the current AG/VRVS Gateway</font> <font size=2>
<dd>has never been in the "Funded Tasks development list". We simply did it:</font> <font size=2>
<dd>because there was a need, and because our architecture was flexible</font> <font size=2>
<dd>enough so that we were able to do it with minimal resources. This is</font> <font size=2>
<dd>just one of several examples where people do not quite realize the</font> <font size=2>
<dd>potential of our realtime software infrastructure, and conversely assume</font> <font size=2>
<dd>limitations in the VRVS system that don't exist.</font> <font size=2>
<dd>In this case if people think that VRVS/AG gateway improvements are</font> <font size=2>
<dd>needed right now, then we could do it in return for a moderate amount of</font> <font size=2>
<dd>funding -- to improve the gateway GUI, to deploy it, and to provide a</font> <font size=2>
<dd>bit of support. The funding is only needed because we would have to push</font> <font size=2>
<dd>other tasks on our long in-progress development list; or (better) we</font> <font size=2>
<dd>would take on another engineer to keep all of our milestones (including</font> <font size=2>
<dd>the VRVS/AG ones) on schedule.</font> <br><br>
<font size=2>
<dd> o 2.4: I don't understand "It is not possible for an Access Grid to</font> <font size=2>
<dd>participate in a VRVS session". This is not true. And the solutions</font> <font size=2>
<dd>proposed do not take in account the current technique used in the</font> <font size=2>
<dd>gateway.</font> <font size=2>
<dd>It would be better to ask us is there is a presumed problem: (a) to</font> <font size=2>
<dd>find out if it is already done, (b) to allow us to propose solutions,</font> <font size=2>
<dd>which in the great majority of cases will be quite straightforward for</font> <font size=2>
<dd>us.</font> <br><br>
<font size=2>
<dd> o 3.4, H.323/H.320: These are Not broadcast quality as mentioned</font> <font size=2>
<dd>here. CIF (352 X 288) is half the resolution of broadcasts. In</font> <font size=2>
<dd>continuous presence (4 videos on a screen) one has QCIF for each video</font> <font size=2>
<dd>which is 4 times less resolution per video than broadcast.</font> <br><br>
<font size=2>
<dd> o 3.7 AG: A solution purely multicast based is surely not "the only</font> <font size=2>
<dd>fully scalable solution for large-scale collaboration" as it is</font> <font size=2>
<dd>mentioned in the report. It has been demonstrated since the early</font> <font size=2>
<dd>1990's that the opposite is true. In contrast, it has been recognized</font> <font size=2>
<dd>(e.g. at high performance network meetings organized by DOE this Summer)</font> <font size=2>
<dd>that using unicast tunnels where needed to interconnect mulitcast</font> <font size=2>
<dd>domains is demonstrably scalable. This is why VRVS has spread to 60+</font> <font size=2>
<dd>countries, and precisely why VRVS is architected the way it is.</font> <font size=2>
<dd> The claim that pure multicast is scalable in this section is in</font> <font size=2>
<dd>contradiction with the following sentence, that mentions that it will</font> <font size=2>
<dd>take few years for having all UK academic LANs with multicast</font> <font size=2>
<dd>capability. We must point out that multicast at network level, with</font> <font size=2>
<dd>large TTL is now considered dead by leading experts such as Linda</font> <font size=2>
<dd>Winkler and Matt Matthis. VRVS's unicast tunnels inter-connecting</font> <font size=2>
<dd>multicast domains has been recognized as the "obvious" globally</font> <font size=2>
<dd>scalable solution; in contrast to what is said in this report.</font> <br><br>
<font size=2>
<dd> o 3.7 H323: Here it is not mentioned that H.323 devices are very</font> <font size=2>
<dd>sensitive to packet loss. The adaptive/predictive protocols used by</font> <font size=2>
<dd>H.323, while bringing higher interactivity and quality per unit</font> <font size=2>
<dd>bandwidth, also are more sensitive to persistent "artifacts" on the</font> <font size=2>
<dd>screen. Especially if packet loss is above a few tenths of one percent.</font> <br><br>
<font size=2>
<dd> o 3.9 AG: An ACL list do not provide a secure way to operate.</font> <font size=2>
<dd>Multicast is definitely not the best technique for providing security</font> <font size=2>
<dd>since almost anyone on the network can sniff the packets. In this</font> <font size=2>
<dd>particular case, the only solution is encryption from the applications</font> <font size=2>
<dd>and today this could be done via the Mbone tools (Vic/Rat). This is not</font> <font size=2>
<dd>particular to the AG, but only related to the Mbone tools. In addition,</font> <font size=2>
<dd>the encryption provided in the Mbone tools is very basic, and anyone</font> <font size=2>
<dd>enough competence will be able to decrypt it.</font> <br><br>
<font size=2>
<dd> o 3.13: I don't understand what is meant by "The AG configuration</font> <font size=2>
<dd>applied at the e-Sciences Centers makes them unsuitable for VRVS</font> <font size=2>
<dd>conferences". Are the AG nodes there different from all the other AG</font> <font size=2>
<dd>nodes elsewhere; if so, in which ways ?</font> <br><br>
<font size=2>
<dd> o 5.2: I am surprised by the cost for the AG. Platinum 30,000 LS</font> <font size=2>
<dd>(UK), One could equip a whole site with desktop cameras !.</font> <br><br>
<font size=2>
<dd> o 8.1: Again, we see an incorrect comparison of VRVS (a realtime</font> <font size=2>
<dd>software infrastructure integrated with many clients) with the</font> <font size=2>
<dd>application tools themselves: Netmeeting, QT, CUSeeMe,...etc.. Once</font> <font size=2>
<dd>again this is a failure to understand that VRVS is not a client</font> <font size=2>
<dd>application, but rather provide the infrastructure upon which many</font> <font size=2>
<dd>client applications can run, and interoperate. It is like comparing</font> <font size=2>
<dd>Cisco routers and the Cisco IOS software, with FTP or Telnet.</font> <br><br>
<font size=2>
<dd>-----------------------------------------------------------</font> <br><br>
<br><br>
<br><br>
<dd>**********************************************************************
<dd>Internet communications are not secure and therefore EPSRC
<dd>does not accept legal responsibility for the contents of this
<dd>message. Any views or opinions presented are solely those
<dd>of the author and do not necessarily represent those of EPSRC
<dd>unless otherwise specifically stated.<br><br>
<dd>All EPSRC staff can be contacted using Email addresses
<dd>with the following format: 'firstname'.'lastname'@epsrc.ac.uk
<dd>**********************************************************************
</dl></blockquote></html>