Forums before death by AOL, social media and spammers... "We can't have nice things"
|    comp.protocols.tcp-ip    |    TCP and IP network protocols.    |    14,669 messages    |
[   << oldest   |   < older   |   list   |   newer >   |   newest >>   ]
|    Message 13,070 of 14,669    |
|    Anne & Lynn Wheeler to All    |
|    Re: Follow up    |
|    26 Sep 09 13:17:31    |
      From: lynn@garlic.com              bits & pieces from (20yrs ago):              Transfer vol. 2 no. 5       Protocol Engine Information September/October 1989              In this issue:              XTP Electronic Mail Group       XTP Multicast Group Management       X3S3.3 HSP Meeting       Thoughts on Fast Protocols       Use of the XTP Path Command       TAB Activity       On the Road with XTP1              ... snip ... and ...              XTP Multicast Group Management and Data Transfer        by Alfred C. Weaver, Bert J. Dempsey, W. Timothy Strayer              (editors note: this is the second installment on a five part series on       Multicast)              Group Management              In developing any multicasting facility one must first decide on the       nature of the multicast groups and how they will be created and       modified. In this section we look first at group attributes of       possible interest, then at the functionality of the group management       protocol, and finally at a classification of the ways to implement       group management.              ... snip ... and ...              To Probe Further ....              BIRM87 - Birman, Kenneth, and Josepth, T., "Reliable Communication in       the Presence of Failures", ACM Transactions on Computer Systems,       Volume 5, Number 1, pp 47-76, February 1987              CHER85a - Cheriton, D., and Deering S.E., "Host Groups" a Multicast       Extension for Datagram Interneworks", Proceedings of the Ninth Data       Communications Symposium, IEEE?ACM, Whistler Mountain, BC, Canada, pp.       172-179, Sept. 1985              CHER85b - Cheriton, D. and Zwaenepoel, W., "Distributed Process Groups       in the V Kernel", ACM Transactions on Computer System, Volume 3,       Number 2, pp. 77-107, May 1985              CHER89 - Cheriton, D. and Williamson, Carey L., "VMTP as the       Transport Layer for High-Performance Distributed Systems", IEEE       Communications Magazine, pp. 37-44, June 1989              GAIT89 - Gait, Jason, "A Kernel for High-Performance Multicast       Communications", IEEE Transactions on Computers, Volume 38, Number 2,       February 1989.              HUGH87 - Hughes, L., "An Introduction to Multicast Communications",       Proceedings of the Special Conference of the International Council for       Computer Communications, New Delhi, India, October 1987              PALI88 - Paliwoda, K., "Transactions Involving Multicast", Computer       Communication Review, Volume 11, Number 6, pp. 313-18, December 1988              ... snip ... and ...              Thoughts on Fast Protocols        by Alfred C. Weaver              Protocol speed is affected by its design philosophy, mechanical       definition, implementation, and the circumstances of its eventual use.       My comments are focused mainly on design philosophy. Since the       transport layer is the traditional provider of reliable       communications, these comments are therefore focused on that layer in       particular. Let's examine some common design philosophies.              ... snip ... and ...              (2) A connection must exist before data can be reliably transferred.              The ISO TP4 philosophy is a six-way handshake: connection request,       connect confirm, data transfer, acknowledgment, disconnect request,       and disconnect confirm. The Xpress Transfer Protocol (XTP) has shown       that data can be reliably exchanged with a three-way handshake: one to       establish the context and transfer the data, one to acknowledge the       context and data, and one to terminate the context. There is little       harm in allowing data to accompany the connection request. If the       connection is accepted then data transfer has already begun; if       refused, the loss is just the investment in transmitting one frame.              ... snip ...              the above multicast article reference has mention of VMTP which has       IETF RFC1045.              for some more drift ... my rfc index       http://www.garlic.com/~lynn/rfcietff.htm              --       40+yrs virtualization experience (since Jan68), online at home since Mar1970              --- SoupGate-Win32 v1.05        * Origin: you cannot sedate... all the things you hate (1:229/2)    |
[   << oldest   |   < older   |   list   |   newer >   |   newest >>   ]
(c) 1994, bbs@darkrealms.ca