NCP Status Report: UCSB/Rand
A status report from UCSB and RAND sites on their network implementation
Harslem, E., Heafner, J.F., and J.E. White, "NCP Status Report: UCSB/Rand", RFC 78, DOI 10.17487/RFC0078, October 1970, <https://www.rfc-editor.org/info/rfc78>.
@misc{rfc78,
series = {Request for Comments},
number = 78,
howpublished = {RFC 78},
publisher = {RFC Editor},
doi = {10.17487/RFC0078},
url = {https://www.rfc-editor.org/info/rfc78},
author = {E. Harslem and J.F. Heafner and J.E. White},
title = {{NCP Status Report: UCSB/Rand}},
pagetotal = 1,
year = 1970,
month = oct,
} RFC text
Network Working Group E. Harslem
Request for Comments: 78 J. Heafner
NIC 5199 J. White
NCP Status Report: UCSB/RAND
NCP Checkout
UCSB an RAND recently conducted UCSB console to/from RAND console
validation fo teh respective NCPs. The exercise was simply to verify
that the NCPs behaved according to Document #1 protocol.
What was Learned
The UCSB and RAND NCPs responded as specified in Document #1. RAND
uncovered a few minor problems with their Network Services Program
(NSP) that interfaces video graphics consoles to the NCP. The nature
of these problems were data format conversions and human factors
aspects of display presentation of status information.
The Immediate Future
1. Both UCSB and RAND will be happy to participate with other sites
in verifying their NCPs.
2. We intend to conduct additional exercises involving data re-
configurations and file transmissions.
JFH:hs
[ This RFC was put into machine readable form for entry ]
[ into the online RFC archives by Josh Elliott 1/98 ]
AI ELI5
On this page
§1NCP Checkout Results
RFC 78 reports on validation testing conducted jointly by UCSB and RAND on their Network Control Program implementations. The exercise involved console-to-console testing to verify that both NCPs behaved according to the protocol specification in Document #1.
The RFC was authored by E. Harslem, J. Heafner, and J. White from UCSB and RAND respectively, providing an early account of intersite protocol verification on the ARPANET.
Two network teams tested their programs by sending messages back and forth through their network boxes.
§2Testing Outcomes and Minor Issues
Both UCSB and RAND NCPs responded correctly to the testing and behaved as specified in Document #1. RAND discovered a few minor problems, but these were not in the NCP itself. Instead, the issues appeared in the Network Services Program (NSP), which bridges the gap between the NCP and video graphics consoles.
The problems were limited to data format conversions and human factors aspects of displaying status information. No fundamental protocol problems were discovered.
The core programs worked fine. The only problems were in how the computer showed what was happening on the screen.
§3Next Steps
Both sites expressed willingness to participate in further NCP validation exercises with other ARPANET sites. They planned additional testing involving data reconfigurations and file transmissions to expand their understanding of protocol behavior under different conditions.
This report exemplifies the collaborative, pragmatic approach taken by early ARPANET implementers as they worked to ensure protocol correctness across different sites.
The teams said they would be happy to test with other sites and try more complicated message flows.