Naming and Synchronization in a Decentralized Computer System (1979) [pdf]

dtic.mil

54 points·da02··9 comments

David P. Reed has a very impressive CV.

Designer of UDP/IP, per above thesis also inventor of MVCC, and a co author of the (imho) important "End-to-end arguments in system design" whitepaper, etc ...

https://en.wikipedia.org/wiki/David_P._Reed

[edit: alt url for OP: https://dspace.mit.edu/bitstream/handle/1721.1/16279/0533164...]

This won't download for me but I was able to get it quickly here: https://dspace.mit.edu/handle/1721.1/16279#files-area

This version is closer to the dtic.mil OP version: http://www.bitsavers.org/pdf/mit/lcs/tr/MIT-LCS-TR-205.pdf

It looks visually almost identical; one neat thing about the dtic.mil version is that it is a reproduction from the microfiche scan of the document (!).

The version at dspace however is paginated and laid out very differently. The TR/dtic.mil version has around 187-188 pages, depending on how you count; the dspace version has 217.

Dave Reed (author of the linked paper) also gave a fantastic presentation on the End-to-End Arguments a few years ago: https://www.usenix.org/conference/usenixsecurity10/end-end-a....

(1978)

September, 1978.

It's David P. Reed's PhD thesis, which led among other things to the fascinating Croquet [0,1] system. When I visited Viewpoints Research in 2014 while working on my own approach to concurrent programming, Reed's thesis was specifically recommended to me as well worth studying.

[0] https://en.wikipedia.org/wiki/Croquet_Project#Synchronizatio...

[1] https://dx.doi.org/10.1109/C5.2003.1222325

On the title page, below the “September, 1978” date, it is handwritten “(i.e. February, 1979)”.

It was signed and publshed in September 1978, says I, the author. The February date is the date of the Ph.D. degree award, which officially happens at the end of each semester.

I stand corrected.

Reply on news.ycombinator.com