From: Gang Zhao (galaxy_at_cs.washington.edu)
Date: Wed Jan 07 2004 - 16:00:35 PST
This paper reads like a contemporary patent application.
While most of the instructions seem obvious to a person who has programmed on a modern computer, some of them require further thought.
The join procedure is the equivalent of the WindowsXP WaitForMultipleObjects and signal semantics. The semantics of lock have been picked up largely unchanged in the recent C# programming language.
I liked seeing the "Exceptional Conditions" section. (Was this the first use of the term?) It is interesting to see that at the time it was believed that all of the possible exceptions could be enumerated.
The concept of protected entry points appear useful. Only recently have I seen this concept be truly enforced in a meaningful way (COM interfaces accessed through dotNET).
Naming appears to excite the authors of several papers. In hindsight we see that the naming schemes (including the time based scheme) discussed do not avoid collisions in a network of machines.
Issues with the linking described in the directory structure section are outlined in the article "Protection and the Control of Information Sharing in Multics"
This archive was generated by hypermail 2.1.6 : Wed Jan 07 2004 - 16:00:40 PST