The Plan 9 Network Operating System the OTHER OS Original Photo

Total Page:16

File Type:pdf, Size:1020Kb

The Plan 9 Network Operating System the OTHER OS Original Photo KNOW-HOW Plan 9 The Plan 9 network operating system THE OTHER OS original photo: www.sxc.hu photo: original In the eery, distant days before the birth of Linux, a strange alien sys- tem set out to fulfill the promise of Unix. Descendents of that system are still living. We caught one and dissected it. BY OLIVER FROMMEL inux has its roots in the famous and develop Plan 9. The license was lib- install is spartan, but well-organized and AT&T Bell Labs, home of the orig- eralized over the years, and in 2003, more or less foolproof (Figure 1). The in- Linal Unix system. The Bell Labs Plan 9 was finally released under a rec- staller checks the current status after programmers have been busy ever since, ognized open source license. A commer- each step. An image for graphics adapt- and one of the fruits of their labors is the cial offshoot, known as Inferno, is also ers with Vesa mode support is also avail- distributed operating system Plan 9 [1]. available under a free license [2]. able. Plan 9 began in the late 1980s as a We had a look at the current “4th Edi- If you prefer not to experiment with a new system designed to address some tion” release of Plan 9 to see if this leg- full installation right now, you might like problems with Unix that the Bell engi- endary alternative is still alluring. to try your luck with a virtual machine. neers considered “too deep to fix.” This See the “Virtual Plan 9” box for details. new operating system did indeed come Getting Started If you are interested, you can also try out with some innovations that had an influ- The distribution takes the form of a a userspace-only install including Plan ence on later systems. But until recently, 65Mbyte packed ISO image, which is 9’s most important tools [3]. Plan 9 was under a proprietary license also suitable for use as a live CD. If you that never really caught on with users. have difficulty booting from the CD, you More Consistent than Unix Bell Labs and a small community of can check the website for a boot floppy The underlying concept for Plan 9 is that programmers have continued to perfect generator to match your hardware. The it is a distributed operating system, not 56 ISSUE 62 JANUARY 2006 WWW.LINUX - MAGAZINE.COM Plan 9 KNOW-HOW kernel mainstream vastly simplifies programming models. in version 2.6.14, The underlying model is based on thus removing any Niklaus Wirth’s Oberon [5], which in obstacles to Linux turn draws on the Xerox Cedar system. and Plan 9 interac- Plan 9’s desktop provides a simple pro- tion. gramming model with a file interface and elementary operations. For example, Namespaces a simple cat /dev/screen > output cre- Plan 9 takes this ates a screenshot of an application. In file-based design a contrast to this, the Rio implementation step further with is quite sophisticated: it comprises con- the concept of current processes and threads that call namespaces. Tra- each other reciprocally. ditionally, Unix Rio can be seen as a departure from manages all its re- the legacy, line-oriented terminal model sources in a single that harks back to the days when print- Figure 1: The so-called graphical installation displays the system load namespace, in ers really did echo user input. Linux still and a log window below the text-based menu. The tool suggests the which /dev/tty1 has residual traces of this line-oriented four required partitions. always represents heritage in the form of the many termi- the same terminal. nal window applications, from Xterm to like Unix, where network functionality is In contrast to this, Plan 9 applications the KDE console. This makes it possible added by mechanisms such as remote have their own namespaces, so that to enter input at any position in a win- login and networking filesystems. In /dev/window will always point to an ap- dow: users simply select a command Plan 9, networking is built into the foun- plication’s own window. and use the mouse to execute. As an ad- dations of the operating system. For ex- Union mount is another important fea- vanced windowing system, Rio makes ample, all resources can theoretically be ture. The union mount feature helps heavy use of any mouse buttons you distributed transparently across a Plan 9 simplify the management of resources have available. network. The system hides the fact that such as files and directories by mounting The Acme program is an unusual mix resources are non-local from the user. multiple directories in a single directory. of shell, editor, and windowing system. Plan 9 not only supports file servers, but This makes it possible to mount directo- Acme follows the principle we just out- also authentication servers and CPU ries with executables on the file server in lined but adds elementary menus to ter- servers. You can use Plan 9 to implement the local /bin directory, which removes minal windows and redirects output to grid systems such as 9grid. the need for typical PATH variable tricks. new subwindows (Figure 2). An X11 im- Plan 9 is based on the Unix paradigm Linux just recently adopted this design plementation of Acme, dubbed Wily, is that “Everything is a file.” Due to the with the overlay file system, Union-FS. based on the 9lib library [6]. huge collection of add-ons, Unix was in- Despite its advanced age, Rio can also creasingly forced to accept ad-hoc Graphics with Rio handle multi-language programs, as it changes that contravene this original In contrast to X11 on Unix, Plan 9 di- relies on Unicode character sets. Plan 9 principle. One example of this is the rectly integrates a GUI desktop, and this uses UTF-8 encoding, which was actu- ghastly socket interface, which uses read and write functions that are different The Name from normal files. Plan 9 puts an end to Plan 9 derives its name from the title of a It is interesting to consider whether Bell this, providing file-based system inter- cult trash film “Plan 9 from Outer engineers chose the name because they faces, for example, /net/tcp and /net/ Space,” by the genial Hollywood felt they were resurrecting the mortal re- udp, as network interfaces. dilettante, Ed Wood. According to the mains of Unix to attack the world. The Most system services follow the server FAQ [8], the name upholds the Bell Labs’ tantalizing similarity of "Solaranite" with principle and interact via file-based in- tradition of “selecting names that make "Solaris," another enemy weapon that terfaces. Plan 9 does not have the usual marketeers wince.” derives its power from a different Sun, FTP program. Instead, the ftps server In the film, which is often considered is perhaps stretching the metaphor, al- mounts the directory in /n/ftp on the Wood's greatest contribution to cinema, though the first versions of Solaris ap- server. The 9660fs server is responsible a pair of aliens launch a plan to resurrect peared when Plan 9 was in early devel- for mounting CD ROMs. Plan 9 was one dead people and create an army of opment. of the first operating systems to imple- zombies that will attack the capitals of Other Plan 9 names also contain refer- ment the Proc filesystem to support file- the Earth. The purpose of this gruesome ences to pop culture, such as the Acme based process control. Basic network mission is to stop humans from com- editor, named for the company in the functionality is provided by the Plan 9 pleting their work on the Solaranite, an Warner Brothers Roadrunner cartoon. file protocol, 9P, which comprises about exotic weapon that will "explode the par- The old windowing system was known ticles of sunlight" and destroy the uni- as 8 1/ 2, based on the name of a Fellini 30 protocol messages. A Linux imple- verse. movie. mentation [4] of 9P was added to the WWW.LINUX - MAGAZINE.COM ISSUE 62 JANUARY 2006 57 KNOW-HOW Plan 9 ally invented by the Plan 9 programming team led by Rob Pike. Security Plan 9 does not have a root superuser, and thus it does not have the kind of SUID programs that have caused no end of security problems on Unix. Following a similar approach to Kerberos, a distrib- uted Plan 9 system does not transmit passwords over the wire but uses en- crypted tickets instead. One way to cre- ate a user account, is to enable the Fossil fileserver by entering con /srv/fscons, and then give the uname user user com- mand. New users can initialize their en- vironments the first time they log in by entering /sys/lib/newuser. Of course, Plan 9 has developer tools for the C programming language. The compiler for the x86 architecture is titled 8c, the linker is 8l, and the build pro- gram is mk. To compile a new kernel Figure 2: A Plan 9 system in action. A multiply-tiled Acme editor window is hiding behind the with these tools, the sysop changes the mascot, Glenda, as is evidenced by the menu items: New, Cut, Paste, and so on. directory to /sys/src/9/pc and edits the corresponding configuration file, which able online and in the Plan 9 Wiki. Un- which continues to work hard on devel- can be named differently depending on fortunately, a lot of changes were intro- oping Plan 9, fights a constant battle the environment: in the simplest of duced with the 4th Edition, and this against the lack of resources.
Recommended publications
  • A Novel Scheduling Framework Leveraging Hardware Cache Partitioning for Cache-Side-Channel Elimination in Clouds
    A Novel Scheduling Framework Leveraging Hardware Cache Partitioning for Cache-Side-Channel Elimination in Clouds Read Sprabery Konstantin Evchenko Abhilash Raj University of Illinois, University of Illinois, Oregon State University Urbana-Champaign Urbana-Champaign [email protected] [email protected] [email protected] Rakesh B. Bobba Sibin Mohan Roy H. Campbell Oregon State University University of Illinois, University of Illinois, [email protected] Urbana-Champaign Urbana-Champaign [email protected] [email protected] ABSTRACT on the Last-Level-Cache (LLC) that is shared across multiple cores While there exist many isolation mechanisms that are available [18, 23, 43, 44, 48] – these make defenses much harder. to cloud service providers, including virtual machines, containers, Many defenses against cache-side-channel attacks in cloud envi- etc. , the problem of side-channel increases in importance as a re- ronments have also been proposed (e.g., [6, 19, 21, 22, 24, 25, 28, 31, maining security vulnerability – particularly in the presence of 32, 37–40, 45, 51]). However, the proposed solutions suer from a shared caches and multicore processors. In this paper we present variety of drawbacks: (a) some are probabilistic [25, 37, 51]; (b) oth- a hardware-software mechanism that improves the isolation of ers do not protect applications when SMT is enabled [51]; (c) some cloud processes in the presence of shared caches on multicore require developers to re-write applications [19, 22, 31], (d) while chips. Combining the Intel CAT architecture that enables cache others require hardware changes [39, 40] impacting deployability; partitioning on the y with novel scheduling techniques and state (e) some depend on violating x86 semantics by modifying the reso- cleansing mechanisms, we enable cache-side-channel free comput- lution, accuracy or availability of timing instructions [21, 24, 38] ing for Linux-based containers and virtual machines, in particular, and consequently require signicant changes to the applications.
    [Show full text]
  • Knowledge Management Enviroments for High Throughput Biology
    Knowledge Management Enviroments for High Throughput Biology Abhey Shah A Thesis submitted for the degree of MPhil Biology Department University of York September 2007 Abstract With the growing complexity and scale of data sets in computational biology and chemoin- formatics, there is a need for novel knowledge processing tools and platforms. This thesis describes a newly developed knowledge processing platform that is different in its emphasis on architecture, flexibility, builtin facilities for datamining and easy cross platform usage. There exist thousands of bioinformatics and chemoinformatics databases, that are stored in many different forms with different access methods, this is a reflection of the range of data structures that make up complex biological and chemical data. Starting from a theoretical ba- sis, FCA (Formal Concept Analysis) an applied branch of lattice theory, is used in this thesis to develop a file system that automatically structures itself by it’s contents. The procedure of extracting concepts from data sets is examined. The system also finds appropriate labels for the discovered concepts by extracting data from ontological databases. A novel method for scaling non-binary data for use with the system is developed. Finally the future of integrative systems biology is discussed in the context of efficiently closed causal systems. Contents 1 Motivations and goals of the thesis 11 1.1 Conceptual frameworks . 11 1.2 Biological foundations . 12 1.2.1 Gene expression data . 13 1.2.2 Ontology . 14 1.3 Knowledge based computational environments . 15 1.3.1 Interfaces . 16 1.3.2 Databases and the character of biological data .
    [Show full text]
  • Computerized Patient Record System (CPRS) Setup Guide
    Computerized Patient Record System (CPRS) Setup Guide June 2021 Department of Veterans Affairs Health Data Systems Computerized Patient Record System Product Line Revision History Project Technical Date Patch Page Change Manager Writer 6/17/2021 OR*3*0*547 30 Added row to GUI Parameters SHRPE SHRPE Menu Options for OTH: GUI Add/Edit Local Message for OTH Button. Added Subsection for GUI 37-38 Add/Edit Local Message for OTH Button 10/16/2019 OR*3.0*397 All Added Revision dated REDACTED REDACTED 3/25/2019 (see below). Checked for 508 Compliance. Removed extra space between some pages. Removed extra Table of Contents from Pharmacy Packages section (G). 07/31/2019 OR*3.0*510 88 Added two NOTE: REDACTED REDACTED 3/25/2019 OR*3.0*397 21 Added ORSUPPLY Key REDACTED REDACTED 08/20/2018 XU*8.0*679 28 Added note regarding Electronic REDACTED REDACTED Signature Block restrictions. 03/30/2011 OR*3.0*272 5, 77, Changed references from REDACTED REDACTED 79, Duplicate Drug Class to 81, Duplicate Drug Therapy. 82, 93, 01/25/2007 OR*3.0*245 123 – Added information about the REDACTED REDACTED 125 new option, ORCM QUICK ORDERS BY USER 11/27/06 OR*3.0*242 182 Added information about the REDACTED REDACTED new way to edit items for the Nature of Order file. 12/30/04 36 Added new information about REDACTED REDACTED document templates. Computerized Patient Record System (CPRS) ii June 2021 Setup Guide Project Technical Date Patch Page Change Manager Writer 10/14/98 138 Added information about ORMTIME 10/15/98 14 Added explanation of global journaling 10/20/98
    [Show full text]
  • Acme As an Interactive Translation Environment
    Acme as an Interactive Translation Environment Eric Nichols and Yuji Matsumoto {eric-n,matsu} ߞatߞ is.naist.jp Computational Linguistics Laboratory Nara Institute of Science and Technology ABSTRACT Translation is challenging and repetitive work. Computer-aided transla­ tion environments attempt to ease the monotony by automating many tasks for human translators, however, it is difficult to design user inter­ faces that are easy to use but that can also be adapted to a dynamic workflow. This is often a result of a lack of connection between the inter­ face and the tasks that the user wants to carry out. A better correspon­ dence between task and interface can be achieved by simplifying how software tools are named. One way of accomplishing this to embrace text as the interface. By providing a simple and consistent semantics for interpreting text as commands, the Acme text editor [1] makes it possible to build a system with a text-centered interface. In this paper we explore the implications this has for translation aid software. 1. Motivation Translation is an essential and important part of human communication. However, it is a very challenging task, requiring the translator to have a complete understanding of and a deep familiarity with the source and target languages. This difficulty is not eased by the fact that fundamentally it is an inherently repetitive task, consisting of looking up unfamiliar words and doing large amounts of editing to produce a good translation. Given these demands, computers provide a good way to ease this repetitiveness by automating lookup and editing; by converting resources like dictionaries, other transla­ tions, and collections of example sentences to a computer-readable format, lookups can be performed much faster.
    [Show full text]
  • Header Manipulation Rules Resource Guide
    Oracle® Communications Session Border Controller Header Manipulation Rules Resource Guide Release S-CZ8.0 September 2017 Oracle Communications Session Border Controller Header Manipulation Rules Resource Guide, Release S-CZ8.0 Copyright © 2014, 2017, Oracle and/or its affiliates. All rights reserved. This software and related documentation are provided under a license agreement containing restrictions on use and disclosure and are protected by intellectual property laws. Except as expressly permitted in your license agreement or allowed by law, you may not use, copy, reproduce, translate, broadcast, modify, license, transmit, distribute, exhibit, perform, publish, or display any part, in any form, or by any means. Reverse engineering, disassembly, or decompilation of this software, unless required by law for interoperability, is prohibited. The information contained herein is subject to change without notice and is not warranted to be error-free. If you find any errors, please report them to us in writing. If this is software or related documentation that is delivered to the U.S. Government or anyone licensing it on behalf of the U.S. Government, then the following notice is applicable: U.S. GOVERNMENT END USERS: Oracle programs, including any operating system, integrated software, any programs installed on the hardware, and/or documentation, delivered to U.S. Government end users are "commercial computer software" pursuant to the applicable Federal Acquisition Regulation and agency-specific supplemental regulations. As such, use, duplication, disclosure, modification, and adaptation of the programs, including any operating system, integrated software, any programs installed on the hardware, and/or documentation, shall be subject to license terms and license restrictions applicable to the programs.
    [Show full text]
  • Protected Database Configuration Guide
    Zero Data Loss Recovery Appliance Protected Database Configuration Guide Release 12.2 E88069-06 January 2019 Zero Data Loss Recovery Appliance Protected Database Configuration Guide, Release 12.2 E88069-06 Copyright © 2014, 2019, Oracle and/or its affiliates. All rights reserved. Contributing Authors: Glenn Maxey, Terence Buencamino, Padmaja Potineni Contributors: Andrew Babb, Anand Beldalker, Jin-Jwei Chen, Tim Chien, Sean Connelly, Donna Cooksey, Sam Corso, Steve Fogel, Muthu Olagappan, Jony Safi, Daniel Sears, Lawrence To, Steve Wertheimer This software and related documentation are provided under a license agreement containing restrictions on use and disclosure and are protected by intellectual property laws. Except as expressly permitted in your license agreement or allowed by law, you may not use, copy, reproduce, translate, broadcast, modify, license, transmit, distribute, exhibit, perform, publish, or display any part, in any form, or by any means. Reverse engineering, disassembly, or decompilation of this software, unless required by law for interoperability, is prohibited. The information contained herein is subject to change without notice and is not warranted to be error-free. If you find any errors, please report them to us in writing. If this is software or related documentation that is delivered to the U.S. Government or anyone licensing it on behalf of the U.S. Government, then the following notice is applicable: U.S. GOVERNMENT END USERS: Oracle programs, including any operating system, integrated software, any programs installed on the hardware, and/or documentation, delivered to U.S. Government end users are "commercial computer software" pursuant to the applicable Federal Acquisition Regulation and agency- specific supplemental regulations.
    [Show full text]
  • Oracle CSM-SBC with Broadworks.Pdf
    Oracle Communications Core Session Manager SCZ 7.1.5m1p1 and Oracle Communications Session Border Controller SCZ 7.2.0 with Broadworks Platform R19SP1 Technical Application Note Disclaimer The following is intended to outline our general product direction. It is intended for information purposes only, and may not be incorporated into any contract. It is not a commitment to deliver any material, code, or functionality, and should not be relied upon in making purchasing decisions. The development, release, and timing of any features or functionality described for Oracle’s products remains at the sole discretion of Oracle. 2 Table of Contents INTENDED AUDIENCE ...................................................................................................................................................... 5 DOCUMENT OVERVIEW .................................................................................................................................................. 5 INTRODUCTION ................................................................................................................................................................. 6 AUDIENCE ............................................................................................................................................................................................. 6 REQUIREMENTS ................................................................................................................................................................................... 6 ARCHITECTURE
    [Show full text]
  • Lifetime Support Policy: Oracle and Sun Systems Software
    Oracle and Sun System Software and Operating Systems Oracle Lifetime Support Policy September 21, 2021, 2021 | Version 1.00 Copyright © 2021, Oracle and/or its affiliates TABLE OF CONTENTS This brochure provides lifecycle timeframes for system software that is typically sold Additional Sun software is located in the standalone, but used in conjunction with specific Oracle or Sun products. Product Oracle Fusion Middleware, Oracle areas include server and storage management, Sun third-party, communication and Technology and Oracle Applications collaboration and Sun Ray. Lifetime Support Policy brochures located on www.oracle.com. ORACLE LIFETIME SUPPORT: 2 Please see the Oracle Software Oracle Communications Session Border Controller (Formerly Acme Packet Net-Net) 5 Technical Support Policies document Oracle Communications Tunneled Session Controller (Formerly Acme Packet NetNet) 5 located on www.oracle.com for Oracle Communications Application Session Controller (Formerly Acme Packet Net-Net) 5 additional terms and descriptions. Oracle Communications Unified Session Manager (Formerly Acme Packet Net-Net) 6 Oracle Communications Subscriber Aware Load Balancer (Formerly Acme Packet Net-Net) 6 Oracle Communications Border Gateway (Formerly Acme Packet Net-Net Border Gateway) 6 Oracle Communications Core Session Manager 7 Oracle Communications Session Router (Formerly Acme Packet Net-Net Session Router) 7 Oracle Communications Security Gateway (Formerly Acme Packet Net-Net Security Gateway) 7 Oracle Communications Session Monitor Family of
    [Show full text]
  • Implementing a Parallel World Model Using Linux Containers for Efficient System Administration
    Implementing a parallel world model using Linux containers for efficient system administration (Paper) Yasushi Shinjo and Wataru Ishida Jinpeng Wei Department of Computer Science School of Computing and Information Sciences University of Tsukuba Florida International University 1-1-1 Tennoudai, Tsukuba, Ibaraki 305-8573, Japan Miami, Florida, USA Email: [email protected] Abstract—This paper describes the implementation of a par- software updates. However, this method does not help to run allel world model using Linux containers. A parallel world the production environment while testing. (or parallel universe) model allows a user to create multiple To address these problems, we propose providing a parallel execution environments, called worlds, in a single operating system and to manipulate these worlds. This model enables world model at the operating system level. A parallel world or a system administrator to create a new test world that looks parallel universe model is a model to describe time machines like a production world. The system administrator upgrades in science fiction (SF) and to interpret observations in quantum fundamental software and tests applications in the new test physics. In our parallel world model at the operating system world while running the production world. If the applications level, a world is a container of files and an execution environ- do not pass the tests, the administrator deletes the new world. If the applications pass the tests, the administrator merges the ment of processes. Using this parallel world model, a system test world into the production world. Prior to the merge, the administrator can upgrade fundamental software, as follows.
    [Show full text]
  • The Virtual Filesystem Interface in 4.4BSDI
    The Virtual Filesystem Interface in 4.4BSDI Marshall Kirk McKusick Consultant and Author Berkeley, California ABSTRACT: This paper describes the virtual filesys- tem interface found in 4.4BSD. This interface is de- signed around an object oriented virtual file node or "vnode" data structure. The vnode structure is de- scribed along with its method for dynamically expand- ing its set of operations. These operations have been divided into two groups: those to manage the hierarchi- cal filesystem name space and those to manage the flat filestore. The translation of pathnames is described, as it requires a tight coupling between the virtual filesys- tem layer and the underþing filesystems through which the path traverses. This paper describes the filesystem services that are exported from the vnode interface to its clients, both local and remote. It also describes the set of services provided by the vnode layer to its client filesystems. The vnode interface has been generalized to allow multiple filesystems to be stacked together. After describing the stacking functionality, several examples of stacking filesystems are shown. t To appear in The Design and Implementation of the 4.4BSD Operating System, by Marshall Kirk McKusick, publisher. et al., @1995 by Addison-Wesley Publishing Companf Inc. Reprinted with the permission of the o 1995 The USENIX Association, Computing Systems, Vol. 8 ' No. 1 ' Winter 1995 I. The Virtual Filesystem Interface In early UNIX systems, the file entries directly referenced the local filesystem inode, see Figure I [Leffler et al. 1989]. This approach worked fine when there was a single filesystem implementation. However, with the advent of multþle filesystem types, the architecture had to be generalized.
    [Show full text]
  • A NAS Integrated File System for On-Site Iot Data Storage
    A NAS Integrated File System for On-site IoT Data Storage Yuki Okamoto1 Kenichi Arai2 Toru Kobayashi2 Takuya Fujihashi1 Takashi Watanabe1 Shunsuke Saruwatari1 1Graduate School of Information Science and Technology, Osaka University, Japan 2Graduate School of Engineering, Nagasaki University, Japan Abstract—In the case of IoT data acquisition, long-term storage SDFS, multiple NAS can be handled as one file system and of large amounts of sensor data is a significant challenge. When a users can easily add new NAS. cloud service is used, fixed costs such as a communication line for The remainder of this paper is organized as follows. In uploading and a monthly fee are unavoidable. In addition, the use of large-capacity storage servers incurs high initial installation section 2, we discuss the issues associated with the sensor data costs. In this paper, we propose a Network Attached Storage storage. Section 3 describes the overview and implementation (NAS) integrated file system called the “Sensor Data File System of the proposed file system, SDFS. In section 4, we evaluate (SDFS)”, which has advantages of being on-site, low-cost, and the performance of SDFS and existing file systems. Section 5 highly scalable. We developed the SDFS using FUSE, which introduces related work, and Section 6 summarizes the main is an interface for implementing file systems in user-space. In this work, we compared SDFS with existing integrated storage findings of this paper. modalities such as aufs and unionfs-fuse. The results indicate that II. ISSUES ASSOCIATED WITH THE SENSOR DATA STORAGE SDFS achieves an equivalent throughput compared with existing file systems in terms of read/write performance.
    [Show full text]
  • Design and Validation Computer Protocols
    DESIGN AND VALIDATION OF COMPUTER PROTOCOLS Gerard J. Holzmann -----------, PRE NTI CE HALL SOFTWARE SERI ES www.spinroot.com DESIGN AND VALIDATION OF COMPUTER PROTOCOLS Gerard J. Holzmann Bell Laboratories Murray Hill, New Jersey 07974 PRENTICE-HALL Englewood Cliffs, New Jersey 07632 www.spinroot.com Prentice Hall Software Series Brian W. Kernighan, Advisor Copyright 1991 by Lucent Technologies, Bell Laboratories, Incorporated. This book is typeset in Times Roman by the author, using an Linotronic 200P phototypesetter and a DEC VAX 8550 running the 10th Edition of the UNIX operating system. DEC and VAX are trademarks of Digital Equipment Corporation. UNIX is a registered trademark of AT&T. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior written permission of the publisher. Printed in the United States of America 10 9 8 7 6 5 4 3 2 1 Prentice-Hall International (UK) Limited, London Prentice-Hall of Australia Pty. Limited, Sydney Prentice-Hall Canada Inc., Toronto Prentice-Hall Hispanoamericana, S.A., Mexico Prentice-Hall of India Private Limited, New Delhi Prentice-Hall of Japan, Inc., Tokyo Simon & Schuster Asia Pte. Ltd., Singapore Editora Prentice-Hall do Brasil, Ltda., Rio de Janeiro www.spinroot.com CONTENTS Foreword ix Preface xi Part I Ð Basics 1. Introduction 1.1 Early Beginnings 1 1.2 The First Networks 9 1.3 Protocols as Languages 12 1.4 Protocol Standardization 13 1.5 Summary 15 Exercises 16 Bibliographic Notes 16 2.
    [Show full text]