Operating Systems Structure
Session Chair: Wu-chi Feng, Ohio State University

A Scalable and Explicit Event Delivery Mechanism for UNIX
Gaurav Banga, Network Appliance, Inc.; Jeffrey C. Mogul, Compaq Computer Corporation, Western Research Lab; Peter Druschel, Rice University

The Pebble Component-Based Operating System
Eran Gabber, Christopher Small, John Bruno, José Brustoloni, and Avi Silberschatz, Lucent Technologies-Bell Labs

Linking Programs in a Single Address Space
Luke Deller, and Gernot Heiser, The University of New South Wales

Big Data and the Next Wave of InfraStress Problems, Solutions, Opportunities
John R. Mashey, Silicon Graphics/Cray Research

Data storage is growing at a higher rate than ever before, and coupled with rapidly increasing demand for instant access, will cause great stress on both the physical and the human infrastructure of computing. System planners and administrators will soon face the interesting challenge of dealing with network and backup issues when office systems hold 100s of GB of disks, and larger servers reach 10s and 100s of TB and even PB. There will also be great opportunities in both research and commercial applications, but the problems must be understood, and solutions anticipated. This talk will give some examples, including some large customer problems that Silicon Graphics has been working on; and examine technology trends in storage capacities, access times, computer architectures, and bandwidths, to see what these portend over the next few years.

Applications
Session Chair: Jason Thorpe, NetBSD

Berkeley DB
Mike Olson, Keith Bostic, and Margo Seltzer, Sleepycat Software, Inc.

The FreeBSD Ports Collection
Satoshi Asami, FreeBSD

Multilingual vi Clones: Past, Now and the Future
Jun-ichiro itojun Hagino, KAME Project; Yoshitaka Tokugawa, WIDE Project

Storage Systems
Session Chair: Mirjana Spasojevic, Hewlett-Packard Labs

The Design and Implementation of DCD Device Driver for UNIX
Tycho Nightingale, University of Rhode Island; Yiming Hu, University of Cincinnati; Qing Yang, University of Rhode Island

An Application-Aware Data Storage Model
Todd A. Anderson and James Griffioen, University of Kentucky

What's Wrong with HTTP And Why It Doesn't Matter
Jeffrey Mogul, Compaq Western Research Lab

HTTP quickly grew to become the dominant protocol on the Internet, but its maturation as a protocol design hasn't been as speedy. The HTTP/1.0 specification was written only after the protocol had been deployed, and the IETF working group chartered to design HTTP/1.1 took 4 years to produce a Draft Standard. What we have now is a useful but still seriously flawed protocol.

Jeffrey Mogul was one of the primary authors of HTTP/1.1. This talk will give his personal view of what is still wrong with HTTP, and what we can learn from these mistakes. These include fundamental conceptual errors (the lack of true extensibility, the inappropriate analogy to MIME, and the confusion around caching) and some other problems with the standardization effort. This talk will explain why he doesn't think these errors matter and how HTTP, flawed as it is, still solves problems. This talk will also describe why various efforts to extend or replace HTTP may not pay off.

Kernel
Session Chair: Jason Thorpe, NetBSD

Improving Application Performance through Swap Compression
Raul Cervera, Toni Cortes and Yolanda Becerra, Universitat Politècnica de Catalunya--Barcelona

New Tricks for an Old Terminal Driver
Eric Fischer, University of Chicago

The Design of the Dents DNS Server
Todd Lewis, MindSpring Enterprises

Works-in-Progress Session
Session Chair: Keith Smith, Harvard University

WORKS-IN-PROGRESS REPORTS (WIPS)
Do you have interesting work you would like to share, or a cool idea that is not yet ready to be published? The USENIX audience provides valuable discussion and feedback. Short, pithy, and fun, Works-in-Progress Reports (WIPs) introduce interesting new or ongoing work. We are particularly interested in presentation of student work. Prospective speakers should send a short one- or two-paragraph report, to Keith Smith at wips99@usenix.org.

There are a limited number of slots available for work-in-progress presentations. Proposals for WIP presentations will be accepted at the discretion of the WIP chair, with preference given to those that are received earliest.

UNIX to Linux in Perspective
Peter Salus, UNIX Historian

Born in 1969, UNIX grew, matured, morphed and was even cloned. Its maturation cycle created international standards as well as multiple for-profit and not-for-profit companies. It became the lingua franca of the computer research and development community. Today, the many variants of UNIX claim 30 million users worldwide.

UNIX was 22 when Linus Torvalds created Linux, a UNIX clone. By 1998, this clone had 5 million users in its own right. Earlier decades had seen successful UNIX strains arise, but as of today fewer than 5 major variants survive. This talk will briefly recap 1969-89, concentrating on the exfoliation of UNIX and its clones over the past 10 years.

USENIX '99 Technical Program    [Wednesday, June 9]    [Thursday, June 10]    [Friday, June 11]