Geoffrey Allen Collyer

2565 Mountain Ave.

Scotch Plains, NJ 07076

908 517 9109 (home)

Mail address: geoff.work@collyer.net

Citizenship: Canadian, but also US Permanent Resident

Job Goals: Making system software more flexible, simple and powerful.

Work Experience:

● Member of Technical Staff, Research, Bell Laboratories, Lucent Technologies, Murray Hill, NJ

full-time from June 2006 to the present. I supported and assisted with research projects in Wireless Research and Computing Research, including upgrading the Plan 9 infrastructure and keeping it running smoothly. This included migrating terabytes of data from older to newer file servers. I ported Plan 9 to two Power-PC-based Xilinx evaluation boards, several ARM-based systems, including the Sheevaplug, Gumstix Overo, and dual-core Compulab Trimslice, and the MIPS-based Routerboard RB450G. I also improved the implementation of IPv6 in Plan 9, notably connecting better to systems with IPv4 and IPv6 addresses.

● System Programmer, 9netics, Vashon, WA

from April 2006 to May 2006, working for equity only. Continuing work on Shared Name Spaces, getting it into shape for Angel and Venture funding.

● Senior Engineer, Infotone Communications, Redwood City, CA

full-time from May 2005 to November 2005 and as a contractor from November 2005 to March 2006. Maintained complex integration software written in Java. Wrote a compiler to simplify this task by generating the more difficult parts of the Java code.

● System Programmer, 9netics, Vashon, WA

from March 2004 to April 2005. Implemented a multiplexing and caching file server as a component of Shared Name Spaces, an application to enable collaborative work. Extended the Plan 9 file server to use 64-bit block numbers and file sizes (from 32-bit ones). Wrote a small XML parser and a server plug-in for Real Networks. Debugged another Real Networks server plug-in.

● Unix/Linux consultant

from February 2003 to February 2004. Consulted to individuals needing assistance with Unix and Linux, notably modifying and configuring sendmail to throttle spam. I also configured FreeSwan opportunistic IPSEC VPNs on Linux.

● Senior Engineering/Scientist II, BSD Technology group, Apple Computer, Cupertino, CA

full-time from March 2002 to August 2002. Fixed bugs in the release that became Mac OS X 10.2.

● Senior Software Engineer, Teleias International, Walnut Creek, CA

full-time from April 2001 to July 2001. I led the software development of a new remote system management product, written in Python. I set up an IPV6 network of IPV6-over-IPV4 tunnels, using IPSEC to secure the tunnels.

● Member of Technical Staff, Research, Bell Laboratories, Lucent Technologies, Murray Hill, NJ

full-time from August 1995 to April 2001. I supported the staff of the Computing and Communications Research Divisions for a year, notably setting up the bell-labs.com mail gateways, on Plan 9 systems, and the web servers, inside and outside the firewall. I supported the SGI multiprocessor servers and network gear of the Math center (Ethernet, FDDI, ATM, and a Research-backbone Cisco router with all those interfaces), and the Network Appliance file servers of the Communications division. I was the primary contact for Linux support.

In 1996 I transferred into a department of the Speech Research center. I set up the Plan 9 cluster in my department’s lab, ported the upas mail system from the Plan 9 distributed operating system to UNIX, and wrote a scalable, multi-threaded multi-media messaging system for the Inferno distributed operating system, including designing and implementing a replacement for the SMTP protocol, making UNIX applications (OCR and text-to-speech) accessible from Inferno by writing a Styx server for them, and writing a general DNS client for Inferno.

I wrote a diagnostic for the Ethernet boards in Lucent’s PathStar router, helped set up the documentation infrastructure for PathStar, and provided a secure connection client and library for PathStar’s management command-line interface.

● Member of Technical Staff, AT&T Consumer Products Division, Murray Hill, NJ

full-time from July 1994 to August 1995. I wrote the operating system for the AT&T HomeCenter set-top box, including DSP communication. This operating system borrows from Plan 9 and UNIX but is for the AMD 29200 processor. Hardware limitations led to the use of software interrupt queueing and a paper in Computing Systems describing same.

● Systems Analyst for Software Tool & Die, Brookline, MA

full-time from March 1991 to July 1994. I further developed the C News network news software on behalf of ST&D’s client, UUNET Communications Services of Falls Church, VA. Much of this work was performance tuning for sites with massive fan-out and writing of generally-useful library functions. I invented, implemented and got the rest of the Usenet community to adopt the ‘news overview’ (nov) database for news articles. During 1993 and 1994, I also did some work on behalf of ST&D’s clients, the WELL of Sausalito, CA, and Clarinet Communications of San Jose, CA.

● System Programmer for University of Toronto Department of Computer Science, Toronto, Ontario

part-time from November 1990 to March 1991. I wrote UNIX system software while waiting for my delayed American work visa to appear.

● System Programmer for University of Toronto Department of Statistics, Toronto, Ontario

full-time from November 1986 to October 1990. I administered and supported the department’s UNIX systems: a shared v7 PDP-11/70, one Silicon Graphics 4D220, seven Sun 3s and one Sun 2. These systems were shared with the Department of Biostatistics. I installed Sun OS on all our Suns, configured the network, and moved the users and applications from the v7 PDP-11/70 to the Suns. I adapted and installed support for the AT&T DMD 5620 and MTG 630 terminals, providing bitmapped graphics before X11 was usable and before the UNIX window system wars had died down. I replaced stock Sun software, including sendmail and uucp, with faster and more reliable software, notably the upas mailer. I contributed much of the first SMTP implementation for upas. I debugged the tbl text formatting software and the standard command interpreter (sh) and replaced its internal memory management to make it more portable. I wrote a manual-page ‘decompiler’ called nam to convert formatted manual pages back into troff -man input.

● System Programmer for University of Toronto Computing Services, Toronto, Ontario

full-time from August 1981 to October 1986. I administered and supported UNIX systems: seven VAXes, an IBM 3033U16 running Amdahl’s UTS under VM on multiple virtual machines, a PDP-11/70, and several Suns. The VAXes and 3033 were used by undergraduate students for course work; the PDP-11/70 and Suns were used for statistical computing (using S) by the Departments of Statistics and Biostatistics, and for commercial typesetting.

I tightened up security on the PWB/UNIX PDP-11/70 system when I arrived, and later on the 4.1BSD VAX systems. I back-ported the uucp and news (Usenet) network software to PWB/UNIX. I debugged and maintained a half-dozen local spoolers and device drivers. I fixed the accounting system and modified it to run much faster.

I converted the PDP-11/70 from PWB/UNIX to Seventh Edition (v7) UNIX. I modified v7 UNIX and improved its performance in part by putting temporary files and pipes in memory. I rewrote the driver for a serial-line terminal multiplexor to improve speed and generality. I repaired damaged file systems. I put routing for UUCP mail in sendmail. I squeezed a bootstrap program to fit more intelligence in and boot an order of magnitude faster.

I installed 4.1BSD (Berkeley) UNIX on all our VAXes and later converted them to 4.2BSD, the first release with TCP/IP and Ethernet support. I moved the typesetting users from our PDP-11/70 to UTCS’s 4.2BSD system. I helped establish the campus IP network, including installing Ethernet, Hubnet and serial-line IP networks. I helped debug the 4.2BSD kernel IP point-to-point packet routing code for our DEC DMR-11 network links. I rewrote the network news receiver, rnews, to consume less than one-twentieth the CPU and elapsed time (this software became the core of C News). In the process, I rewrote the standard I/O library functions fgets, fputs, fread and fwrite to make them much faster.

In my spare time, I wrote an ANSI-compatible standard I/O library that influenced the 4.4BSD one, and helped to design a cheap DMD-5620-like terminal and message-passing operating system for same.

● Programmer/Analyst for the Vancouver Stock Exchange, Vancouver, B.C.

full-time from March 1980 to June 1981. I maintained front-end serial line concentrators to the IBM mainframe, wrote RT-11 device drivers in the C language rather than assembler, wrote new applications and ported assembly-language tasks from the old PDP-8 line concentrator to a new PDP-11.

● System Programmer for the Department of Nuclear Medicine, St. Joseph’s Hospital, London, Ontario

full-time from May 1978 for one year. I maintained RT-11 and UNIX systems, and ported RT-11 programs to v6 UNIX. I wrote device drivers for both operating systems for such devices as a colour television, a storage monitor, radiation scintillation counters, and a gamma-radiation camera.

Education Synopsis:

● Student in the Bachelor of Science Program (Major - Computer Science)

The University of Western Ontario, London, Ontario, CANADA, September 1975 to April 1978.

I persuaded the Computer Science department to obtain its first UNIX licence, for Sixth Edition UNIX, and wrote the first widely-used mail system for the local DECsystem-10 timesharing system.

● Ontario Secondary School Honours Graduation Diploma (Grade 13)

H. B. Beal Secondary School, London, Ontario, CANADA, June 1975.

I wrote an interpreter for a simple assembly language (ICL) for a hypothetical machine, which saw daily production use in teaching for about ten years, and which I ported to another model of machine at the request of Honeywell Information Systems. The first version was written in COBOL and the second in Honeywell H-200 assembler.

Languages:

I have programming experience in many languages, including the following: C, Limbo, awk, Java, Simula, Pascal, Ratfor, and Fortran; and the assembly languages of these machines: ARM, MIPS, PowerPC, AMD 29000, Intel x86, PDP-11, PDP-10, PDP-8, and PDP-15.

Computers and Operating Systems:

My experience has been on such computer systems as:

Patents:

I am an inventor of US patent #6,519,327, System and Method for Selectively Retrieving Messages Stored on Telephony and Data Networks. This patent has been assigned to Lucent Technologies and arose from my work at Bell Labs.

I am an inventor of US patent #7,103,635, Really Simple Mail Transport Protocol, describing a protocol I designed. This patent has been assigned to Lucent Technologies and arose from my work at Bell Labs.

Professional Societies:

I am a member of the Institute of Electrical and Electronics Engineers (IEEE).

Awards:

● Winner, Usenix Quiz Show, 1996 San Diego Usenix Conference.

● Honeywell Information Systems Data Processing Award, 1975.

● Asteroid 129101 Geoffcollyer named for me, 2006.

Publications:

1. Geoff Collyer, ‘‘Recent Plan 9 Work at Bell Labs,’’ Fifth International Workshop on Plan 9, Seattle, invited talk (October 2010).

2. Geoff Collyer, Charles Forsyth, ‘‘A Cache to Bash for 9P,’’ Fifth International Workshop on Plan 9, Seattle (October 2010).

3. Sape Mullender, Geoff Collyer, Second International Workshop on Plan 9, Bell Labs, Murray Hill, NJ (December 2007).

4. Skip Tavakkolian, Geoff Collyer, ‘‘Rangboom,’’ First International Workshop on Plan 9, Madrid (December 2006).

5. Geoff Collyer, ‘‘Setting Interrupt Priorities in Software via Interrupt Queueing,’’ Computing Systems 9(2), pp. 119-130} (Spring 1996).

6. Geoff Collyer, ‘‘A Comment on Open Systems,’’ ;login: 17(6), p. 22} (December 1992).

7. Henry Spencer, Geoff Collyer, ‘‘#ifdef Considered Harmful, or Portability Experience with C News,’’ Proc. Summer Usenix Conf. San Antonio 1992, pp. 185-197} (June 1992).

8. Geoff Collyer, ‘‘A Partial Tour Through the UNIX Shell,’’ Proc. Winter Usenix Conf. San Diego 1989, pp. 343-353} (January 1989).

9. Geoff Collyer, Henry Spencer, ‘‘News Need Not Be Slow,’’ Proc. Winter Usenix Conf. Washington 1987, pp. 181-190} (January 1987).

10. Ian Darwin, Geoff Collyer, ‘‘Can’t Happen or /* NOTREACHED */ or Real Programs Dump Core,’’ Proc. Winter Usenix Conf. Dallas 1985, pp. 136-151} (January 1985).

11. Ian Darwin, Geoff Collyer, ‘‘The Evolution of UNIX - 1974 to the Present,’’ Microsystems 5(11), pp. 44-56} (November 1984).

12. Geoff Collyer, ‘‘Writing Position Independent Code for the PDP-11 in C,’’ Whitesmith’s C letter 2(2), pp. 12-14} (April 1981).

13. Geoff Collyer, ‘‘Reply to Brett Hallett,’’ Simula Newsletter 8(4), p. 7} , Letter to the editor (November 1980).

14. Geoff Collyer, ‘‘A Proposal for an Othello Referee,’’ ;login: 5(6), pp. 10-13} (August 1980).