Juli Mallett

FreeBSD

I have not regularly contributed to FreeBSD for a long time. Previously, I did a lot of work related to FreeBSD on MIPS, especially Cavium Octeon, and various odds and ends related to networking. I am thoroughly out-of-the-loop and FreeBSD has been a part of my life for decades. If I can help you with something, I'm happy to try, especially if you're just getting started with making your own dent in the world of operating systems.

Life

I no longer work in tech professionally. Like many of my peers, the current (as of 2026) trajectory of technology innovation leaves me feeling pretty pessimistic. I find more hope in people and communities. Technology can be a great tool for enabling, connecting, and mobilizing individuals and communities, but it can also do a fantastic job of sending them chasing ghosts and fighting their own shadows, while indifferent to the fallout in the people around them of whose real existence they are ignorant.

Non-Projects

Sometimes I throw together little tools or modify others' tools to suit my own needs. You're probably interested in them. (At least, I can't imagine why else you might actually be reading this.)

• blindvnc, a modified version of TightVNC's vncviewer that is stripped down to support only input.
• diff-filter, a program to select individual files to extract or remove from a diff based on a regex applied to their file names.
• buildomat, a job-controlling, web-displaying front-end to the FreeBSD build process.
• sboxes-1, some randomly-generated 8-bit s-boxes with useful properties. (And some without many useful properties.) I really wish there were an s-box database like the OEIS where people could annotate them with useful properties like "no fixed points" "has fixed points when used in a substitution-rotation cipher" or even "has fixed points when used in a 16-bit substitution-rotation ciphers along with [some other s-box]." I will, of course, be glad to add annotations to this list of s-boxes if anyone has any interesting observations that they decide to share with me.
• Octeon dmesg(8)s, a collection of boot output logs from various Octeon systems running FreeBSD/MIPS.
• Pharaoh's Shuffle, a listing of results from a Pharaoh's Shuffle done on decks of cards with various sizes, demonstrating the incorrectness of assertions that 8 shuffles will preserve the order of a deck of any size. Results are also shown for doing in-shuffles rather than the standard out-shuffles, as well as for alternating between doing in- and out-shuffles.
• BSD+, a set of libraries and utilities meant to be useful on BSD systems, or to add new functionality in the BSD tradition. At the time of writing here it is still fairly trivial, including diff-filter mentioned above and tools for getting the device name associated with a (possibly-renamed) network interface.
• Key flopping, a proposal to increase robustness of cryptosystems which use public keys to identify a host.

Projects

I have released a WAN optimization proxy which does dictionary compression (also called "deduplication" in the storage industry) under the name WANProxy. It is available under a BSD license. It is relatively stable and supports a number of platforms.

I do love to tinker, and used to maintain a list of little projects I'd worked on recently here, and longer-term research interests, but it's difficult to keep up with it, and it didn't seem to interest anyone much other than me.

Impossible Algorithms

I'm interested in collecting a list of algorithms which can't exist, but might sound like they could. Algorithms which would make you rich beyond reason if you could discover them, but which you know violate basic axioms. You know, the kind of things Hollywood might presume to exist.

• Proof-of-deletion; a cryptographic algorithm which can be used to prove to another party that all copies of a piece of information can be deleted. (NB: That sounds even more narrow than it is; proving that any copy of a piece of information has been deleted can be extended to proving that all copies of a piece of information have been deleted. You can model it as a refcount, so the number of readers doesn't matter while it's over 1; verifying the number of copies in existence is the same as verifying that there are 0 copies.) The hard part is to do this without relying on any service, i.e. some kind of information escrow which uses DRM or the like to resist copying. Cryptographic proof of non-copying and of deletion is the hard part. Might be feasible with quantum computing, since you can determine whether quantum information has been observed or erased, right? (I'm no expert on that.) You can prove that it's impossible just by virtue of the fact that a determined human can remember anything, whatever demands you might impose on a computer. Still, think of the possibilities! Great for blackmail, all kinds of privacy issues, Bitcoin escrow, etc. Too bad it's impossible. But what interesting subsets aren't?

Contact

Feel free to email me jmallett@FreeBSD.org. I get a ton of spam and non-spam mail, but try to respond to messages sent to me.

Remembering

I wish to remember Alan Eldridge, who died in 2003. He and I had been talking about the dire job situation just before his death, and I got a great job very shortly after. I wanted so badly to tell him how much better the market was getting, and that there was still hope, and I couldn't. Many have mentioned his love of Warren Zevon, and when I (however infrequently) hear a Warren Zevon song, I still think of him.

I wish also to remember Cameron Grant, who died in 2005. He was a great friend, and I still miss him acutely. In 2007, I had the pleasure of meeting his father, and my partner and I bought a car that had been his to do the Mongol Rally. The trip was in many ways a chance to reflect on Cameron's passing, and to try to do some good things in his memory, and to have fun in his spirit. His impact on my life was major; in the short time we knew eachother, he greatly helped me in coming to terms with myself and the difficulties of life and other people. He was a truly charming man.

I wish also to remember Andy Purshottam, who died in 2007. His presence was always strange and awkward, delightful and unexpected. He regularly shared his thoughts with friends via email, and I was always pleased to find a new one greeting me in my inbox, with far-flung links and thoughts on technology. He moved with a gentle self-assuredness and had a fierce inability to deal with people who were there to waste his time. He despised fools and corruption, and was always glad to explain himself. His emails stopped, and I was informed that he had died, unexpectedly and sadly. I find myself wanting to share thoughts and links with him, and struggle to remind myself that he is no longer there.

I remember also Robert Mallett, my grandfather, who died in 2013. He was an engineer and an avid Unix fan, and had no small influence on my own passion about computers and technology, and the shape of my career. There is too much to say about him personally, but in this list in particular it seems like I must say that he was always curious, and encouraged me to be as well; he imbued me with a love of science, math and experimentation. As my great engagement with Unix really flourished, he gave me use of piles and piles of old Sun and SGI equipment, and would remind me how to get out of vi(1) when I got stuck. He was so much more than that to me, of course; I grew up very close to him. I've always hoped to please and impress him with my career and my work, and I think it important to say to those who value my technical output that I could not possibly have done any of it without him.

I wish to remember Alex Botero-Lowry, who died in 2012. We had been friends since the late 1990s, but never quite managed to meet up, always barely passing each other in space. We met and bonded around a shared obsession and fascination with computers of a certain era, the hopeful and expansive aesthetic that permeated the late '80s and early '90s, an era which was clearly formative for us both. Being able to share that imagination and enthusiasm for technology with someone was very important to me. As we grew up, we shared our struggles and the many changes in our lives with each other. He was always bright, energetic, sweet and reflective. It pains me to know that we will not manage to meet up some day in San Francisco or Portland after all, and the future seems a little bleaker for its lack of him.