Programmer at HexstreamSoft (self-employed)

I open-sourced 100% of my completely undocumented and effectively unusable code from my Hexstream Software (BahagonTools) era, first under the MIT license and then quickly under the Unlicense (public domain).
For several years I planned to improve these codebases to make them usable, but my skills having substantially improved, I eventually reviewed and deleted 48 old projects I no longer care about.
I redirected my efforts towards my newer, suitably engineered, documented, actually worthwhile projects.

Having learned from past mistakes, I now specialize in micro-libraries, with growing success in adapting my extreme modularity skills to bigger, higher impact projects.
I am finally achieving traction and accelerating momentum in completing my exciting roadmap.

Creating important Common Lisp community resources has been a particular focus of my work, and I will continue to invest heavily in this endeavor.

I am spearheading the nascent movement to make Common Lisp a top 5 programming language by 2040. I believe this is eminently achievable given proper approaches.

See my Accomplishments section below for a summary of my extensive Common Lisp Open Source contributions.

Programmer at Hexstream Software (self-employed)

Weird era of surprisingly advanced programming using nascent software engineering and project management skills, entirely on my own. Several pearls in a sea of dreck.

While working on my BahagonTools mega-project (tools for players of the Bahagon MMORPG), I spent almost all my time doing pure Common Lisp R&D while building my own advanced (but badly engineered) web framework from scratch (using Hunchentoot for basic HTTP). I became a Common Lisp expert almost entirely by reading the CLHS and writing my own web framework.

I single-handedly designed and implemented the following from scratch:

- Several DSLs, for HTML, CSS, HTML rewriting, form validation and a bunch of other things, compiling to closures being a favorite technique. Declarative Programming is the ultimate paradigm!

- An HTML rewriting and injection system. This greatly simplified and automated the writing of new static and/or dynamic pages while eschewing unnecessary layers of fake "abstraction".

- A server-side form validation system notably featuring efficient declarative rete-based merging of error messages.

- "flexiconf", a fully declarative configuration management system notably featuring bijective macros and the ability to load a configuration, modify it in-memory and then dump it back in the appropriate files, preserving all the appropriate structure and abstractions, thus supporting the full configuration evolution cycle.

- A 25-tables SQL schema using Postgres.

- An authentication system. The passwords were stored hashed and salted.

- Tons of other cool stuff!

Version control evolution: none, subversion monorepo, darcs monorepo, git monorepo, one git repo per project.

I did successfully launch BahagonTools in early 2010, the site worked but was an utter commercial failure. The eventual goal was going freemium but no premium features were ever implemented. I took the site down after a few months due to lack of interest and prohibitive hosting costs: 80$/month for a dedicated server.

Computer Programming at Collège de Bois-de-Boulogne (completed 4 terms out of 6)

Grade: Best of my class in programming

Activities and Societies: Club Info

I programmed even more BETWEEN courses (at home and at school, and between semesters) than within my programming courses.

I started building crude GUI apps on my own while others were still using the console and JOptionPane. My BahagonTools app eventually reached a fairly high level of polish, the most advanced tool being a combat report analyzer featuring a custom list item renderer.

I fairly often helped others with programming, especially in the beginning, and I once gave an impromptu presentation to most of the class about the clear superiority of Eclipse over NetBeans, which was very well-received and also impressed the teacher. (I proceeded to install Eclipse dozens of times on school workstations over the years.)

I introduced and explained the concept of an event loop to one of my computer science teachers. 😂

I dropped out for personal reasons.
(And the 2 hours per day commute was death.)

30+ ready-to-use libraries in Quicklisp

I am the sole author and maintainer of 30+ ready-to-use Public Domain Common Lisp libraries in Quicklisp. (Excluding CLHS part of CLHS ASDF wrapper.)

Most of them are simple, some of them are highly innovative, all of them are useful in some way.

I love making highly focused, lightweight, single-issue libraries. I have started adapting my extreme modularity skills to bigger, higher impact projects.

Some top highlights:

- compatible-metaclasses completely solves an important and traditionally nearly intractable problem using a very innovative and simple solution. Features a metametaclass!

- canonicalized-initargs finally provides a pleasant standard solution after decades of awkward ad-hoc canonicalization. It uses a clever novel technique to remove the traditional requirement that subclasses of instances of metaclasses that customize the inheritance of slot options must also be instances of that metaclass. In other words, the metaclass is fully metaclass-compatible with cl:standard-class. This provides tremendous convenience for the user, and allows subclasses to retain all the special optimizations that direct instances of cl:standard-class usually benefit from.

- cesdi seamlessly solves a set of grating API design problems in a specific part of the MOP, by implementing a more ergonomic API on top of the original.

- definitions-systems finally extracts an extremely common design pattern (traditionally written out by hand ad-nauseam or using brittle macros) into a proper simple reusable unified extensible object-oriented system offering tons of convenience features and a greatly reduced documentation burden.

- place-modifiers essentially gives access to hundreds of modify-macros through one single macro: modify. This is a natural complement to setf.

I'm also quite proud of my various trivial libraries seamlessly addressing various minor but grating issues in Common Lisp, which in aggregate greatly increase my satisfaction and productivity with the language.

Modern public domain CLOS MOP specification

Chapters 5 and 6 of The Art of the Metaobject Protocol, adapted to modern web standards.

Notes and tips: Standard Common Lisp symbols

Informal yet helpful information on the standard symbols and their bindings.
Definitely one of my biggest projects yet. Still incomplete, but already useful.

Epic FORMAT table

Summarizes all the features of Common Lisp's FORMAT facility in one table.

Common Lispers list

Discover Common Lisp open-source contributors and their best contributions! Add yourself!

I am particularly proud of the extensive set of robust policies I quickly created for the project.

Single-handedly creating the initial list of 100 people was quite a feat, in my humble opinion.

Awesome Planck EZ keyboard configuration

This is an extremely advanced presentation for my highly innovative Planck EZ keyboard configuration.

This highly interactive presentation is implemented in pure HTML5 and CSS3, with no JavaScript involved.

I implemented this in 5 days. The CSS is a bit scary, but probably compares favorably to the equivalent JavaScript.

I showed this to the great folks at Ergodox/ZSA and they were so impressed that they quickly offered to interview me!

TypeMatrix 2030 written by hand in raw SVG

Have a look at the very clean hand-written source of the image with Ctrl-U. CHECKMATE, INKSCAPE!!

I am very proud of my design for the green LED lights in particular, for aesthetic and technical reasons.

Note that this is a DRAFT, the ridge things in particular are incomplete and ugly.

English

Native or bilingual proficiency

French

Native or bilingual proficiency

Spanish

Professional working proficiency

Portuguese

Limited working proficiency

Triplebyte quiz (generalist track)

Score: 21/25 (84%)

Strictly-timed multiple-answers online test using multiple programming languages, with some language-agnostic (e.g. software design) questions.

Automated test assessment said I did "exceptionally well", and that I could be matched with "approximately 12 top tech companies".
I got 5/5 in "Low-level Systems" and "Back-end Web", and 4/5 in "Academic CS" and "System Architecture".

These results eventually prompted Triplebyte to contact me regarding a potential 100$/hour flexible part-time remote job.

Pascal Costanza

I'd like to highlight one of Pascal's many important contributions to the Common Lisp community.

Pascal is the author and maintainer of closer-mop, demonstrably one of the most important Common Lisp libraries. Indeed, according to official Quicklisp downloads statistics, closer-mop was #2 in terms of direct and indirect downloads for 5 consecutive months beginning October 2017. As of January 2020, more than 100 libraries directly depend on closer-mop, which represents more than 5% of libraries in Quicklisp.

closer-mop is a compatibility layer that smoothes out many unfortunate differences between implementations of the Metaobject Protocol (MOP) across several Common Lisp implementations. The CLOS MOP is one of Common Lisp's crown jewels, it is the single most important de-facto standard extension to Common Lisp, and it remains one of Common Lisp's most distinguishing features to this day. Were it not for Pascal's diligent work, the MOP would probably be impractical enough to use to be mostly of academic interest, whereas closer-mop enables everyone to confidently innovate using this feature almost as if it had been part of the Common Lisp standard all along.

Pascal's work deserves greater recognition.

Alexander Artemenko

Alexander is a very productive Common Lisp engineer who single-handedly liberated the Common Lisp community from Quicklisp's monthly release cycle with Ultralisp, an alternative Quicklisp dist which features instant updates without prior review needed. I still submit my libraries to Quicklisp, but personally I entirely switched from the Quicklisp dist to the Ultralisp dist.

Alexander's Common Lisp Project of the Day initiative is an interesting step towards fixing our Discoverability Problem, already featuring more than 200 projects! This is a truly outstanding effort!

Alexander is a great asset to the Common Lisp community, so please consider supporting him on Patreon!

Vincent Dardel

Vindarel is a very valuable Common Lisp Open Source contributor and advocate.