Because working for 20+ years in avionics using niche technologies and industry specific regulation and processes is a bit alienating, having few side tech projects is sometimes refreshing.
simple stupid static site generator in python, 2022~2023.
this site has moved to a minimalist and (mostly) static approach.
my oldschool demo effects experiments in C (working/in-standby)
(quick overview available in demoscene)
m3u mirroring, 2022.
I use this to mirror my iTunes playlist into my smartphone (awk+shell)
a tiny instruction-stepped MOS 6502 emulator (working/in-hibernation)
my humble C lisp interpreter (incomplete/paused)
Shopping cart using the Code Igniter framework on top of the LAMP stack, 2002~2010.
eCommerce website for niche products.
Full homemade frontend/backend multi-lingual website with customer management, shopping cart management, loyality card, secured payment system, orders shipping and tracking, with data export to accounting.
Based on the famous Linux Apache MySQL PHP (LAMP) stack.
I started with stock homemade PHP: it was somewhat pretty difficult to develop and maintain both the core "framework" and the product website itself at the same time.
Then I heard about Ruby and Ruby on Rails (RoR), but at that time, finding a decent web hosting service that support Ruby was pretty hard.
I discovered Code Igniter, a PHP framework inspired by RoR. It was good enough and I redesigned everything using Code Igniter (v2 and v3), and I finally never jumped to Ruby.
I sold the website in 2010.
Chemical production using Omron CQM1H Programable Controller, 2006
Automated chemical production control using one Omron CQM1H Programmable Controller. I used Omron CX Programmer for the logic. It has principles similar to the Grafcet language (synchronous). I had to control pumps, motors, heaters, digital displays and use input sensors to schedule the production.
They were no offline simulation at that time. I had first to live run the full logic using the CQM1H controller and check outputs switches thanks to the leds available on the output interface cards.
Then comes the first real system integration. Checking each controlled element one by one, before the first full production cycle.
The first full production cycle was near to a failure, mainly due to badly calibrated probes. Things went crazy, with smoke everywhere in the factory. Thanks to the emergency switch I was able to stop everything quickly with no big damage.
Still in production. Safe and secure!
Digital transformation of a paper-based planning process, 2005.
Done with a colleague.
It is a collaborative tool used to allocate machines to construction sites, replacing an ad-hoc paper based process. It helped to both optimize the transport of big drilling machines and collaborate for contracts signing.
Stayed in production untouched for a handful of years before one new local recruit took the handover to push evolutions/improvements.
Text-based basic stock management, 2002.
Simple stock management and control deployed on a legacy Pentium PC under FreeBSD through a text-mode interface using ncurses and an ad-hoc database (not a very good idea...)
Configure the products, the current stock, and the alarm limit before shortage for each product. Configure dependencies between elements to setup group alarms.
Dashboard showing current stock and allowing to add/remove stock on a per product basis (mainly when you receive raw material) or per group (mainly at the end of the manufacturing process).
Stayed in production untouched for at least 15 years.