In this post, I'll take a look at first converting the [existing weighted model](https://congx.dev/posts/multiobjective-optimisation-with-ibm-ilog-cplex-1/#much-faster-model-with-weights-cplex-optimizer) in part 1 to Python, proceed with adding additional weights to the model and finally deploy the model to run on IBM Cloud with Watson Machine Learning.
Eons ago, I played this online game called Tribal Wars 2. Recently, Innogames wiped their online forums to its pre-historic ages, which happens to include my report on army sizes to build. Back then, I used MATLAB Optimization Toolbox. Now I'm going to give IBM ILOG CPLEX a try and see how it goes.
I really like [Panr's Hugo theme Terminal](https://github.com/panr/hugo-theme-terminal). It's beautiful. For long texts, the contrast can make texts a bit hard to read however. This post outlines the files I made customisations to, to create a light theme with a dark blue accent. This is in no way a complete customisation of the Terminal theme, just enough of it that fits my current needs.
There are 4 types in XPSL: user, room, track, playlist. This blog post lists the functional requirements for each of these types.
Some notes on how I set up Travis-CI to build and ship content to another Github Pages repository
A closer look at how Tribal Wars 2 client and server communicates
LogDNA on IBM Cloud could be better
Tribal Wars 2 Manager is now called Toddler!