I made lots of mistakes and in hindsight, put and tried to play an humongous amount of games following my own tournament format that bite me to this day.
This was just one of a few very big mistakes, I now think that is better to run a small number of games and found my self on the other extreme... growing the bot to play bo3 focus only on one (2), (3) and (4) player maps. I'm in between a code refactor moving bot to a entity-component-system architecture that play nice with a simulator that I have running Luerl (Lua on the Erlang VM) this is one of my main project goals and objectives this integration of different versions and implementations of the Lua language in tandem so instead of fixing the container tournament manager that explode on my face as reported before or finishing the actual tournament as planned, is not that I abandoned torch-up, if anything I want to approach things differently, I still plan to run games with the registered by now old bots following the previous stated rules, a first round and announce results I believe that as my goal with torch-up, that in order to move on to future things without old mistakes hunting me.. I'm a sloth on hell already, I don't have anything new to report around tournament results, other than something completely different, and probably not what you are wondering, like:
Originally I had this crazy idea of a future where bots could play the old proleague format including 2v2 games, all wishful thinking, since after some real world experience as 2v2 player and a lite dive into bwheadless.exe or C++ for that matter I learned the hard way how crazy my intention was.
I'm not going to continue with the Python tutorial and no matter that I can code some lite examples with TorchCraft's Python API the idea to build also a Terran bot, a race that I don't have any real experience with for a future 2v2 combo with my Zerg bot its completely out of my capacity, I wanted to get the attention of existing Python devs that could want to build a bot using Python, I hope that someone write a great bw bot with Python eventually I bet a scary bot could be made just with Python alone, but I'm not going to try to be that champion, I have my plate full with Luerl and I'm making a really bad job about it, I'm going to return to write about the Python side of things, for anything PyTorch running on another process talking TCP with the rest of the bot system as it should instead of using Python to write a second bot, Blueberry is dead long live Ophelia.
Python is slow, people say that to keep the meme alive, the reality is that is fast enough, the speed of Python have never been an issue for writing a StarCraft bot, deer reader take a look at the StarCraft 2 AI scene for examples.
Q. Is torch-up vaporware? is it ever going to happen?
A. I guess is vaporware until is not?
Q. What is my problem? ^^
A. StarCraft: Brood War.
Q.What happened? is there any update on this?
A. My goal still is built, compete and write about my bot and the system of systems core of the fixing of the StarCraft: Brood War problem instead of a main focus as tournament organizer, but I also want to compete in a world where you are not forced to use Windows because the game release originally for the 1998 version of that OS. So the idea of become an event organizer, sadly apparently the worst at tournament organizing is really not that bad, mistakes where made, my apologies to all involved that want to see this finally out of the deeps of hell or all of you competitive beautiful people that wanna a piece of my bot, until then, this will hunt me until I hunt it.