Steam Deck, Linux, Github Copilot and more!
This is my first time writing a blog so expect some rough edges.
So the new steam deck got announced.
It’s basically a gaming console made by steam. I’ll leave a link in case you want to read more about it but I’m not gonna be talking about the hardware of the steam deck. I’m more excited for the implications it has on linux gaming.
Why? You ask me. Let me tell ya.
Because steam chose Arch linux as a base for Steam OS 3.0 which is absolutely brilliant!! Steam choosing linux to be running their game console means that game devs will be pushed to support linux or at the very least support proton (a windows compatibility layer for linux) on their games. Valve also announced that they are working with Battleye and EAC, the two biggest hindrances to linux gaming, in bringing proton support to their anti-cheat software. So finally I can play Apex or Rainbow Six without having to boot into windows.
And I’m so very stoked for it, because it brings us to our next topic,
I switched to linux this week!
No dual boot, no Virtual machines, just plain old Manjaro on bare metal. This isn’t my first time on linux, I’ve always had linux on dual-boot along with windows so I can play games or use the Adobe software suite. But having just linux feels liberating in a weird way.
So far it has been very fun. There were some challenges getting things working the way I wanted to but that’s why I love using linux so much, there is always a way to fix things and there is always someone there to tell you how to fix it. It’s been a blast just tinkering with it and fixing things.
(On to the nerdy part)
Coding on linux hasn’t been very different than on windows other than software dependencies resolving themselves automagically (most of the time).
Ruby on windows is a nightmare, and getting this website which you are reading right now was so difficult on windows. I had to jump through so many hoops and had to go through dependency hell to just get the website running with the layout I wanted. It was waaaay easier on linux, working with ruby and jekyll on linux feels fun, not frustrating (80% of the time)
I’ve done a lot of things on linux since I installed it, but writing them all here would just bog this post down. Maybe I’ll write another blog talking about my linux adventures.
On the topic of coding, I discovered,
Github Copilot
Github Copilot is an AI based autocomplete for your code. You can do crazy things with it like write a comment saying what you want the code to do and it writes the code for you. It also works as an easy template bot that self inserts long snippets of code after you write just the start of it resulting in more productivity and a lot less grunt work of writing the same functions again and again for projects.
Github Copilot was apparently trained on the publicly submitted code on all of Github, which raises a few questions on plagiarism and/or indirect plagiarism. But Microsoft can always throw money at the problem like always.
I, personally, am excited to get a hands on with GIthub Copilot. There was another AI based software that I found out this week, which is way more malicious than Github Copilot,
AI Cheating Software
So how this cheating software works is that it looks at the game via a capture card and then inputs mouse movements via a usb device.
How normal cheats worked is that they accessed game files to find out the location of enemies in the game, and snapped the crosshair to their head to kill them. This cheat bypasses that need to access game files.
So how you run it is, you run the cheating software on a PC different from the pc you are playing the game on. You send the “game PC’s” hdmi signal via a capture card to the “Cheat PC” and the cheat PC then does its AI magic and sends mouse signals to the game PC via another USB device.
This is crazy, even though it requires two computers and dedicated hardware to work, considering that cheaters have been buying $300-$400 cheats already, this feels very plausible for them to use and get away undetected.
That’s it for today! New blogpost next week, same bad day, same bad website.