Commute talk: How I learned OS development

[@Walkietalkye]

There's a lot of value in implementing your own solution without looking at a "proper" way of doing it. Then, when you actually look at the "proper" way of doing it, you have a better understanding of why it's done that way. For people starting out or in a rut, don't be afraid of doing "complicated" stuff. You'll learn from your mistakes. If you're not challenged you're probably in a comfortable spot and doing something you're familiar with.

[@_yannis2707]

Implementing something your way without looking for the proper or "best" way to do it is really nice. I often got stuck for hours searching for the "best" solution, it's just depressing.

[@EvineDev]

Hey, What are your thoughts on TempleOS and Holy C? Particularly the C shell scripting environment. Personally I think the C shell is really cool, giving shell scripting some more rigor. I'd love to see it attempted in a more traditional operating system. Has Serenity taken any inspiration from TempleOS?

[@_yannis2707]

This was really interesting and helpful. I have been programming from time to time for a few years and I now feel like I'm just getting started. I love that there's so much to learn.

[@Kenbomp]

Very nice yes had similar experience. Was created simple GUI in lisp env at work for coworkers and it just started a dive in windows and popups and seeing the big picture of the users and system. When I got a great cheer by team when they see it save them 20 minutes of tedious work in 1 click they were amazed and it set off a spark ever since.

[@cstormtech]

I've watched a few of your videos, and as a programming newbie I find them helpful and inspiring. Eventually I want to be able to work on big projects myself (like an operating system!), and I'm taking small steps to get there. Hearing what you say about "complicated" stuff is definitely a confidence booster.

[@AnonYmous-spyonmepls]

Its not that things are complicated the only problem is that there are so many of those things that it takes a lot of time to learn everything. You just need an optimal workflow and you will get there. Always try and jump into subjects you don’t have a clue about, the less you know about something the more you learn while you expose yourself to it.

[@Kniffel101]

Hey Andreas, I've got a question for you. From what I understood you're trying to implement a really fast and responsive OS with Serenity, and I really love that! Though what struck me and confused me a little bit was that you seem to have very OOP-heavy codebases instead of approaching it more in a data-oriented design approach. Why's that? I mean if performance was a main focus, OOP wouldn't be the first choice, right? Because of all the abstractions between the CPU/Memory and the code. And from my experience DOD even results in more readable i.e. more maintainable code than OOP. If you don't know what exactly I mean by this question, I'd like to refer you to the following three videos. I'd love to get your opinion on that! :) The first is an abstract someone uploaded from Casey Muratori's "Handmade Hero" series' Q&A: kzbin.info/www/bejne/kH6Ui3eXlqqsg7s The second is the talk from Mike Acton he refers to at the end: kzbin.info/www/bejne/qImTeqeMerudfsU The third is a discussion of the two at HandmadeCon 2015: kzbin.info/www/bejne/p4itoXxolsl2mdU

[@MaxThrun]

What’s your opinion on micro kernel architecture?

[@MindGameArcade]

Damn you changed side since this video!

[@KIXEMITNOG]

200years

[@michaelnolan1715]

should I learn the same way (as a hobbiest)

[@__gadonk__]

well i'm pretty much a newbie at this point but i feel like actually reading the hardware documentation gives you a really good understanding for implementing your software. I kinda started the worst possible way because i went straight into writing the "kernel" before even knowing how the hardware works on a low level (i literally only knew how to write basic Linux shell code in assembly ). But as time moved on i started to understand more and more about the hardware just by watching some Computer Organization lectures on youtube and playing around with device drivers, that run on realmode. I still know nothing about Computers but at least i now know where to look when i want to research about something new i want to implement inside my "kernel". Implementing an Emulator or similar that forces you to know the architecture inside out is a really good if not the best way. Just get as much information as you can about your target hardware or Computer Concepts in general. i hope this was helpful... as i said i don't know much (i'm still 16 and in School so i have more time on my hands than an Adult) but i research as much as i can. I really hope you succeed in OSdev and wish you luck, my friend! :)

[@__gadonk__]

I also recommend you to find a way to motivate yourself (i know that sounds kinda obvious but stick with me) Everyone has a different way of finding something that motivates one self. Maybe you want to implement a shell interface as fast as possible or you find interest in device drivers or similar. It doesn't matter. Andreas mentions in another video that you should just jump to the thing, that interests you the most and try to abstract things away in some sense so the actual implementation on your system doesn't involve a complete rewrite.

[@uonliaquat7957]

@@__gadonk__ I'm amazed to know that you're just 16 and you're already doing great. I'm a Systems Programmer and a Computer Scientist. Would love to connect with you.