Today I decided to start an interesting project that I hope will turn out to be quite an adventure for those seeking to learn about Linux and Git, and for those that want to work on projects themselves, and to just spur open source development in general.
Right now, it is simply a kernel that I want the community to decide the overall direction of based upon feedback and stuff like that. That said, here is the basic framework that I am adhering to:
The kernel is built on top of Linux-CK from where I merged Torvalds latest commits on his GitHub. The reason I did this is because I believe 4.10 will be the next LTS version of the kernel and we are too close to that milestone to focus on 4.4.
CFS is getting pretty old so I think it is time we use MuQSS (based on BFS) as it is much improved and works better for daily desktop use.
BFQ is pretty awesome on desktops and has a lot of potential for multimedia workloads
That's pretty much it. As for the direction of the kernel and how you can help, simply provide feedback or open a PR on Git.
Moving on to other things besides the Linux kernel (my main focus), I want content creators and people in general to contribute their works on GIt. Rather it be simple scripts, embedded development, programs, games, etc... I feel that you need the chance to express yourself. Simply PM me if you would like to contribute or host a project and I will add you in time
And one last thing... Do not by any means feel intimidated. This is a learning environment. I am no pro nor are several other developers. I too am learning. I want people to innovate and to work on new things never thought of before, and I want developers here to collaborate.
I think this should be a heavily optimized kernel for workstation and gaming PCs.
I think most of the community is pretty enthusiastic about high end computing, so removing things like radio functionality and low power cpu options would be a step in the right direction.
MuQSS is definitely a good idea. Any optimization that would improve EXT4, F2FS, and even XFS would be pretty nice.
And while I am talking about it, I think NTFS drivers should be baked into the kernel if possible. A lot of new users are dual booting, and it would be cool if they did not have to fuss with NTFS drivers.
@Zippy_Parmesian This is a great idea. I've been getting into Linux more and more recently, I tried my hand at it a year or so ago but with the new hardware I had and frustrations I decided to go back to Windows for the simple functionality. Learning git is definitely something I need to do as I am a CompSci major interesting in computer programming mainly so it is something I'm going to need.
@Tjj226_Angel This is definitely the direction I would want the kernel and the distro to go in general. Finding a distro that works well with top end hardware without some excessive preconfigurations is difficult.
I've been a part of this community for a while, found out about Tek syndicate years ago through a friend and have been watching since the early days if the Tek. I've had multiple accounts on the website but I haven't been too involved in the community. I just wanted to say that ever since the split I've been seeing more and more topics just like this and I'm really excited for the future of this community. You guys are really inspiring!
We should NOT use github. I personally use it for some projects, but I believe we as a community can better benefit from the mostly FOSS gitlab, which is self hostable, and "premium" features that Github offers (such as private repos) are free with Gitlab on their own hosted instance.
As for the custom linux kernel, I don't have a opinion.
I'll look into it. I think it might be a good idea to indulge in both platforms as they are both very powerful with their own benefits. I have used GitLab in the past so i might as well create a repo there too
I'd recommend using some of the PaX and grsecurity patchsets. Why? They do almost nothing to upset user experience(you just have to have a properly configured paxd.conf or set all the flags, easy as pie), and they do provide lots of mitigation.
If you do use the grsecurity patchsets, though, you'll want a beastly build server -- they update every few weeks.
oh right, I forgot about that. I've just been using their testing version(with no problems, but still)
Alpine is... really small. Really not a good choice for a desktop OS. I don't think it can even play steam games, since it has a different ABI from the one that most steam games are probably built with. That and its repositories leave much to be desired.
I haven't been on top of it as of late so I don't believe so yet. Most of what I have done up to this point is just optimization for performance and architectural optimization as well. Also added in Exfat support too. We have a long ways to go from this point. I'm taking input from everyone here and working it all into one.
I appreciate everyone's recommendations and insight. I will definitely get around to doing a lot more with the project this weekend.
That would be pretty awesome actually. The issue is it would require a lot of work on my part to bring it up to spec. I'm trying to find others interested in this as well so when I do, that is a definite consideration. Right now, most of the stuff I've been adding is from Zen and Liqourix. I've also heard a lot of good things about the PF patchset as well.
@Zippy_Parmesian how about we do some other smaller stuff instead of just trying to do a distro? such as the example script i posted in the linuxfirewalld thread
being stuff like scripts, package lists/recommendations, tutorials, tips/tricks, examples etc. for stuff like.. making a swap file or container/volume, mounting bind options so you can have /home/username/music be the same folder as c:/users/username/documents/music on a windows partition, wildcards piping etc
edit: dont mean instead of, as in shouldnt do a distro if you want, but supplementally
Oh no, no, no. This is the kernel itself. Not a distribution. Other projects are welcome and even encouraged. I do plan to write a few guides too all in due time. I know I put emphasis on Linux but other projects I feel will allow the community to thrive.
That said though.... A distro would be an interesting project
I'll definitely be posting a few scripts that I use for automated kernel compilation and updating GCC. Stuff like that, and how to make those scripts commands recognized by bash.
Maybe I'll do a GitLab overview as suggested by one of the users here
I was asking permission to put the project with those tweaks on OBS. I prefer it over GitHub. And making the kernel prepackaged it would help newbeans.
@Zippy_Parmesian do you have an issue with me packaging your Kernel on OBS?
Its been a while since I updated this thread lately and here's why:
I am a ROM maintainer for CandyRoms and recently stuff started falling apart and everyone felt opposed to one another so people started to leave the project. One and two of my other friends didn't want to part ways and we didn't want to see Candy die because it was fun to maintain so I took over development of a new ROM I coined Zippy and I've been working on it for a bit trying to get 7.1 up to speed. So of there is a delay in Linux kernel development on my part that would be why. I'm hoping to get back to it in a week or two when I'm done with my finals and hopefully my ROM bring up. Its based on Slim and so far I've just been messing with sdclang and Uber so we'll see how that goes
Any ideas any one has for future projects or what they would like to see in a custom ROM in general?