I Really want to make a distro and i want to know, how would i add a GUI? what (other than C and Assembly) should i learn? Should i take classes? How would i incorporate audio and hardware compatibility? This all hurts my head to think about as i love linux and i want to learn more and more but im too eager to make things. (LAST 3 QUESTIONS) How can i make this distro compatible with older/crappier machines? Should i Build from the Kernel or another distro? (Red Hat, Fedora, Debian ETC.)LAST: What filetypes should i use and how?
thinking of making a distro, how would i go about this?
- Thread starter reonized
- Start date
Perhaps have a look at this thread:
Most one man projects don't last long, you would contribute more to Linux if you joined and existing project and gain experience there so that if you at one point still wanted to create your own Linux distribution you can then use the experience you gained from helping with the other project you helped work on.
Last edited:
Depending on your reasons for wanting to create your own distro, it might be a lot of fun, a satisfying long term project, an exercise in futility or a brief but satisfying ego trip.
Reply #3, above, by f33dm3bits was spot on advice -unless- you are in it primarily for the learning experience -and/or- you have some specific and well thought out goal in mind that is not served by some existing distribution.
I tried remastering my favorite distro (microcore, which is designed to make remastering easy) a few years back with the thought of that being a great learning experience and with the not-so-well-thought-out goal of making it even smaller and simpler by eliminating a couple of features that I never use and possibly adding some (nebublous) features that might be useful, if only to me.
From the "learning experience" point of view, the project was a brilliant success. I learned some valuable skills but I also learned that the small team who produces the distro I was starting from -really- know what they are doing! I learned that trimming out the features that I had in mind to eliminate would have made virtually no difference in the size or complexity of the final product. I learned that virtually every feature that I would have added was either already present (my bad for not knowing that ahead of time) or was actually not appropriate for inclusion in the base system (my whole project revolved around only the base OS, not the various user applications and desktop environments which, IMHO, are not to be considered part of the OS even though, in most cases, they are part of the "distribution").
So for all my effort (which wasn't really all that much), I ended up with a "distro" that had the name I gave it but was almost completely the same as the distro on which it was based. Then I realized I would have to maintain it, either by keeping it up to date myself or by re-applying my changes every time the upstream distro was updated. And there was no reason whatsoever for anyone besides myself to want to use my distro. That's where the "brief but satisfying ego trip" ended.
I have neither the dedication to maintain a distro nor the imagination to come up with something that would fit into the fundamentals of the OS and differentiate it enough to make worth a fork.
So the take-aways:
You'll need the underlying graphical system, like Xorg (and all of its dependencies, of course), a window manager or DE (and all of its dependencies) and possibly a bunch of graphical utilities. Don't be too concerned at the get-go about having a graphical tool for every little setting - because, where your going, your probably not going to be too concerned about (gasp!) editing a few config files with your favorite text editor. That was a gross oversimplification.
Reply #3, above, by f33dm3bits was spot on advice -unless- you are in it primarily for the learning experience -and/or- you have some specific and well thought out goal in mind that is not served by some existing distribution.
I tried remastering my favorite distro (microcore, which is designed to make remastering easy) a few years back with the thought of that being a great learning experience and with the not-so-well-thought-out goal of making it even smaller and simpler by eliminating a couple of features that I never use and possibly adding some (nebublous) features that might be useful, if only to me.
From the "learning experience" point of view, the project was a brilliant success. I learned some valuable skills but I also learned that the small team who produces the distro I was starting from -really- know what they are doing! I learned that trimming out the features that I had in mind to eliminate would have made virtually no difference in the size or complexity of the final product. I learned that virtually every feature that I would have added was either already present (my bad for not knowing that ahead of time) or was actually not appropriate for inclusion in the base system (my whole project revolved around only the base OS, not the various user applications and desktop environments which, IMHO, are not to be considered part of the OS even though, in most cases, they are part of the "distribution").
So for all my effort (which wasn't really all that much), I ended up with a "distro" that had the name I gave it but was almost completely the same as the distro on which it was based. Then I realized I would have to maintain it, either by keeping it up to date myself or by re-applying my changes every time the upstream distro was updated. And there was no reason whatsoever for anyone besides myself to want to use my distro. That's where the "brief but satisfying ego trip" ended.
I have neither the dedication to maintain a distro nor the imagination to come up with something that would fit into the fundamentals of the OS and differentiate it enough to make worth a fork.So the take-aways:
- 1) The "anti-Nike" ( Eekyne (tm) ) motto: "Just don't"
- 2) Or, just do it anyway - it's fun and educational
- 3) If you want your distro to be more than a "flash in the pan"...
- a) Have well thought out goals
- b) Before learning C or assembly (which are great in their own right), learn team building
- c) Revisit your goals with your team and always -focus- on the goals
- d) Get advice from someone besides me because the anecdote above is the extent of my experience with such things.
You'll need the underlying graphical system, like Xorg (and all of its dependencies, of course), a window manager or DE (and all of its dependencies) and possibly a bunch of graphical utilities. Don't be too concerned at the get-go about having a graphical tool for every little setting - because, where your going, your probably not going to be too concerned about (gasp!) editing a few config files with your favorite text editor. That was a gross oversimplification.
arochester
Well-Known Member
This is something which occasionally crops up. "I want to create a distro, but I have no I have idea how to do it, what should I do".
Is there a course somewhere in the world where they say "Create a distro"?
Is there a course somewhere in the world where they say "Create a distro"?
Last edited:
This is going to sound rough, but if you don't know how to make a distro then making a distro might just not be for you. It's complicated. It's tedious and thankless work. You can make a distro based on something else with what I'll term 'relative ease' - but that's REALLY making light of the situation.
From your post, I'm under the impression that you're not even all that familiar with Linux at this time.
Spend a few years learning Linux (and not learning a distro) and that'll suit you better than any other suggestions I could give you.
From your post, I'm under the impression that you're not even all that familiar with Linux at this time.
Spend a few years learning Linux (and not learning a distro) and that'll suit you better than any other suggestions I could give you.
