[manjaro-dev] Manjaro Driver Support
philm at manjaro.org
Sat Oct 3 09:00:03 CEST 2015
let me answer your e-mail by answering in sections ...
> I was hoping to get a little insight from the manjaro-dev expertise. First
> of all, Manjaro is great! When I was rounding up every Linux distro I could
> get my hands on a few months back, Manjaro was one that really stood out
> and impressed me. (Up there with Zorin & Fedora, IMO.)
Great to hear that you like us so far. Zorin and Fedora are also great
projects. It is fantastic to be compared with them. As there are so many
distributions out there, we thought lets do it simple for the customer
but be as much flexible as possible.
> I know you all do hard-ware/driver detection/installation on top of the
> Arch base.
Yes, Arch Linux is great. Fast, flexible but not so easy to install
if you don't know how. A key point was also the hardware detection.
> I, too, am doing a distribution from Arch. Is there any kind of advice you
> would give? There ought to be a place for Distro maintainers and start-ups
> to share knowledge, if there isn't. (How did you all know where to begin?)
Fantastic. What is your idea and why do we need another distribution
based on Arch? My advice. Arch is to fast in adding new packages. You
have to find a way to slow that speed down. Take a look at Antergos
and how often they have to update their packages and stuff. With Chakra
Linux I used the PKGBUILD files and compiled them my own. With time I
had so many packages to compile and so less time to do so that I started
to think about something else. Chakra is still there, but I switched
over to a new project, which was called Manjaro.
Most people start with the book 'Linux from Scratch' to get into the
whole thing. Other take their favourite distro and try to do their
respin. What ever you do. You can either try it on your own and join
several IRC chats and forums or you ask a running project to join.
> There are a lot of major aspects to this project, but the one that actually
> has to do with system is how the software upgrades work. (Oh no, not
> again!) Instead of a package manager, to work with the way the system is
> set up, users will just install preconfigured FS ROMs. (These will be like
> individual "service-packs" for certain occasions. --like an education ROM,
> a graphics ROM, a gaming ROM, etc.) More advanced users could have the
> ability to generate & upload their own ROMs.
I assume you talk about your own project now. Since the idea of
pre-configured, you call them ROMs, we already have developed and tested
in our internal project for desertbit which we might launch mid 2016
based on our hardware we might sell. We have a base layer with all
needed parts like libraries and core drivers like the kernel and on top
you get those which are for office, DE or extra layers you need for
games. Also you can try different layers when they come out and go back
in time if those doesn't play out for you. So what ever you try to do it
might be already done. But nice idea.
> So, in that sense, it's in the same line as distros that are just selective
> pre-configured environments and distros that offer spins. (This is to make
> my task easier.)
Which line do you mean? Manjaro is totally flexible. Our install medias
are just example how we think Manjaro should look like and function.
With 'manjaro-tools' you can configure, theme and create your own
distro. Many other projects already use that tool. Netrunner-Rolling
and Sonar Linux are those big players at the moment.
> As far as drivers, I need them as expected. Would it be as simple as
> including everything for all platforms by default in the stock OS?
Free drivers you can add all together. This way everybody is doing it.
Manjaro is the only distro I know of, who is installing the needed
driver on bootup for live-session and during the system installation to
the target. Sure having all drivers installed might get your distro run
on any other PC by just switching the HDD out, but we have a simple
routine which detects your hardware and installs the right driver just
> Any other advice?
Think before you start just another project - what you might try to
achieve might be already out there.
Manjaro Project Lead
More information about the manjaro-dev