mutithreaded freedom
26 Nov 2010i wrote this article almost 5 years ago for kerneltrap. at that time i titled it "part I" but never wrote a second part.
Some thoughts on the Hurd and community plumbing, part I
While browsing throught the l4-hurd mailinglist archive some days ago i discovered some interesting thread, which seemed quite unusal: In a mail titled "The current hurd development system - or: optimizing for fun" from january 2006 Tom Bachmann asked some interesting questions about the developement of the GNU/Hurd Operating System.
Basically Tom Backmann stresses that there is a vicious circle in GNU/Hurd developement: There is a lack of coders and no new coders join the project because their is not really any community where these coders could come from. On the other hand as long as developing is very slow their won't be a growing number of users.
Let's have a closer look at myself as a typical example: I am what some would people call a geek, although my tech knowledge limited, but computers and internet are part of my lifestysle and biography, from the homecomputer in the 80s, windows in the early n90s and later on Linux , basically Debian GNU. I was interested in GNU/Hurd since i read about it some years ago and felt drawn to the project, also first of all because i feel drawn to the free software philosophy. also there might be some affection towards GNU/Hurd because I simply like to have some alternative to the overwhelming sucess of Linux and it's "commercialisation" in many ways.
I will never have the time to learn enought C to code on the hurd directly (i am basically a web developer), neither to debug or port some applications. What i can do is try the hurd for everyday use (and there is actually something where it can be used) and report experiences back from the view of a normal everyday user. i also can contribute troubeshooting to the project, i.e. all the little stupid thinks and workarounds you have to know to run a computer system especially when it is under developement.
I guess my distinction of people into users and developers is wrong anyway, because everybody can contribute to a project: manuals have to be written, they have to be written many times: There is nothing like THE manual for a problem, there have to be many manuals for one problem, each written in its own language, style, knowledge level and using its own methaphorism. Like some (if i remember right) hindu saying says, there is not one book explaining the world, there have to be billions, so everybody can find his/her own to understand.
This is normally the way communities are build, the wonderful thing about (as an example) the debian community is that if i am running an apache server on debian and want to know some detailed stuff on setting up vhosts people in debianforum.de will help me, althought the problem is definitely not specific to debian or even to linux. They help me because its a way of sharing everyday problems in a community which is glued together on some common basis (debian in that case). if i am in a football club i will talk to people about other things than football and nobody would say "this is not football related" if i ask him how to fix my computer. thats how social community work, and that's great. The lack such of a GNU/Hurd community suprises if you consider that Marcus Brinkmann decribed the aims of the project by "putting an emphasis on the user's freedom"(source). (A statment which i really gave me a kick to have a look again at the Hurd last year)
Browsing the net there is no such thing for GNU/Hurd users, not even the GNU project website seems to have any kind of social community plumbing. You could answer that there are no GNU/Hurd users, but you are wrong. I am one and i meet others in the vast emptiness of cyberspace from time to time. These people seemed lonely to me because on one side the linux community makes fun of them and on the other side many GNU/Hurd developer seem to give them the impression that they are not welcome until they can write some hurd translator. That's actually okay from the developer side, I do not want to claim that developers should care about stupid beginners questions. My idea is more to introsuce more of a DIY approach for GNU/Hurd users. Maybe, if I find some time in the next weeks, i'll start some lonely forum projects for something like that. you never know what happens.
original: http://kerneltrap.org/node/6192