[openbeos] Re: distropia
- From: Adi Oanca <adioanca@xxxxxxxxxxxxxx>
- To: openbeos@xxxxxxxxxxxxx
- Date: Fri, 04 Feb 2005 11:26:14 +0200
Hi all,
Apparently I am not the only one thinking distributions are bad...
Anyway, I think my opinion matters just a bit so it's up to Michael
and team leaders to decide what should be done in this area. I'm waiting for
feedback from you guys.
I would want us to be the *only* ones allowed to make and distribute
the whole Haiku-OS build. *No* other can take that base and ship or the Haiku-OS
base, improve, add and then sell. *HOWEVER* ANYONE is invited to take
*everything*
they want from our CVS tree, *as long as* they don't take more than 50% of the
base
modules (put together), which are: kernel, media_kit, app_server,
application_kit,
interface_kit, networking, BFS.
I think in this way we can have a consistent OS, people would know where
to come to get a free Personal version or buy a Pro version. You would get
the same consistent OS everywhere: one FileManager, same install dirs, same
preference
apps, one window manager, one audio system. This way *we* will be the ones who
decide where our little project goes. This way, we would directly see the
revenue
made from shipping distributions, not just seeing bits of dollars/euros coming
from
a distribution maker/seller.
I'm glad I'm working at Haiku Inc., I am glad I'm doing something for
others,
but I REALLY *DON'T* want others to just take our 4-5 years effort, sell it,
improve
and then make lots of money out of it while pushing Haiku-OS.org in the back
scene.
Regards,
Adi.
Tim Timson wrote:
Hi, I'm a short-time list-lurker, but I wanted to throw in my 2 cents.
Having distropia can be both a blessing and a curse. For example, the
various distros in GNU/Linux fill various niches. Want a
(l)user-friendly desktop (GNU/)Linux? Xandros. Want a linux with
networking support that fits on 2 bootfloppies? Blue Flops. Want to
have commercial support? Red Hat. Want a cluster-capable OS that boots
from cd-rom? Knoppix with openMOSIX. Etc. You got the choice.
But the new user is also clueless about which one to try, where to get
the best distro for him. And it annoys me to no end to see software
packages for download for different types of distros. I don't know
enough of Linux to be sure one of the 5 available packages work with
the distro that isn't listed but the one I'm using.
But having a company such as Redhat backing Linux has done a lot to
further Linux' cause. Linux is now a multi-billion industry with a lot
of professionals working with it, and corporate backing ensuring basic
driver support will be a constant.
But starting a new distro is, like starting a new political party,
really usually done out of discontent with what is available and/or
how things are run. NetBSD for example was done because people
disagreed with how development of FreeBSD was done. But OpenBSD was
forked off NetBSD because a group of people had different goals in
mind, and thus wanted to make decisions differently.
I disagree with putting out only the most basic of distro's, because
then it's waiting for several different distro's to simultaneously
spring up, splintering resources, the market and interest; simply
because with a most basic of distro distroitis is a requirement.
If the goal is to convert the most people possible to a superior OS,
then the OS distro needs to be superior to begin with, but leave room
for improvements via other people's distros. You don't want to turn
away companies willing to support your product for you (IMHO). But too
much distropia also detriments the user-experience. Ideally,
everything could be done with 'basic Haiku', leaving room for
seperately sold software-packages(because that's what distro's really
are basically) to improve the lot.
If someone wants DVD-support, Partition Magic and Realplayer (lol) out
of the box, then there's not much option other than allowing a
commercial distro. For everything else..
As long as Linux occasionally requires kernel-recompiles for
driver-support, has 5+ different
graphical shells of which the worst ones are the most succesfull, and
as long as Windows is being designed by Microsoft, Haiku will have
plenty of place on the desktop-market.
Sorry if I rambled on,
Ronald Vos
I feel the next logical step is to also be the *one* distributor for the OS too. I realise there is nothing to stop others making distros, but with a good distro available from the "heart of the project" as it were, most others would probably consider making distros not worth the hassle. I point to the fact that it is only since the death of Be Inc that the whole concept of a 3rd-party "distro" of BeOS has appeared.
Distro makers can/should add other apps, art work,
sounds, whatever, but keep the base, core distribution intact.
The only reasonable course of action is to take the lead,
and keep the lead. Presence, perfection, predation.
As long as we can, we'll definitely try :-)
OTOH if there really is a better substitute of us out there, it
wouldn't be for the bad of all.
(I wish I was in a better position, contribution-wise, to speak
my mind on these issues, but nevertheless I believe this is true.)
It would be naive to think that distromakers merely
want to bundle a little this and that, or change some
esthetic aspect of Haiku.
Right. We need to stay in public, so that users will come to us or
"friendly" distributions first.
Thinking ahead, what are likely reasons to fork Haiku?
Can those reasons be preempted?
(I'm not looking for a big discussion, just posing a question.)
In any case, I hope the scenario will be more like that of
the BSDs, where low level as well as high level code
flow between the projects.
Consider how Linux has benefitted from polish and a penguin. The
(superior?) FreeBSD collects dust in comparison. Two essentially
identical platforms!
Nobody can predict what will happen with our little project :-)
Bye,
Axel.
Other related posts:
