[openbeos] Re: Haiku distribution
- From: Simon Taylor <simontaylor1@xxxxxxxxxxxx>
- To: <openbeos@xxxxxxxxxxxxx>
- Date: Fri, 13 Apr 2007 9:40:29 +0000
>
> From: Sogabe <sogabe@xxxxxxxxx>
>
> Ingo Weinhold wrote:
> >> What is after all the common (official?) position on Pingwinek Haiku and
> >> third party distros in general?
> >>
> >
> > We're currently preparing to take the final vote on the matter. Now it's
> > hopefully really only a few more days. :-)
> >
>
> One more Haiku distro coming. ;)
>
> http://osnews.com/comment.php?news_id=17686
So if the final vote hasn't happened yet there might still be time for me to
have a little influence on it. I genuinely believe this decision will be the
most important one Haiku ever makes in determining whether or not Haiku is
successful.
I was an almost full-time BeOS user a few years ago, but that was simply the
reason I found out about Haiku and is not the reason I continue to follow the
progress of the project. The reason is because I believe Haiku has a huge
chance to become "The Open Source OS, done right."
One of my main reasons for disliking the Linux approach is the lack of an
overall consistent plan for the whole system. Haiku has neatly sidestepped the
problem by sticking to the API which Be did a very nice job with all those
years ago. It just makes the whole thing feel as though it was designed
together.
Another, related, issue facing Linux as a desktop platform is that all the
components are developed independently with no coordination of priorities,
roadmaps or releases. Haiku is a huge step forward with all the components
being developed in the same tree to the same schedule and for the same users -
even things borrowed from other OSS projects are imported into the tree so
their versions are also fixed with a Haiku release. That is a *huge* advantage
for Haiku, for both users and developers (no need to target millions of
possible configurations).
That's two big reasons why Haiku just makes so much more sense as a proposition
for a desktop OS. The next problem is how to communicate those advantages to
users. I would argue that a free distribution policy does nothing but confuse
the issue for users and implies that Haiku will suffer from the same problems
as Linux with regard to versioning. Imagine the difference in the build up to
R1:
Feb: "Haiku 0.9 beta released"
March: "Haiku release first RC for R1"
April: "Anticipation builds for Haiku R1"
June: "Haiku Reaches R1, The OS Desktop Comes of Age!"
vs
Feb: "Official Haiku.org sources reach 0.9 beta"
Feb: "GNU/Haiku 0.9.9.9.4 released"
Feb: "Hi-ku 1.4 Released, still full of bugs"
etc, etc...
Encouraging 3rd party distributions is usually defended on the basis of
providing users with more choice. Really all that amounts to is installing a
few 3rd party applications. The R5 distributions (MAX etc) had two other
reasons for their existence - 1) R5 stopped booting on newer hardware and 2) R5
was missing a lot of useful drivers that were developed after the release by
3rd parties. Haiku will not suffer from either of those issues. I very, very,
very (as if you couldn't tell by now) strongly believe that the additional
benefit of having a few 3rd party applications pre-installed in no way
outweighs the disadvantages caused by muddying the waters around the "unified
desktop platform" message that is, for me, Haiku's strongest selling point.
A couple of weeks ago I suggested having a base R1 and allowing third parties
to produce Application Bundles to be installed over the top. I also suggested
allowing the Bundle producers to distribute an ISO with their bundle
pre-installed. To me that gives the advantage of user choice and pre-installed
apps, but makes it clear that Haiku is the core OS independent of the apps
included in your particular download.
I really wish I could write short, persuasive emails. Sorry about that. The
most important bit that I'm interested in hearing replies about is why anyone
would favour a free Linux-like distribution policy over my App Bundle
suggestion.
I realise being MIT licensed anyone can do what they want with the code, but
Haiku Inc do have some control over the use of the Haiku name - and the
position of the Haiku admin team will directly affect whether shoot-off distro
projects are seen as unofficial experimentation (which I have no problem with,
as long as they are viewed as such) or whether they are seen as officially
sanctioned distributions of the OS. It's all about perception.
And for those who think all this talk of selling points and perception has no
place in OSS - the simple argument is a well-marketed release brings more
users, more developers, more apps, and more commercial interest - which is
definitely a good thing for the platform in the long term.
Simon
-----------------------------------------
Email sent from www.virginmedia.com/email
Virus-checked using McAfee(R) Software and scanned for spam
Other related posts:
