[openbeos] Re: Bounties
- From: "Karl vom Dorff" <karl@xxxxxxxxxx>
- To: openbeos@xxxxxxxxxxxxx
- Date: Wed, 18 Jul 2007 10:54:12 -0400
Hi,
Thanks again for the reply.
On 7/18/07, Stephan Assmus <superstippi@xxxxxx> wrote:
Hi Karl,
> I learned a while ago not to consult this list for anything other than
> dev questions. And that's what my post was primarily about. Getting
> answers to what bounties are needed. I don't want to put pressure on
> the developers (as you say below) or 'bug' them with this issue, which
> was why I didn't mail the list about bounties in the first place. I
> leave developing to you, let us worry about this. There are several
> members at Haikuware that are Haiku developers, and I thought news
> would travel through them anyways. Seems I can't do anything right!
Please don't take it like that. I'm sorry if I sounded harsh, I didn't
actually mean to.
No worries.
With "after the fact" I only meant the vote for the bounty place. I wasn't
aware of the voting and would have voted to put any bounty effort on the
official Haiku site.
Well, I gave my explanation about this already. I'm sorry I can't change the
results.
I'm not asking how to run it really, I just wanted to know from the
> inside what the developers felt was missing and the bounties that
> could be of use.
As the list of possible bounties to chose from, I would _not_ use the
obvious wishlist (JAVA, WebKit, ...). Instead I would present a *short*
list of small items, which are realistic to interest an existing or new
developer. In another words, put the goal of "resulting in the completion
of one or more items" at the front while taking the current developer
situation into account. That means to avoid destroying the possible
success
of the bounty project by any illusions of what is possible with our
current
developer situation. Remove any items from the list, where you have any
doubts that the item might be completed, because then it is simply a wish
item, but would take away donations from actually achievable items. Like I
said, the last four items are good ones IMHO.
It's funny you mention the obvious wish list. You may see why later.
Anyways, I've contacted Bryan Varner about the Java issue just yesterday,
because I had an individual who wanted to contribute a fair bit of money to
such a bounty. Bryan was very helpful and explanatory, he is such a jovial
fellow and always in good spirits, I enjoy our correspondence. Anyways, it
boils down to a very complicated issue with Sun and Bryan advised me to wait
with this.
> Yes, people realise that a bounty system can be helpful if done
> > > right, we said that before, but your list again tells me, that you
> > > have not
> > > listened to the input you did get from Haiku devs before.
> >
> >
> > I have listened to input from both sides and made compromises, and am
> > listening to it now. If you
> > read<
http://www.haikuware.com/index.php?option=com_content&task=view&id=
> > 43&Itemid=1>the original post where I made this list, it says that I'm
> > not sure these are the right things, they are merely ideas, and that I
> > would consult Haiku developers first before starting anything. And so
> > far I haven't really got any other input outside of that list. The
> > particular news article got around
> > 500 hits, of the ~70 members at Haikuware (of which several members
are
> > Haiku devs) nobody had any input either. So... I do so see mention of
a
> > virtual memory manager here. Is this possible/achievable/too
difficult?
That particular item would need clarification IMHO. I think Euan meant a
memory manager that would take care of managing graphics memory too? I
don't know, in any case, the virtual memory subsystem in Haiku is actively
being worked on. But yes, individual pieces of useful code for the Haiku
tree (without visible results like an application or driver) are possibly
a
good candidate as well. For example, I'd like a faster BRegion
implementation. But too many of such items, and you spread the funding too
much, or risk some people not understanding the usefulness of some items.
See, 'in any case, the virtual memory subsystem in Haiku is actively
being worked on.' that's why it's good to ask about these things first!
One more thing, eventually Haiku should name an official 3rd party
> repository, because although an operating system is nice, you need
> software for it. Maybe people don't want to develop for the OS itself
> but applications for it. There's no info on Haiku about how to do that,
> where to do that etc. This issue was raised a while ago in this
> thread<//www.freelists.org/archives/openbeos/06-2007/msg00149.html>
> . I've made the announcement that OSDrawer would handle the bounties.
> Can't Haiku do the same for 3rd party apps? Just a suggestion!
I agree it might be nice (in terms of generating a feeling of "unity"),
but
there are lots of services already out there, I mean to say, the fact that
Haiku does not endorse any service should not be stopping anyone from
writing Haiku apps. And the Haiku website already has the infrastructure
to
serve as a communication platform about such efforts.
I see. To each his own right?! That may also be a problem though. So many
choices, scattering of projects... Maybe a Haiku recommends x...?
Bye for now,
Karl
Best regards,
-Stephan
Other related posts:
