FreeLists / openbeos / [openbeos] Re: status of OpenBeOS

["Michael Phipps" <mphipps1@xxxxxxxxxxxxxxxx>]

>Personally, i believe OpenBeos is walking *really* slowly... isnt any way to
>speed up the processes?

I was actually just talking (voice) to one of the team leaders.
We really have three issues.
1) The progress that does happen isn't obvious to everyone. We are working on 
this for the new website, but we could use a "reporter". Someone who will 
(dilligently) subscribe to the CVS list and write up a description of the 
changes for the newsletter. If you are interested, give "Aaron Kavile" 
<akavlie@xxxxxxxxxxxxx > a shout.

2) The progress isn't milestones. If I spend 4 hours optimizing some function, 
it doesn't blow your mind. But it is necessary and important work. 

3) Not enough people. We have something like 10-15 active developers at any 
given time. Yeah, really. If you divide that up, that is 1-2 developers per 
kit. In some cases, that is a developer alone (Translation Kit is a good 
example). In other cases, you have 2 or 3. That is *NOT* a lot of people.  Look 
at it this way. How many developers does XFree86 have? We have 2 or 3 on the 
Interface Kit. How many people does libpng have? Libjpeg? We have 1, total. 

That we have made the progress we have is, in my opinion, stunning. OBOS has 
the best group of volunteers I have ever seen. We are all working very hard. 
Not that we are looking for a pat on the back, or anything, but most of us put 
in 10 or more hours a week. On top of jobs and stuff. We will not go any faster 
with the volunteers that we have today. It just can't happen - there is only so 
much code that one can put out in a day. 

If you have programming skills, we need you. Most of us have put aside our 
personal projects to finish the OS first. If you are sitting on the fence 
wondering if we will finish and wondering if your time would be wasted, well, I 
guess I would point this out to you - more than half of the kits are in alpha. 
At least a couple more are almost ready. We have produced a number of things 
already - PDF writer, HP PCL driver, OpenBFS, the new translator kit, several 
new translators, AC97 audio, 41 newsletters, the new midi server, the new print 
server, most of screen saver, the 6 app server prototypes, most of the IK 
widgets, and more. Most of the best talent in the BeOS world is either writing 
for OBOS or directly using our code. 

Our biggest needs right now are in kernel and networking. Our next big 
milestone is to boot the kernel (for real) off of a hard drive. No BIOS hacks, 
but real PCI/IDE booting. After that comes a whole number of milestones - we 
can integrate servers into the system. This is sort of the quiet, dark period. 
When we first started, there was a lot of hoopla and excitement. We could show 
rapid progress (twice as much code as we had last week!) and show that it was 
and is possible. But we also knew all along that there were big, hard pieces to 
build. Kernels don't grow on trees. If we had a few people that could work on 
it full time, we would be done by now. We don't. That is, by the way, how both 
Linux and 386BSD were done - by people hacking on it full time. And NewOS, for 
that matter. But we don't have that. We have part time volunteers. So we need 
more people than they had/have. 

I don't want to paint too gloomy a picture. Work is progressing and as well as 
can be expected for the level of help that we have. But 10 people can only take 
us so far so fast. If you want more done, the only way to get there is to help 
out. :-)

Michael