[openbeos] News, Views and Overdues

  • From: "Michael Phipps" <mphipps1@xxxxxxxxxxxxxxxx>
  • To: openbeos@xxxxxxxxxxxxx
  • Date: Mon, 20 Aug 2001 09:40:10 -0400

Good evening, troops!

There are now more than 60 people on this list. WOW! I am consistantly 
impressed more and more with the BeOS community.
I have no news to report from the Palm folks, other than the survey that BeFAQs 
is doing has had a massive response!

Many of you have read the IRC log on here that was posted today. Some very good 
thoughts and analysis.

The time is coming very shortly, I think for us to split up into our respective 
kits. There are many people chomping at the bit. :-)

Please, everyone, get me some clues as to what you want to work on! I have 
about 12 of these so far. If that means that only 12 people want to work on 
this and 50 people want to watch, we will release 1.0 in 2029! :-) 

This is a list of the tasks, as I see them. No particular order.

Printing        OpenGL          file system     app_server      kernel          
game_kit                midi            
mail            media           translation     input_server    networking      
screensaver             OpenTracker/Deskbar

Those of you who have already mailed me, I will sign you up for items. Those 
who have not, let me know. 

I am not publishing a lot at this point. There are a couple of reasons for 
this. One is that I don't want to run too far down this path until I hear from 
Palm.
Another is personal - I am moving at the end of this week, so we are packing, 
etc. My email should remain the same, but I may be out of touch for a day or 
two...

For what my approval is worth, the charter that was posted here a few days ago 
(cut and paste coming) is very good. Thanks to Daniel Reinhold.
----------------------------------------------------------------------------------------------------------

OpenBeOS Charter 
------------------------------ 
 
Description: 
OpenBeOS is an open source implementation of the BeOS operating system. 
 
Mandate: 
To provide complete source and binary compatibility for all  applications 
written for BeOS R5. This means that: 
a) any application that runs on BeOS R5 can be run unaltered on  OpenBeOS 
b) any source code that compiles and runs on BeOS R5 can be compiled  and run 
unaltered on OpenBeOS 
 
Architectures: 
The first goal is to achieve the Charter Mandate on intel x86  platforms. This 
may be followed by other machine architectures in the future. 
 
Design: 
OpenBeOS will implement the entire BeOS R5 API. This API can be extended over 
time, but will alway retain backwards compatibility. The BeBook and other 
documentation from Be Inc. will serve as the official guideline for the 
OpenBeOS API. As extensions are made, the documentation will be updated to 
reflect this. 
 
Source Code: 
The OS kernel is written in C. All remaining source code is written in  C++ and 
complies with the OpenTracker guidelines. 
----------------------------------------------------------------------------------------------------------

Other issues:
License - on source-forge, I chose MIT License. This is always a religious 
issue. Suffice it to say that a number of people have expressed their desire to 
avoid GPL. I tend in that direction myself, but if there was some MASSIVE 
swelling of opinion the other way, I would go with the majority. I doubt it. We 
aren't generally here because we are open source zealots, but because we want 
BeOS.

Sources - As an open source project, we can borrow code with a lot more abandon 
than Be could. We need only be careful of GPL'ed code. If there are any actual 
lawyers here, please talk to me. If not, I would rather avoid GPL unless 
someone has some really outstanding reason not to. That leaves us free to use a 
lot of code, though.

Leadership - I think (and others can certainly talk me out of this), what I 
would like to do is break into groups. Groups that communicate, sure. But 
groups none the less. In 3 days, we have produced more than 130 messages. That 
is a lot. Small groups centered around kits would help in this regard. I am 
looking for motivated, mature people to lead groups. I personally will lead the 
kernel group for now, as well as the project as a whole (unless the majority 
disagrees). That just gives a single point of contact for people to get to us, 
as well as dealing with administrative things. I am not a dictator. Promise. 
:-) 

Final note:

Please forgive the length here. There are a lot of people interested in the 
kernel. Suggestions have ranged from using Linux to using the Be kernel. I want 
to draw a line in the sand. Short term, we have to use Be's. I see only two 
possibilities for a kernel on a long term basis. Either we modify NewOS or we 
make one from scratch. I have emailed Travis as a representitive of the group 
asking him what he would like to do about this. I would love to get him on 
board. If not, we will write our own with NewOS supplying code and ideas. 
Either way, I want to rule out all of the other alternatives now. If you 
object, let's take it off line. We have to make some decisions some time or we 
will never get anywhere. This is one that I want to make sooner and more 
concisely than others. :-)

Thanks, all, for reading this far. :-)

Michael

Other related posts:

  1. Home
  2. haiku
  3. Archive
  4. 08-2001