Mattia Tristo | 2 Feb 2004 15:55
Picon
Favicon

License

Hi to all,
            i have read the issue 23 newsletter article entitled "Divisions 
of Pain" and i think that the GPL license is more "secure". What about making 
a survey on what license OpenBeOS has to use and take a democratic decision 
based on the survey?

                                                                                
Mattia

Bruno G. Albuquerque | 2 Feb 2004 17:11
Picon
Gravatar

Re: License

Mattia Tristo disse:

> Hi to all,
>             i have read the issue 23 newsletter article entitled
> "Divisions
> of Pain" and i think that the GPL license is more "secure". What about
> making  a survey on what license OpenBeOS has to use and take a
> democratic decision  based on the survey?

No.

Yeah, that was blunt but this is the only possible answer to that. We, the
developers working on the OpenBeOS project, decided a long time ago we
would use the BSD/MIT license. That won't change. In fact, I am really
glad we decided to go this way.

-Bruno

Matthijs Hollemans | 2 Feb 2004 16:16

Re: License

Mattia wrote:

[the GPL license]
> ... and take a democratic decision based on the survey?

There is nothing democratic about the OpenBeOS project. The people who write
the code make the decisions. And this particular decision was made long ago.

-Matthijs

Michael Phipps | 2 Feb 2004 17:56
Picon

Re: License

I find the wording particularly interesting. What do you mean by "secure"? I 
would guess probably not the computer security meaning. The usual complaint 
that people have about MIT/BSD (with respect to GPL) is that people might 
come along and "steal" our code and use it in a commerical product. Is this 
your concern?  

If it is, let me state clearly: "Let Them!". 

In any case, the amount of time and effort to re-license our 2 million lines 
of code from (guessing) 40 different authors would not be worthwhile without 
some radically earth shaking news.

On 2004-02-02 at 09:55:04 [-0500], Mattia Tristo wrote:
> Hi to all,
>             i have read the issue 23 newsletter article entitled "Divisions
> of Pain" and i think that the GPL license is more "secure". What about 
making
> a survey on what license OpenBeOS has to use and take a democratic decision
> based on the survey?
> 
>                                                                             

> Mattia

Adi Oanca | 3 Feb 2004 00:39

Re: License

Michael Phipps wrote:
> I find the wording particularly interesting. What do you mean by "secure"? I 
> would guess probably not the computer security meaning. The usual complaint 
> that people have about MIT/BSD (with respect to GPL) is that people might 
> come along and "steal" our code and use it in a commerical product. Is this 
> your concern?  
   Yeees! I want to help a friend to implement a large distributed 
system! And what better way than relying on a modified
   BHandler/BLooper/BMessage/BMessenger
architecture? It's PERFECT!

> If it is, let me state clearly: "Let Them!". 
   Yeeeeees, LET them! :-)

Adi.

Andrew Bachmann | 3 Feb 2004 07:15
Picon

initialize_before, _init, etc.

Hello all,

There's been some discussion on the net list about a problem stemming from a situation where 
initialize_before wasn't called when a library was loaded.  I figured it was worth starting a 
thread in the main list about it since it seems to be something that we should (all) discuss.  In 
particular it seems that we can't generally rely on initialize_before.  It's not particular portable. 
(read: at all)  I thought maybe we could use these macros from kernel/image.h:

#define	B_INIT_BEFORE_FUNCTION_NAME		"initialize_before"
#define B_INIT_AFTER_FUNCTION_NAME		"initialize_after"
#define	B_TERM_BEFORE_FUNCTION_NAME		"terminate_before"
#define B_TERM_AFTER_FUNCTION_NAME		"terminate_after"

But you can't very well write anything like this:

void B_INIT_BEFORE_FUNCTION_NAME (void)
{
....
}

Because the name is quoted.  I tried looking a bit on the net about this, and in particular I looked 
for _init and _fini.  I found an interesting message:

http://mail.gnu.org/archive/html/libtool/2001-03/msg00017.html

... which recommended the use of __attribute__((constructor)) and __attribute__((destructor))  I 
did not test these on our gcc.

I think we should come up with one strategy for doing this and hopefully it will be one that we 
can reasonably port to other architectures, eh?  Recommendations?  Maybe even just a set of 
(Continue reading)

Bruno van Dooren | 3 Feb 2004 08:40
Picon
Favicon

Re: License

I am a professional software developer, and some time ago i had to implement
a crc32 checksum algorithm on an embedded DSP. i had to search the net for
some time to find an implementation that was not GPL. GPL was unacceptable
because it was a proprietary application.

Vice versa i don't mind if people use my public code and use it in their
application when they need it.
To me the BSD/MIT license is the real FREE software license.

> > If it is, let me state clearly: "Let Them!".
>    Yeeeeees, LET them! :-)
Yay

kind regards,
    Bruno.

Mat Hounsell | 3 Feb 2004 10:48
Picon

Writing an Audio Driver

I was thinking I need to develop for my own good and I thought a sound driver
would be make the most difference to me.  

I was looking to write a sound card driver for my chipset Realtek ALC650 (an
AC'97 device).

Remember I have a second install that I play on.

I hoped that I could copy over the audio stuff from a openbeos build but the
audio_module crashed from a media_addon whatsit. Is the audio driver interface
the same? should I try moving the media kir?

I have scoured bedrivers and I was wondering if there was anymore to start
(besides the AC'97 stuff in the repository) ???

http://greetings.yahoo.com.au - Yahoo! Greetings
Send your love online with Yahoo! Greetings - FREE!

Jérôme Duval | 3 Feb 2004 11:55
Favicon

Re: Writing an Audio Driver

Hi Matt,

AC97 devices usually are managed by a motherboard chipset. We have now Intel 
based (possibly nForce and SiS as well) and Via based audio southbridge devices 
handled in CVS. If yours doesn't work, support should be added in one of these 
existing drivers (auvia, ich).

You should identify your audio device by its PCI vendor id/device id (in the 
Devices preference panel). You should then know which driver can handle your 
device, and verify it is well handled.

Hope it helps,
Jérôme 

Selon Mat Hounsell <mat_geek@...>:

> I was thinking I need to develop for my own good and I thought a sound
> driver
> would be make the most difference to me.  
> 
> I was looking to write a sound card driver for my chipset Realtek ALC650 (an
> AC'97 device).
> 
> Remember I have a second install that I play on.
> 
> I hoped that I could copy over the audio stuff from a openbeos build but the
> audio_module crashed from a media_addon whatsit. Is the audio driver
> interface
> the same? should I try moving the media kir?
> 
(Continue reading)

Mattia Tristo | 3 Feb 2004 12:29
Picon
Favicon

Re: License

>I find the wording particularly interesting. What do you mean by "secure"? I 
>would guess probably not the computer security meaning. The usual complaint 
>that people have about MIT/BSD (with respect to GPL) is that people might 
>come along and "steal" our code and use it in a commerical product. Is this 
>your concern? 

Yes, something like this.

>If it is, let me state clearly: "Let Them!". 

Yes, why not, but what i was thinking is that it's a great thing to have a 
lot of user but it is also a great thing to have a community that grows 
always.

>In any case, the amount of time and effort to re-license our 2 million lines 
>of code from (guessing) 40 different authors would not be worthwhile without 
>some radically earth shaking news.

Ah, i didn't know that change license would be a big work.

                                                                            
Mattia


Gmane