The goal of the BSD license is to promote the development of useful software. How it is used and what benefits the end users gain (or lose) are not relevant to the programmers who want to contribute code under this license.
If I want to see my software widely used and don't care about the fact that the end user may not be able to modify it, I would choose the BSD license.
are you sure bsd-style licenses forbid enduser modification? Most of the internet is down for me, so i couldnt verify this
As far as I know, BSD allows 'someone' to make a derivative work out of a 'free' work and then prohibit the community from looking at the source, changing, modifying or distributing it further. In fact Microsoft did just that with the TCP/IP stack developed by BSD.
Note that the license does not prohibit from making enduser modification, but allows others to prohibit, whereas the GPL prohibits the others from prohibiting.
Cheers, Debarshi
The BSD, Apache, MIT, etc. licences work because of a simple model. They allow you to take without giving back, but they ensure that unless you give back, your code will be so far out of sync with HEAD that you spend more money merging the latest changes into your code than you would spend in merging your changes back to HEAD.
On 8/17/06, Philip Tellis philip.tellis@gmx.net wrote:
The BSD, Apache, MIT, etc. licences work because of a simple model. They allow you to take without giving back, but they ensure that unless you give back, your code will be so far out of sync with HEAD that you spend more money merging the latest changes into your code than you would spend in merging your changes back to HEAD.
The same applies to GPL code as well doesn't it ? A case in example is vendor supplied drivers in the Linux kernel. The choice for them is to maintain it themselves or give it back to the community and let them maintain it.
-- Vinayak
Quoting Philip Tellis philip.tellis@gmx.net:
The BSD, Apache, MIT, etc. licences work because of a simple model. They allow you to take without giving back, but they ensure that unless you give back, your code will be so far out of sync with HEAD that you spend more money merging the latest changes into your code than you would spend in merging your changes back to HEAD.
true - and they are also very easy to understand, and also most of them are recognised by the FSF itself
On 17/08/06 12:31 +0530, Philip Tellis wrote:
The BSD, Apache, MIT, etc. licences work because of a simple model. They allow you to take without giving back, but they ensure that unless you give back, your code will be so far out of sync with HEAD that you spend more money merging the latest changes into your code than you would spend in merging your changes back to HEAD.
Which works on the assumption that people using/deriving BSD code want to stay in sync with HEAD.
I don't know of too many projects doing this. They usually use the BSD stuff as a way to get to a running start, but not much beyond that.
Devdas Bhagat
Quoting Devdas Bhagat devdas@dvb.homelinux.org:
On 17/08/06 12:31 +0530, Philip Tellis wrote:
The BSD, Apache, MIT, etc. licences work because of a simple model. They allow you to take without giving back, but they ensure that unless you give back, your code will be so far out of sync with HEAD that you spend more money merging the latest changes into your code than you would spend in merging your changes back to HEAD.
Which works on the assumption that people using/deriving BSD code want to stay in sync with HEAD.
I don't know of too many projects doing this. They usually use the BSD stuff as a way to get to a running start, but not much beyond that.
again postgresql, django, zope, lighttpd, apache?, subversion?
On 17/08/06 02:40 -0700, lawgon@au-kbc.org wrote:
Quoting Devdas Bhagat devdas@dvb.homelinux.org:
<snip>
Which works on the assumption that people using/deriving BSD code want to stay in sync with HEAD.
I don't know of too many projects doing this. They usually use the BSD stuff as a way to get to a running start, but not much beyond that.
again postgresql, django, zope, lighttpd, apache?, subversion?
And how many of those have been rolled into closed products with significant changes, and kept on going that way?
The only one I can think of is Bizgres. The biggest exceptions I can think of are Cisco, Juniper, Microsoft and Apple Computers.
Devdas Bhagat
Quoting Devdas Bhagat devdas@dvb.homelinux.org:
On 17/08/06 02:40 -0700, lawgon@au-kbc.org wrote:
Quoting Devdas Bhagat devdas@dvb.homelinux.org:
<snip> > > Which works on the assumption that people using/deriving BSD code want > > to stay in sync with HEAD. > > > > I don't know of too many projects doing this. They usually use the BSD > > stuff as a way to get to a running start, but not much beyond that. > > again postgresql, django, zope, lighttpd, apache?, subversion? > And how many of those have been rolled into closed products with significant changes, and kept on going that way?
does it matter - if software is free shouldnt people be free to roll them into their closed source projects. Not a nice thing to do, i agree, but certainly allowable. The reason why mysql doesnt give commit rights to people who are either not employed by them or not on contract with them is because of the whole hassle of who owns what is committed - which means they lose a lot of potential committers and commits. Postgresql doesnt have that problem. If you are good enough, you get commit rights. After all in the last analysis, if i 'write' code and own copyright in it, the question is 'how much did i write?'. At the most just an incremental improvement on the work of lakhs of others. So why worry overmuch about someone 'stealing' it. These things may have been relevant 10 years ago - they are not relevant now.
kg
On Thursday 17 August 2006 02:07 pm, Devdas Bhagat wrote:
On 17/08/06 12:31 +0530, Philip Tellis wrote:
The BSD, Apache, MIT, etc. licences work because of a simple model. They allow you to take without giving back, but they ensure that unless you give back, your code will be so far out of sync with HEAD that you spend more money merging the latest changes into your code than you would spend in merging your changes back to HEAD.
Which works on the assumption that people using/deriving BSD code want to stay in sync with HEAD.
I don't know of too many projects doing this. They usually use the BSD stuff as a way to get to a running start, but not much beyond that.
Market share that's all tey care about. Finished Closed apps would not even qualify as beta. So working bsd app would outlive the product life cycle. After which u launch a new model with the latest and greatest from HEAD. If the source was open any reseller with gcc would be able to fix bugs effectively killing the new improved model.
-
Debarshi 'Rishi' Ray -
Devdas Bhagat -
jtd -
lawgon＠au-kbc.org -
Philip Tellis -
Vinayak Hegde