[os-infrastructure] Re: [os-engineering] Closed source components

[Joe Abbate]

Daryl Monge wrote:
> We clearly do not have the same mental model of these processes.   
> :-)   You don't build branch "CR1" with "Project D support".  You 
> cross-integrate head-rev changes into the Project D branch and build 
> Project D.  You could even cross-integrate changes from branch "CR1" 
> into branch "Project D" but I don't see the advantages to using such a 
> process. (It would be a valid process however.)  If I had a special 
> project I would just keep pulling changes I wanted from the head-rev 
> into my project.
>
> Project D is, well Project D.  Its branch is completely self 
> contained.  It is NOT cross-integrated into the head-rev and therefore 
> will not be cross-integrated into stable code branches as described in 
> CR1.
>
> Definitely a fun scenario to analyze........   Even useful

But then Ingres Database Community Edition Release 1, built from CR1, 
will *not* have D Language support as was my premise from the 
beginning.  Someone from the community who wants D Language support has 
to build the PROJD branch on his own or get a build from that branch.  
When some feature, like a new filesystem, is accepted by the Linux 
community as no longer experimental, Fedora (and other distributors) 
offer it as part of their regular releases, without having to ask 
dispensation from Red Hat, even if RH decides *not* to offer the feature 
to its commercial customers.

Joe