[openib-general] progress...but opensm crashed

[Grant Grundler]

On Wed, Jan 05, 2005 at 11:14:17AM -0800, Tom Duffy wrote:
> This is true for Linux kernel code as Linus requires it,

Linus in fact does not _require_ it. He just wants a name and
email address associated with each patch and who has reviewed it.
It's all based on trust.

He strongly *prefers* a "Signed-off-by" line and will get cranky if
folks submit too many patches without the S-O-B line.
Been there, done that.

> but do we require such things for the userland stuff?
> I see that it says so on openib.org.  Just making sure.

It's probably overkill for userspace stuff.

But I think it's good practice to track who submitted a patch
in the SVN logs as well as archiving the email.
Tracking this stuff down after the fact is just a PITA and
if someone else wants to adopt the code, they may in fact
need to have such a list of submitters.

grant