[mythtv] status of MythTV wrt Coverity Scan

[Erik Hovland]

On Mon, May 7, 2012 at 8:34 AM, Eric Sharkey <eric at lisaneric.org> wrote:
> On Mon, May 7, 2012 at 11:07 AM, Stuart Morgan <stuart at tase.co.uk> wrote:
>> On Sunday 06 May 2012 23:00:36 Eric Sharkey wrote:
>>> I'll see if I can get a build done with the Coverity analyzer
>>> tomorrow.  Unless someone has another opinion, I'll use the fixes/0.25
>>> branch built on x86-64.
>>
>> We should use master, since it's better to fix new bugs before they make it
>> into the next release.
>
> Coverity has a concept that they call "streams" which are kind of like
> branches, but often a single branch might have multiple streams if it
> can be compiled more than one way (e.g., compiled with/without
> debugging, etc.).  Coverity can test all these streams and coordinate
> the defect reports across streams.  Ultimately you'll want to have a
> stream for master and another stream for the latest fixes branch.  I
> thought it would be simpler, on the first run, to start with a fixes
> branch as that should be more stable, then add a stream for master
> once the mechanics of using Coverity are better understood.

You are better off starting w/ master. It is the branch that will
matter going forward and it is easier to back port fixes then work like
heck to get defect fixes into a mostly frozen tree and then forward
port.

E

-- 
Erik Hovland
erik at hovland.org
http://hovland.org/