continuum-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Jason van Zyl <>
Subject Re: Continuum 1.1 roadmap
Date Mon, 03 Jul 2006 09:17:28 GMT

On 3 Jul 06, at 9:30 AM 3 Jul 06, Brett Porter wrote:

> On 28/06/2006 8:40 PM, Jason van Zyl wrote:
>>> When we last discussed it on the development process, we talked  
>>> about instead doing regular promotion of the automated builds  
>>> (eg, roughly once a week we say "this is a stable build with  
>>> something new/a good fix/etc, let's ask people to test it").
>>> I'd really like to try that (and add anything to continuum to  
>>> make it easier :)
>> Yes, but we still have to make some official public releases. I  
>> don't think we can just release Continuum generated builds and  
>> then just release 1.1? Or do I misunderstand what you're talking  
>> about.
> I think we are talking about the same thing, but it's just mechanics.
> So a fundamental thing I think we need is regular builds that at  
> least passed the basic integration tests (ie, they compiled and the  
> server started). These are what we have now, though I'd rather they  
> came from Continuum and were more easily accessible to joe schmo  
> just coming to the web site.
> The next thing is regularly approved builds. Currently we schedule  
> and do the alpha/beta thing, but what I'm thinking is not going  
> through the whole release process that is time consuming and doing  
> something more regular - ie every week or two we say "there's been  
> a few new features, or a significant internal change, or some bad  
> bugs fixed and we want people to test it", so we check a particular  
> build is reasonably stable and then vote (or something) to promote  
> it - and we put that up on the web site as the latest "test" build.  
> It's somewhere between unstable nightlies and really stable releases.

So you're suggesting we not doing any official alpha/beta releases?

> I guess what I'm thinking of is something like IDEA's EAP program.
> Then the final release would what we do now: produce an RC which we  
> intend to be the actual binary, call for testers and vote. Release  
> that, or produce another one, then push it out to the mirrors and  
> announce.
> Cheers,
> Brett

Jason van Zyl

View raw message