[cosmo-dev] 0.9.0 update

[Ted Leung]

On Nov 14, 2007, at 10:16 AM, Jared Rhine wrote:

> A recap of our 0.9.0 status.
>
> We have a final build of 0.9.0.  It has not been announced to lists  
> and the blog.
>
> We discovered there was a significant performance regression for  
> some operations with Cosmo 0.9.0, to the point that we rolled-back  
> the Hub.
>
> How should we set priorities, in particular with Ted mostly  
> unavailable this week?

My cell phone# is on the internal wiki - you should not hesitate to  
use it...

>
> Most devs are already into a high-urgency performance-review phase.
>
> I suggest the following for priorities:
>
> * Leave the tarball and tag in place; it's not really hurting  
> anyone right now.  Assume we're doing a 0.9.1 release fixing this  
> specific issue; today if it works out.

We have in the past had people grab tarballs without a public  
announcement.   Pulling the tarballs and just doing 0.9.1 is an option.

>
> * Do a all-hands-on-deck sort of thing for characterizing and  
> fixing the performance regression with /atom requests.  Defer any  
> other specific milestone, like 0.10 branching/code-complete until  
> we have a grip on this situation and probably an updated 0.9 pushed  
> to Hub.  Fixes from this work go into 0.9 branch.

+1

>
> * Tell Travis to go ahead with putting-the-A-back-in-Ajax work to  
> make our XHR calls not block.  That work can go into 0.9 branch if  
> it's completed early.

I'd prefer to fix one problem at a time.   This is something that I  
think we want to test more carefully -- I'd prefer to see this go  
into 0.10

>
> * Developers not tied up in the above continue work on 0.10 updates  
> to trunk.

+1

Ted