[Chandler-dev] Checking background sync into the trunk [Poll]

[Reid Ellis]

+1 on checking bgsync into the trunk now. The more eyes looking at it  
for various side-effects, the better.

Reid

On 6-Jun-06, at 7:08 PM, Morgen Sagen wrote:
> I still have one or two loose ends to wrap up, but we're getting  
> close to landing the background sync feature to the trunk.  The  
> biggest issue we know about is stamping/unstamping a shared item is  
> not yet supported. One reason for this is background sync makes  
> heavy use of the repository's view-merging functionality, and has  
> pushed view-merging more than any previous code.  View-merging has  
> changed enough that we would like it to be exercised and stabilized  
> before adding features to it (including stamping support).   Also,  
> since stamping may be implemented via annotations in the future,  
> that would obsolete any view-merging code Andi would have to write  
> to support stamping now.
>
> So the choices are:
>
> 1) Check bgsync code into the trunk now, do without stamping of  
> shared items for some amount of time, but get the rest of view- 
> merging solid first.
>
> 2) Keep the bgsync branch separate from the trunk until we can add  
> stamping support to the view-merging code, which means view-merging  
> won't be fully exercised before we add further complexity, and the  
> continuing head-aches of maintaining a branch (which Andi has been  
> mostly doing -- thanks Andi!)
>
> I vote for #1, but I would like to get feedback on how painful it  
> will be for people if shared-item-stamping isn't supported for a  
> while.  "A while" is a function of how much work it first takes to  
> stabilize sharing once we land the background sync changes.  After  
> that, supporting stamping sounds like it would be on the order of 2  
> to 3 weeks.
>
> ~morgen