On 05/03/2012 09:31 AM, Alexander Sack wrote:
On Thu, May 3, 2012 at 6:10 PM, John Stultzjohn.stultz@linaro.org wrote:
On 05/03/2012 07:15 AM, Alexander Sack wrote:
On Thu, May 3, 2012 at 4:00 PM, Andrey Konovalov andrey.konovalov@linaro.org wrote:
Sorry for the delay.
I've updated the linux-linaro-core-tracking, but it currently misses the linaro-android-3.4 topic. Will put the linaro-android-3.4 topic back after resolving the merge conflicts; guess later today.
Is the resolution simple enough? Otherwise, feels that the better scalable approach is that the topic owner (e.g. John Stultz for the time being) gets notified and asked to fix the merge conflict on his side. Your daily merge scripts would then pick it up automatically as soon as the merge conflict is resolved.
So Andrey mailed me that there was a conflict between Linus' v3.4-rc5 and the Android tree, but when I went to updated the tree sort it out, v.3.4-rc5 merged in without any issues. I suspect the collision is with something else in the linaro tree, but haven't yet gotten any feedback as to what that might be.
OK, feels we could benefit from getting Andreys tools out and enable you to easily fold linux-linaro with your your own local topic?
I think in this case we've already cleared it up as a slight communication issue.
Would this help you to more easily converge, prepare and maintain topics that coexist in linux-linaro?
Probably not. I'm unable to dedicate as much time to merging items as Andrey, so I'm not likely able to keep pace. As it stands now, I update the Android tree at least ~once a week (sometimes more). So Andrey notices collisions before I do. Usually he then pings me and that prompts me to try to solve it.
Although I've had a few cases where at Andrey's prodding I spend a few hours resolving the collision and testing the resulting tree, only to find the Android team did the same in parallel. Usually then I just pick up the Android teams' work, but I'm not happy about burning my time that way.
This is one of the drawbacks of being always too close to the edge. We end up trying to address issues that others in the community also are addressing in parallel. I think its great to be proactive this way (especially if you're really the patch-queue owner), but it can keep us busy and we miss out on leveraging others work.
thanks -john