sijie commented on issue #570: Multiple active entrylogs URL: https://github.com/apache/bookkeeper/issues/570#issuecomment-369243964 > Ah, I didn't go back as far as July. I don't see any concensus on design or a design document though. In fact, the discussions highlighted aren't even about per ledger ledger storage, but a question about synchronization. In fact, I see a lot of the issues I am raising being risen there. > There's nothing in the notes about a decision on a design. That's not to say that it wasn't discussed, but there's no record of a design discussion/decision, so it may as well not have happened. The Apache Board highlighted exactly this issue a year ago (feedback on 2017-02-27 report). > As far as the community is concerned, what happens in direct meetings may as well not have happened. what I was trying to explain here is there is already an effort since last May, which I wish you respect to the fact that three people have being working on this topic. I never used "decision" in my comments or said it is a "decision". I am not sure why you are going to pull the whole conversation to "decision" and "ASF policy" which sounds political and doesn't make any sense to me. A final decision for a code change or feature is made through approvals or +1s on the pull requests following the bylaws. This is the easy thing to reason about; all the others are efforts/discussion, which can happen in the community meetings, direct messages/chats, these are records not decisions, which should be kept in ASF infra at their best efforts. for feedback 2017-02-27, I have responded that we have all the meeting notes recorded. The board doesn't have any objections to that. And this is irrelative to the topic/questions here, which I would prefer leaving out of this thread. > To summarize, I'm -1 on the changes in their current form. This doesn't mean they can't go in obviously, just that you'll need two +1 from elsewhere. first of all, I am not the author of this feature. I am not sure why your statements sound like I am pushing this change in. what I have been done in this thread is just to explain the fact to you, my view of current approach. anyway, I have tried to explain what I can explain here, would like to defer any technical questions to Charan, since he is the best person to answer it. regarding the approach and the implementation, I am fine with current approach implemented in the pull request and am okay to continue my review in that form.
---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [email protected] With regards, Apache Git Services
