Hi, On Tue, Aug 04, 2009 at 12:30:30AM +0200, Thomas Schwinge wrote:
> Please, please, please, let's try to finally get some of these patches > installed before discussing matters to death. The real problem is not patches being discussed to death, but rather patches that have been reviewed still not getting comitted... Like some of Zheng Da's patches for example. I know that my reviews are too pedantic at times -- partially on purpose, trying to establish some good practices; partially because of my annoying perfectionism. But here I didn't just talk about formalities like variable names or comments, but raised a fundamental concern! The ironic thing here is that once you actually started looking at this patch, you did the same kind of "discussing to death" :-) > but I'm totally losing track of all these emails and huge discussion > threads and proposals and misunderstandings and clarifications of them > and further possibilities. Don't worry -- I keep track of the patches I reviewed. Better worry about those I didn't review... (Usually because they were put only in the patch tracker, not sent to the list.) > So. Sergiu, if you need specific review of some parts of this patch > (or any other patches), then please say so, otherwise please get it > installed. I'd be very grateful for you not to tell people to ignore my remarks and commit broken patches. The origianl patch is wrong and needs to be reverted, and a new one comitted once all concerns actually have been addressed. > I assume that you can confirm in some way (testing, staring at the > code, ...) that it does the correct thing. And should there be any > breakage, and we discover it later on, we'll repair it later on. The correct functioning of filesystem syncing is very hard to test; and any errors will be extremely hard to track down later on. Which makes it all the more important that we be able to say with some confidence that the code is really doing the right thing. -antrik-
