Hi Guozhang, Ok, I was initially confused as someone mentioned in the comment of KAFKA-4772 that such a KIP is required there. I therefore tackled both issues at the same time and prepared a KIP and created PR to discuss this. However, since the KIP slows things down I guess its smarter to solve them independently. Therefore I'm going to close https://github.com/apache/kafka/pull/2669 But before I do so, would you mind giving a feedback on the printAction ? I struggle with the deserialization (previously introduced maybeDeserialize) part and whether this is really needed. If only instance of byte[] is checked, then we could directly apply key and value as far as I can see.
Marc On Fri, Mar 10, 2017 at 4:47 PM Guozhang Wang <wangg...@gmail.com> wrote: > KAFKA-4830 though, would require a KIP. > > On Fri, Mar 10, 2017 at 7:47 AM, Guozhang Wang <wangg...@gmail.com> wrote: > > > Hi Marc, > > > > I have added you for wiki access. On the other hand KAFKA-4772 would not > > require a KIP as it does not change any public APIs, it just clean up the > > internal implementations of these public APIs. > > > > > > Guozhang > > > > > > On Fri, Mar 10, 2017 at 7:10 AM, Marc Juchli <m...@marcjuch.li> wrote: > > > >> Hi all, > >> > >> I am preparing a KIP for KAFKA-4772. May I have permissions to create a > >> child page of the KIP page? > >> > >> Username: mjuchli > >> > >> Thanks and regards, > >> Marc > >> > > > > > > > > -- > > -- Guozhang > > > > > > -- > -- Guozhang >
