Thanks everyone for the comments and votes.
KIP-79 has passed with +4 binding (Jason, Neha, Jun, Ismael) and +1
non-binding (Bill).
On Mon, Sep 19, 2016 at 1:54 PM, Ismael Juma wrote:
> Thanks for the KIP, +1 (binding) to the latest version.
>
> Ismael
>
> On Sat, Sep 10,
Thanks for the KIP, +1 (binding) to the latest version.
Ismael
On Sat, Sep 10, 2016 at 12:38 AM, Becket Qin wrote:
> Hi all,
>
> I'd like to start the voting for KIP-79
>
> In short we propose to :
> 1. add a ListOffsetRequest/ListOffsetResponse v1, and
> 2. add
Hi, Jiangjie,
For 1, yes, since this api is about timestamp, returning any other metadata
more than the timestamp may not make sense any way. So, we can leave the
api as it is.
Thanks,
Jun
On Mon, Sep 19, 2016 at 12:57 PM, Becket Qin wrote:
> Thanks for the comment Jun.
Hi all,
A few comments below.
On Mon, Sep 19, 2016 at 5:49 PM, Jun Rao wrote:
>
> 1. I thought at some point you considered to only return offset in
> offsetsForTimes instead of offset and timestamp. One benefit of doing that
> is that it will make the return type consistent
Thanks for the comment Jun. Please see the reply in line
On Mon, Sep 19, 2016 at 10:21 AM, Jason Gustafson
wrote:
> +1 on Jun's suggestion to use "beginning" and "end". The term "latest" is
> misleading since the last message in the log may not have the largest
> timestamp.
+1 assuming the minor comments are addressed.
Thanks,
Jun
On Mon, Sep 19, 2016 at 9:49 AM, Jun Rao wrote:
> Hi, Jiangjie,
>
> Thanks for the proposal. Looks good to me overall. Just a couple of minor
> comments.
>
> 1. I thought at some point you considered to only return
+1 on Jun's suggestion to use "beginning" and "end". The term "latest" is
misleading since the last message in the log may not have the largest
timestamp.
On Mon, Sep 19, 2016 at 9:49 AM, Jun Rao wrote:
> Hi, Jiangjie,
>
> Thanks for the proposal. Looks good to me overall.
Hi, Jiangjie,
Thanks for the proposal. Looks good to me overall. Just a couple of minor
comments.
1. I thought at some point you considered to only return offset in
offsetsForTimes instead of offset and timestamp. One benefit of doing that
is that it will make the return type consistent among
Thanks everyone for the votes, I want to mention a minor change made to the
interface.
To ensure the accordance with KIP-45, we have changed the
"earliestOffset()" and "latestOffset()" method to take
Collection instead of Set.
Thanks,
Jiangjie (Becket) Qin
On Sun, Sep 18, 2016 at 5:53 PM,
+1 (binding)
On Wed, Sep 14, 2016 at 9:57 AM Bill Bejeck wrote:
> +1
>
> On Tue, Sep 13, 2016 at 11:08 PM, Jason Gustafson
> wrote:
>
> > +1 and thanks for the great proposal!
> >
> > On Fri, Sep 9, 2016 at 4:38 PM, Becket Qin wrote:
+1
On Tue, Sep 13, 2016 at 11:08 PM, Jason Gustafson
wrote:
> +1 and thanks for the great proposal!
>
> On Fri, Sep 9, 2016 at 4:38 PM, Becket Qin wrote:
>
> > Hi all,
> >
> > I'd like to start the voting for KIP-79
> >
> > In short we propose to :
> >
+1 and thanks for the great proposal!
On Fri, Sep 9, 2016 at 4:38 PM, Becket Qin wrote:
> Hi all,
>
> I'd like to start the voting for KIP-79
>
> In short we propose to :
> 1. add a ListOffsetRequest/ListOffsetResponse v1, and
> 2. add earliestOffsts(), latestOffsets() and
Hi all,
I'd like to start the voting for KIP-79
In short we propose to :
1. add a ListOffsetRequest/ListOffsetResponse v1, and
2. add earliestOffsts(), latestOffsets() and offsetForTime() methods in the
new consumer.
The KIP wiki is the following:
13 matches
Mail list logo