@jincheng sun Thanks for your kind sharing.
Regards,
Lei Rui
On 1/3/2020 16:01，jincheng sun wrote：
I see. Thanks for your feedback! Lei Rui & Justin!
As I know, Both Apache Beam[1] and Apache Flink[2] carry website updated
PRs when bring up the release votes. This approach benefits Apache
I see. Thanks for your feedback! Lei Rui & Justin!
As I know, Both Apache Beam[1] and Apache Flink[2] carry website updated
PRs when bring up the release votes. This approach benefits Apache Beam and
Apache Flink as the update changes of Beam and Flink are a little more
complicated compare with
Hi,
No need +1.
I don't think it's a good practice, because if the released file doesn't exist,
how can people review a PR like [1] where
there are download links in it waiting to be checked, e.g.,
Hi,
> I mean that prepare the PR when bring up the VOTE, then people can review
> the PR ASAP. we can merge it when vote has passed and 24 hours later after
> the release .
You could but I don’t see any great need, webpages are easily fixed if there is
an issue.
Thanks,
Justin
> No. You should only update the website after the IPMC vote has passes,
the file put into the Apache mirrors and you waited 24 hours or so for the
release to propagate.
I mean that prepare the PR when bring up the VOTE, then people can review
the PR ASAP. we can merge it when vote has passed
Thank Justin for the quick reply.
Regards,
Lei Rui
On 1/3/2020 11:43，Justin Mclean wrote：
Hi,
Do you think that when bring up a release VOTE, we need to add the PR of
update the website in VOTE thread? The changes of the PR would be similar
with
No. You should only update the website after
Good find. I like the idea of a ticket to keep improving these types of
things.
On 1/1/2020 10:01 PM, Xiangdong Huang wrote:
> Hi,
>
> using `find **/src -type f -exec grep -i "copyright" {} \; -print | sort
> -u`, many apache header format issues can be found..
>
> e.g.,
> 1. showTTLPlan.java
>
Hi,
using `find **/src -type f -exec grep -i "copyright" {} \; -print | sort
-u`, many apache header format issues can be found..
e.g.,
1. showTTLPlan.java
2. all files for QueryMetricUI function...
3. All scala files are using /** so the blank space is two chars while it
is one char in all java
