Fix data_collection_server

[k-okada]

same as #1542

• Sleep less time in data_collection_server.py
• Return false response in data_collection_server

[wkentaro]

FYI: you can push to the branch in #1542. Even force push.

[k-okada]

• I think you'd better to create branch on your repository, create branch on origin runs two different travis test, and that makes less chance of passing the tests.
• more over, test on origin branch somehow failing on test

[wkentaro]

I agree with some of your points, but there are some merits when we create branch on jsk-ros-pkg: - Save time of `git remove add` and `git fetch`. - No need to change URI in the .rosinstall The problems you pointed out can be solved, I think:

create branch on origin runs two different travis test, and that makes

less chance of passing the tests. For this, we can ignore one of the two tests. PR and Push build.

more over, test on origin branch somehow failing on test

Maybe because of the load on the Jenkins server, so we can filter the branch to run the Push build. One idea is disabling **Build branch updates**, and add **Cron Jobs* for *master* branch.

…

On Mon, Aug 21, 2017 at 7:26 AM, Kei Okada ***@***.***> wrote: Merged #1549 <#1549>. — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#1549 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AEHFkyOymuhVjjKuircVPGtCtNRudMTTks5saLKEgaJpZM4O8jqH> .

-- 和田 健太郎 / Kentaro Wada http://wkentaro.com -- 和田 健太郎 / Kentaro Wada http://wkentaro.com

[wkentaro]

@k-okada I found a solution. start-jsk/jsk_apc#2579

[k-okada]

I agree with some of your points, but there are some merits when we create branch on jsk-ros-pkg: - Save time of `git remove add` and `git fetch`. - No need to change URI in the .rosinstall

you can use `hub remote add -p k-okada && git fetch --all` I would not say everyone should not create branch on origin, if they know what's happens and take care that branch and do not forget to remove that, it's ok, but in general if people start create branch on origin, they leave as it is. the result would be something like https://bitbucket.org/osrf/gazebo/branches/ and it is hard to understand what is the active one.

> create branch on origin runs two different travis test, and that makes less chance of passing the tests. For this, we can ignore one of the two tests. PR and Push build.

github do not ignore one of these build

> more over, test on origin branch somehow failing on test Maybe because of the load on the Jenkins server, so we can filter the branch to run the Push build. One idea is disabling **Build branch updates**, and add **Cron Jobs* for *master* branch.

this is not so easy, https://travis-ci.org/jsk-ros-pkg/jsk_common/builds/266690514 My guess is something related to https://blog.sqreen.io/freeze-python-str-encode-threads/

On Mon, Aug 21, 2017 at 7:26 AM, Kei Okada ***@***.***>

wrote:

> Merged #1549 <#1549>. > > — > You are receiving this because you commented. > Reply to this email directly, view it on GitHub > <#1549 (comment)>, > or mute the thread > <

https://github.com/notifications/unsubscribe-auth/AEHFkyOymuhVjjKuircVPGtCtNRudMTTks5saLKEgaJpZM4O8jqH

…

> . > -- 和田 健太郎 / Kentaro Wada http://wkentaro.com -- 和田 健太郎 / Kentaro Wada http://wkentaro.com — You are receiving this because you modified the open/close state. Reply to this email directly, view it on GitHub, or mute the thread.

[wkentaro]

the result would be something like
https://bitbucket.org/osrf/gazebo/branches/ and it is hard to
understand what is the active one.

Maybe GitHub solves that problem

[wkentaro]

But maybe your point is that there can be someone who won't delete their stale branch on the origin, and you and other maintainers can hesitate to delete them..
And if I create branch on the origin, others will do..