Time
S
Nick
Message
13:06 donsizemore joined #dataverse
13:16 pdurbin
The failing test (AccessIT) was a pre-existing condition on the "develop" branch. :)
13:29 donsizemore
there was also a MoveIT failure? https://jenkins.dataverse.org/job/IQSS-Dataverse-Develop-PR/view/change-requests/job/PR-6333/2/execution/node/3/ws/target/site/surefire-report.html#edu .harvard.iq.dataverse.api.MoveIT
13:32 pdurbin
Huh. 500 error. That's no good. :)
13:48 donsizemore
some of those are intermittent
13:50 pdurbin
database deadlocks, maybe? :)
14:41 pdurbin
Fingers crossed that the API test suite passes when this completes: https://jenkins.dataverse.org/job/IQSS-Dataverse-Develop-PR/view/change-requests/job/PR-6334/1/
14:41 pdurbin
donsizemore: I ended up just forwarding your email to the people who hung around after standup.
14:57 pdurbin
passing: https://github.com/IQSS/dataverse/pull/6334#issuecomment-548819150
14:58 pdurbin
donsizemore: this is such a fun new toy. Thanks again. :)
15:03 donsizemore
@pdurbin look at PR6334! love the screenshot
15:04 donsizemore
@pdurbin i'm stumbling towards setting up a test job and branches to test fixing that fork variable. but we've got a lunch talk here and kasha wanted hot chocolate
15:07 pdurbin
Merce is giving a lunch talk in about an hour. I'm planning on going.
15:07 pdurbin
yeah, that screenshot makes me very happy
15:07 pdurbin
phoenix should be worried though... its days are numbered
15:33 pdurbin
donsizemore: if you want I can say a little more about https://github.com/IQSS/dataverse-ansible/issues/129 while it's fresh in my mind. I just opened it. :)
18:27 donsizemore
@pdurbin hey hey.
18:28 pdurbin
donsizemore: check this out: https://dev2.dataverse.org/file.xhtml?fileId=66&version=1.0
18:28 donsizemore
@pdurbin i can do that
18:28 pdurbin
What do you think of the new "Preview" tab? :)
18:28 donsizemore
yis yis i've seen some of your iterations come through
18:29 pdurbin
juancorr might be interested in this too
18:29 pdurbin
donsizemore: anyway, what's on your mind?
18:30 donsizemore
about to submit another PR since jenkins changed the output of CHANGE_FORK (to match their API , apparently)
18:30 donsizemore
just finished our meeting with a new member of the UNC board of trustees
18:34 pdurbin
cool, cool
18:34 pdurbin
donsizemore: do you want me to make a pull request for the previewers?
18:35 donsizemore
you're welcome to or i can start on that next
18:35 pdurbin
let me assign it to myself for now
18:36 pdurbin
I made a new "don't forget" column on my board. :)
18:38 donsizemore
speaking of, i haven't covered my eyes and rebased our fork yet (maybe Monday)
18:39 donsizemore
so if you want to do your stealth PR again: https://github.com/OdumInstitute/dataverse/blob/6335_hard_code_repo/tests/jenkins/ec2/Jenkinsfile
18:39 pdurbin
We should probably rope Akio into that discussion. :) Is there stuff in your "develop" that you need? If not, I think I'd just force push whatever is in "develop" on IQSS to your "develop".
18:41 donsizemore
in theory he's got everything in IntelliJ anyway
18:41 pdurbin
heh
18:41 pdurbin
Well, I can tell you what I did with dataverse-installations.
18:41 donsizemore
anyway, i need to change one line in the Jenkinsfile
18:42 pdurbin
yeah, I can see the one line in https://github.com/IQSS/dataverse/compare/develop...OdumInstitute:6335_hard_code_repo
18:42 pdurbin
What if you make a new branch called "upstream" or something?
18:43 pdurbin
The idea would be that you keep upstream the same as "develop" on IQSS and always branch from upstream.
18:43 pdurbin
Does that make sense?
18:45 donsizemore
i can do that
18:45 pdurbin
https://github.com/pdurbin/wt-design-docs for example is 20 commits behind https://github.com/whole-tale/wt-design-docs ("master" branch for both)
18:45 donsizemore
again, i only tested against OdumInstitute/develop thinking i'd just switch the org when the time came
18:45 pdurbin
right
18:46 pdurbin
maybe we just need to get over this hump
18:46 donsizemore
i just need to rebase (or force)
18:48 pdurbin
I usually don't suggest force pushing but I think that's what I'd do in this case. I typically want the default branch (usually "master") of any fork I have (under "pdurbin", say) to be an exact copy of whatever is in the default branch of the upstream repo.
18:51 pdurbin
donsizemore: now I see plugins! https://github.com/IQSS/dataverse-jenkins/pull/20#issuecomment-548908003
18:52 donsizemore
so you think i should push --force instead of rebase?
18:53 pdurbin
donsizemore: there's a safer way
18:53 pdurbin
Or at least there's something I do that makes me more comfortable.
18:54 pdurbin
Here are the steps in my mind (I hope this makes sense):
18:54 donsizemore
well, i just cloned a current copy of his trsa-api branch
18:54 donsizemore
so we've got the current snapshot even if we lose the git history
18:54 pdurbin
- copy develop to develop2
18:54 donsizemore
or make it odum-develop
18:54 pdurbin
- delete develop (using the GitHub web gui)
18:54 pdurbin
- fetch the latest from IQSS
18:55 pdurbin
- push IQSS develop to develop
18:55 pdurbin
- delete develop2 if you don't need it
18:55 pdurbin
That's what I've done a couple times in a couple repos.
18:56 pdurbin
Sure, you could call it odum-develop instead of develop2 if you want.
18:56 donsizemore
Akio is watching over my shoulder, BTW
18:56 * pdurbin
waves at Akio
18:56 donsizemore
(just don't want to detonate a git bomb on a Friday afternoon)
18:56 donsizemore
can i just 'git branch -d develop'?
18:56 pdurbin
yeah, that's why I think this way is safer
18:57 pdurbin
You can delete branches locally and push the deletion to GitHub but I always find this to be a power user thing. I just use the GitHub web interface when I want to delete a branch.
19:01 donsizemore
do i say git config master.merge refs/head/master or git config master.merge refs/head/develop
19:02 pdurbin
Um. I'm not sure. You're making me think you should re-clone at this point. I re clone quite often. :)
19:02 pdurbin
I'm also not quite sure what your immediate goal is. :)
19:03 donsizemore
eh? eh? https://github.com/OdumInstitute/dataverse
19:03 pdurbin
"This branch is even with IQSS:develop." Perfect!
19:10 donsizemore
@pdurbin I didn't change the plugin import code, just down-graded Jenkins at install time
19:10 pdurbin
donsizemore: I'm looking at https://github.com/IQSS/dataverse/pull/6336 and it looks perfect in terms of a clean single commit. I see what you mean that this will only work if the fork is called "dataverse". That's what you were saying, right?
19:13 donsizemore
correct. and until we get a CHANGE_FORK_FULL that may be all we can do
19:15 pdurbin
I think that's fine. Do you want me to let Leonid take a look at this? Or should I just merge it?
19:16 donsizemore
i mean, it worked for me here: https://jenkins.irss.unc.edu/job/IQSS-Dataverse-Develop/
19:17 pdurbin
Yeah, I'm not worried about it working or not. I'm just wondering if we should get more of the team at IQSS looking at these pull requests. Leonid, Gustavo, Steve.
19:18 donsizemore
well, once you all approve it i can rescan the repository and see how far we get
19:19 donsizemore
a couple days ago pkiraly's PRs displayed in the web interface as "from an unknown repo" and returned a 404
19:19 donsizemore
but that seems to be fixed, maybe jenkins will like them better
19:19 pdurbin
yeah, that was weird
19:19 pdurbin
well, I can merge it now (or you can), if you're eager to rescan
19:23 donsizemore
your preview mode is in place
19:24 donsizemore
if you merge it #6336 now (i can't) i can rescan the repo and go to the gym =()
19:24 pdurbin
You can't merge it? Ohhhh, it's in "dataverse".
19:25 pdurbin
donsizemore: developers aren't allowed to merge anything in "dataverse" but I just approve it for QA to merge.
19:26 pdurbin
When you get back from the gym (or next week), please take a look at https://github.com/IQSS/dataverse-jenkins/pull/20/files#r341718673
19:27 donsizemore
now jenkins likes 6183 but not 6177
19:27 donsizemore
i can fix the file mode bit
19:28 pdurbin
donsizemore: thank you!
19:30 donsizemore
i'll just pop in a bash command for now. i'll swap out IQSS-Dataverse-PullRequest.xml for IQSS-Dataverse-Develop-PR.xml next anyway
19:31 pdurbin
awesome
20:45 pdurbin
https://github.com/IQSS/dataverse/pull/6333 merged. Tests passing on the develop branch ("API Test Status: passing" at https://github.com/IQSS/dataverse ). It's been a good week. I'm heading home. Have a great weekend, everyone!
20:46 pdurbin left #dataverse
23:43 nightowl joined #dataverse
23:49 nightowl left #dataverse
