Time
S
Nick
Message
06:47 nils`` joined #dataverse
06:51 nils`` joined #dataverse
07:04 nils`` joined #dataverse
08:12 jri joined #dataverse
09:50 skasberger joined #dataverse
10:44 jri joined #dataverse
10:58 skasberger joined #dataverse
12:41 jri joined #dataverse
12:48 poikilotherm
Ola pdurbin
12:48 donsizemore joined #dataverse
12:56 pdurbin
poikilotherm: ola. I had one more thought about the character bug. Can you please add a dataset that exercises it to https://github.com/IQSS/dataverse-sample-data ?
12:58 poikilotherm
A dataset?
12:58 poikilotherm
This is related to our custom metadata block
12:58 poikilotherm
I can create a dataset with lots of those characters to see if there are more places where this happens
12:59 poikilotherm
pdurbin: have you seen https://groups.google.com/forum/#!topic/dataverse-community/wjH_dPaoSAY ?
13:02 poikilotherm
Oh and I opened issue no 55 for installation no 55 https://github.com/IQSS/dataverse-installations/issues/55
13:04 poikilotherm
And you might be interested about taking a look at https://github.com/IQSS/dataverse-kubernetes/tree/173-dvcli/dvcli-plugin showing a first example of how to add a dvcli plugin for functionality not used by every installation...
13:04 poikilotherm
I think that's it from my side of "what did you miss" ;-)
13:17 Benjamin_Peuch joined #dataverse
13:17 Benjamin_Peuch
Hello everybody!
13:17 Benjamin_Peuch
"At least in Germany, if not also within EU, you need to provide a link to a legal notice for any webservice being public. There are even laws/court decisions how many clicks are appropriate to reach the legal notice from any page of the service (it's 2)."
13:18 Benjamin_Peuch
I was really dumbfounded when I read that, poikilotherm. :D
13:18 Benjamin_Peuch
(Then again, I think that's a very good rule.)
13:33 pdurbin_m joined #dataverse
13:33 pdurbin_m
Benjamin_Peuch: where did you read that?
13:34 pdurbin_m2 joined #dataverse
13:34 pdurbin_m2
poikilotherm: thanks for the update. Please go ahead and create a pull request to add your installation to the map! Exciting!
13:46 Benjamin_Peuch
pdurbin_m2: https://github.com/IQSS/dataverse/issues/6676
13:51 pdurbin_m2
Ah, thanks. Yeah, should be easy to add, I would think.
13:52 pdurbin_m2
poikilotherm: The setting table has a new "lang" column you might be able to use. It was added for terms of use in sign up. For French, originally.
13:58 adrix joined #dataverse
14:09 poikilotherm
pdurbin_m2 I was thinking about providing a setting for a URL and printing the text from bundles
14:09 poikilotherm
That way we don't have to mess with localized settings
14:09 poikilotherm
But good to know anyway!
14:10 poikilotherm
Benjamin_Peuch why where you dumbfounded?
14:12 poikilotherm
Its Germany... But as we learned in Tromso, the Netherlands are even better/worse with bureaucracy than we are :-) Ask Slava for the tale of reimbursement ;-)
14:12 poikilotherm
It's epic 8-)
14:13 Benjamin_Peuch
Oh dear.
14:13 Benjamin_Peuch
Dumbfounded because I thought the legislator had been most sharply precise here, by specifying the maximum amount of clicks.
14:14 Benjamin_Peuch
But I also think it's a very positive thing. ToS shouldn't be buried deep within a website: they should be easy to locate right from the go.
14:15 poikilotherm
Oh that two clicks are not in the law
14:16 poikilotherm
In the law it's just that the legal notice has to be reached immediately from any page of the service
14:16 poikilotherm
And then obviously this went to courts
14:17 poikilotherm
And here in Germany we have kind of second level support for laws
14:17 poikilotherm
They call it courts :-D
14:18 poikilotherm
A few judges decided that 2 clicks are sufficient to fullfill the laws demand of "immediately"
14:19 poikilotherm
So jurisdiction formed the details ;-)
14:21 poikilotherm
pdurbin_m2: do you want me to open an issue about that favicon thing?
14:22 pdurbin
andrewSC Benjamin_Peuch bjonnh bricas donsizemore jri juancorr pmauduit poikilotherm skasberger the community call starts in about 40 minutes: https://dataverse.org/community-calls
14:22 pdurbin
Please add your questions to the bottom of https://docs.google.com/document/d/13-szY0W9aIto1XaOKHS-L-HRktMsdWIX9vjGoH5P1Y8/edit #
14:22 poikilotherm
I'll try to participate. Working from home today, we'll see if kids are compatible :-D
14:23 poikilotherm
pdurbin: do you think I should add my freshly baked issues to the doc?
14:24 pdurbin
poikilotherm: sure! But if you don't call in, we'll probably handle them after the call, like we did with the questions from Singapore last night.
14:26 pdurbin
xarthisius: oh, we'll be discussing this "Computational Reproducibility in Dataverse" doc in our design meeting tomorrow morning, so if you have any bullets to add, please go ahead: https://docs.google.com/document/d/1xG8xAcPSOe1xCWUlhj46AKrK4MAZbY6ed96yBKHCXiA/edit?usp=sharing
14:30 Benjamin_Peuch
Thanks for the details of the story, poikilotherm. Jurisdiction is also a mechanism in Belgium, but I believe it's still more like France than the UK: more on the side of the Roman law than of the common law.
14:31 poikilotherm
Benjamin_Peuch: I have no idea :-D
14:32 poikilotherm
pdurbin: about adding us to the map
14:32 poikilotherm
Do you have a minute?
14:37 pdurbin
Oh! Is there a pull request already? Lemme look.
14:38 poikilotherm
pdurbin: no, I wasn't sure what to do.
14:38 pdurbin
sounds like we need more docs :(
14:38 poikilotherm
Your source of truth is your spreadsheet
14:39 poikilotherm
I can't make a pull request against that
14:39 pdurbin
Actually, it's worse than that (for now). For now you need to follow the instructions just under the map at https://dataverse.org/installations
14:39 poikilotherm
I feel fine to create a PR against the JSON file when you tell me you add the data to the spreadsheet
14:40 poikilotherm
So I need to contact support?
14:41 pdurbin
We can't make a pull request yet. Julians needs to add your installation to the old map first. The one we're going to shut down some day. And the process for that is for you t please email support
14:41 pdurbin
to*
14:42 poikilotherm
Just a thought: maybe that doc section of the guide would be helpfull in the README :-D If you feel lucky and have time, even a process diagram might come in handy ;-)
14:43 poikilotherm
So I will write an email to support, announcing that we are online. For now it's just our beta training facility (that will stay online for exactly this purpose), but real production is coming very soon
14:43 pdurbin
as long as it's your production URL , that's fine
14:56 donsizemore joined #dataverse
14:59 poikilotherm
donsizemore: Mooh!
15:02 pdurbin
poikilotherm: we got your ticket about adding your installation to the map: https://help.hmdc.harvard.edu/Ticket/Display.html?id=286824 Thanks!!
15:03 poikilotherm
Great :-)
15:04 poikilotherm
pdurbin: you still owe me an answer on creating an issue about the favicon :-D
15:05 poikilotherm
I don't know the other guy - do you feel like I just should go ahead?
15:05 pdurbin
poikilotherm: you didn't meet Obi in Tromso?
15:06 poikilotherm
I don't think so
15:10 donsizemore
@poikilotherm in case you were wondering, Thu-Mai only pretended to elbow Jon when he sat down.
15:11 poikilotherm
donsizemore: LOL didn't see that ;-)
15:15 pdurbin
poikilotherm: weird. He works at DataverseNO.
15:15 pdurbin
poikilotherm: that reminds me, is there a group picture from Tromso?
15:17 poikilotherm
Too many people around :-D
15:18 poikilotherm
Stefan has it
15:18 poikilotherm
Although AFAIK there is no complete group picture
15:24 donsizemore
@poikilotherm speak up!
15:24 pdurbin
poikilotherm: please. We can't hear you. :)
15:27 donsizemore
@poikilotherm there!
15:27 pdurbin
much better!
15:28 iqlogbot joined #dataverse
15:28 Topic for #dataverse is now Dataverse is open source research data repository software: http://dataverse.org | IRC Logs: http://irclog.iq.harvard.edu/dataverse/today | Who's who: https://docs.google.com/spreadsheets/d/16h3jv24usMGq18495C-JA-yNcQCKiKDa65MTraNDd7k/edit?usp=sharing
15:45 pdurbin
skasberger: group pic from Tromso, please
17:04 pdurbin
donsizemore: I'm (finally) circling back to the links you posted at http://irclog.iq.harvard.edu/dataverse/2020-02-24#i_119124 . Code Coverage Trend line! Hooray! Great work!
17:09 pdurbin
donsizemore: I'm seeing this: ERROR: HTTP 500 - Dataverse open-source-at-harvard could not be created. MSG: Command edu.harvard.iq.dataverse.engine.command.impl.CreateDataverseCommandhttps://jenkins.dataverse.org/job/IQSS-dataverse-develop/366/consoleFull
17:13 pdurbin
ConstraintViolationException on DataverseServiceBean.save(DataverseServiceBean.java:93) ... from what I can tell
17:14 pdurbin
CommandException: Command edu.harvard.iq.dataverse.engine.command.impl.CreateDataverseCommand
17:14 pdurbin
at edu.harvard.iq.dataverse.EjbDataverseEngine.submit(EjbDataverseEngine.java:267)
17:15 pdurbin
line 93 is Dataverse savedDataverse = em.merge(dataverse);
17:17 pdurbin
Hard to tell which of the methods threw the exception that was caught in line 267.
17:25 pdurbin
donsizemore: I'm confused though. Build 362 was fine ("Tests run: 116, Failures: 0"). Builds 363-366 seem to have been on the same commit as 362 but they all failed. I was worried that a pull request broke something but now I'm less worried about that. But I'm still worried. :)
17:29 poikilotherm
Folks do you have a minute for an unexperienced dataverse user? Regarding ZIP file handling on upload
17:29 pdurbin
poikilotherm: sure, go ahead. But I need some lunch soon. :)
17:30 poikilotherm
He, my wife is preparing pizza for the family ;-)
17:30 poikilotherm
On upload, FileUtil will try to unpack ZIP files, right?
17:30 poikilotherm
I'm not sure I fully understand what happens in there
17:31 pdurbin
it's been a while since I've been in that file
17:31 pdurbin
and I need lunch. but keep talking. I'll catch up
17:32 poikilotherm
If you upload a ZIP file with a directory structure, that will get lost, right?
17:32 poikilotherm
At least when looking at the code it tells me that it's skipping dirs
17:33 poikilotherm
But I remember some discussion that you could upload a directory structure and get the structure in the dataset via ZIP file
17:33 poikilotherm
But maybe that's just wrong in my head
17:34 poikilotherm
And I don't find a thing corresponding to the handling of ro-crate or BagIT
18:01 donsizemore
@pdurbin the API test suite has been failing off and on at different points in EC2
18:01 donsizemore
@pdurbin even on a t2.xlarge
18:04 donsizemore
@poikilotherm file hierarchy for zip files went into 4.11 https://github.com/IQSS/dataverse/issues/5498
18:04 skasberger
I think Philipp Conzett should have it (honestly, I dont remember, if we even made one...) too much upgrading I think. :)
18:25 donsizemore
@pdurbin hmmm, the 500 is a new one
18:27 donsizemore
@pdurbin oh. i see what happened. give me a minute
18:31 poikilotherm
donsizemore: ah ok, so I had that correct. So that ZIP dir removal is most likely because Dataverse does not have a directory DvObject. Good. Thx. I'll take another look and try again
18:37 pdurbin
poikilotherm: there is a recent spreadsheet all about zip and file hierarchy behavior: https://docs.google.com/spreadsheets/d/1vraiL1GoTSylkx9Kr_5-xZMJ-tUSVdRkduKU5_UZ6IM/edit?usp=sharing . It's a little hard to follow but I'm happy to walk you through it.
18:39 pdurbin
skasberger: no group pic is a crime. poikilotherm and I made sure we were in front of a nice old map at the State Archives of Belgium: https://twitter.com/philipdurbin/status/1223334005006794752
18:41 poikilotherm
OK I retested with new ZIP files and it's working just fine.
18:53 pdurbin
good
18:58 donsizemore
@pdurbin Tests run: 116, Failures: 0, Errors: 0, Skipped: 4"
19:00 pdurbin
donsizemore: we got a success? :)
19:03 pdurbin
donsizemore: 41% at https://jenkins.dataverse.org/job/IQSS-dataverse-develop/ws/target/coverage-it/index.html !
19:04 pdurbin
"API Test Status: passing" at https://github.com/IQSS/dataverse ! Nobody touch anything! ;)
19:22 donsizemore
@pdurbin https://jenkins.dataverse.org/job/IQSS-dataverse-develop/368/
19:22 donsizemore
@pdurbin yeah, i enabled the "publish status to github" step this morning, but something else broke the test suite
19:23 pdurbin
I'm so excited I'm making an issue to add a 4th button to the readme.
19:24 donsizemore
@pdurbin i don't mean to bear bad tidings, but we did see intermittent API test suite failures on Friday, in EC2, in t2.xlarge
19:25 pdurbin
yeah, Danny just created an issue to look into all this: https://github.com/IQSS/dataverse/issues/6681 ... we want to help
19:29 donsizemore
@pdurbin the two yellow builds were legit problems in the test suite. this morning's reds you can blame on me
19:30 donsizemore
@pdurbin I made a change in dataverse-ansible this morning, and tested in Vagrant, but didn't test deep enough
19:39 pdurbin
donsizemore: here's the new issue I excitedly just created about the higher 41% number (API test coverage): https://github.com/IQSS/dataverse/issues/6682
19:51 donsizemore
@pdurbin I installed the Code Coverage API Plugin so in theory we should be able to pull directly from Jenkins
19:55 pdurbin
donsizemore: awesome. I was having trouble finding it though.
20:00 donsizemore
@pdurbin yeah most everything addresses PR coverage reports
20:00 pdurbin
ah, not develop
20:00 pdurbin
or master or whatever
20:22 dataverse-user joined #dataverse
20:22 dataverse-user
Hi
20:23 dataverse-user
I hope you are doing well. I would like to know the status of my application and whether you need more information? Thanks
20:23 dataverse-user
my name is mostafa
20:24 dataverse-user
salari
20:25 pdurbin
dataverse-user: hi! Are you talking about https://help.hmdc.harvard.edu/Ticket/Display.html?id=286845 ?
20:26 dataverse-user
I applied for HILDA data set
20:35 poikilotherm
pdurbin: what do you want me to do in the spreadsheet now?
20:35 poikilotherm
Just add us?
20:35 pdurbin
poikilotherm: yes please!
20:36 poikilotherm
I'll add a column "OpenID Connect login", right?
20:36 pdurbin
poikilotherm: sure! Go for it.
20:39 poikilotherm
pdurbin: where do you track the geolocation and friendly name?
20:39 donsizemore
@pdurbin okay, so one of my _other_ changes today did not in fact break the build =) https://jenkins.dataverse.org/job/IQSS-dataverse-develop/369/console
20:42 pdurbin
poikilotherm: for now, here: https://services.dataverse.harvard.edu/miniverse/map/installations-json
20:42 poikilotherm
pdurbin: Oh. OK.
20:42 pdurbin
donsizemore: sorry, what am I looking at here? :)
20:42 poikilotherm
Oh I didn't send a logo...
20:42 donsizemore
@pdurbin now we have a https://jenkins.dataverse.org/job/IQSS-dataverse-develop/369/coverage/dataverse__jacoco_xml/
20:48 donsizemore
@pdurbin and we can get https://jenkins.dataverse.org/job/IQSS-dataverse-develop/369/coverage/dataverse__jacoco_xml/result/api/json?depth=1
20:49 pdurbin
poikilotherm: logos aren't shown on the new map but yeah, you should reply on the ticket and attach it. The RT ticket.
20:50 pdurbin
donsizemore: very cool but I don't see 41% in either of those.
20:55 pdurbin
But I do see 41% (which makes me happy) at the bottom of https://jenkins.dataverse.org/job/IQSS-dataverse-develop/ws/target/coverage-it/index.html
20:56 pdurbin
And scrolling through https://jenkins.dataverse.org/job/IQSS-dataverse-develop/ws/target/coverage-it/edu.harvard.iq.dataverse.api/Datasets.java.html makes sense to me in terms of which lines are being exercised or not. I'm not sure if you can easily see the red vs green.
20:57 pdurbin
If you hover over the diamonds you'll see popups like "All 2 branches missed"
20:58 pdurbin
I can tell at at glance (because it's all red) for example that this method isn't being exercised by our API test suite: @Path("{identifier}/dataCaptureModule/rsync")
20:59 pdurbin
Knowing what code is exercised (or not) by tests is extremely valuable. So thank you!
20:59 pdurbin
:)
