Time
S
Nick
Message
00:33 jri joined #dataverse
01:34 jri joined #dataverse
02:35 jri joined #dataverse
03:36 jri joined #dataverse
04:36 jri joined #dataverse
05:37 jri joined #dataverse
07:00 jri joined #dataverse
08:45 stefankasberger joined #dataverse
09:11 pvx joined #dataverse
09:19 pdurbin
Yes, you need a burrito but if you're using docker-aio it should give you one.
11:39 stefankasberger joined #dataverse
12:35 pdurbin
stefankasberger: not sure if you saw my last message: http://irclog.iq.harvard.edu/dataverse/2019-05-06#i_92470
12:49 donsizemore joined #dataverse
12:55 donsizemore
@pdurbin so the API test suite currently depends on the docker-dcm dockerfiles...
12:56 donsizemore
@pdurbin ima set DV_API_KEY, DV_APIKEY and DVAPIKEY and see how far i get
13:15 donsizemore
@pdurbin also: https://access.redhat.com/products/red-hat-enterprise-linux
13:22 pdurbin
rhel 8 is out for real now?
13:27 donsizemore
so they say
13:29 pdurbin
heh
13:30 pdurbin
Should I go ahead and create the dataverse-ansible issue asking for CentOS 8 support? :)
13:31 donsizemore
you're welcome to, but centos 8 won't be out for a while
13:32 pdurbin
yeah, I'll wait
13:32 pdurbin
please ping me when it's out :)
13:35 pdurbin
donsizemore: I wouldn't say that the API test suite depends on docker-dcm. I've been running the test suite on phoenix for three years. Is that longer than since Docker existed? :)
13:36 donsizemore
@pdurbin so is this create-all-and-test (instead of run_test_suite.sh)?
13:38 pdurbin
no, no, you have the right script
13:39 pdurbin
see "Goals and options" under "Build" at https://user-images.githubusercontent.com/21006/28791389-4797d3ee-75fa-11e7-8269-dca817459a22.png ?
13:41 donsizemore
must it be called on any particular port? i was passing it localhost but can make it 8080 if need be
13:41 donsizemore
(actually, i'm calling it against the siteUrl variable, which in vagrant is localhost.localdomain)
13:43 pdurbin
docker-aio uses a particular port (whatever is in the script) but when I run the api test suite from old Jenkins across the wire to phoenix is use port 80.
13:44 pdurbin
You can't see it in that screenshot above, which is old anyway, so let me grab the latest from old Jenkins.
13:44 donsizemore
so it's just not getting its burrito: "No API key defined for built in user management"
13:45 pdurbin
Here's what I have on old Jenkins under "Goals and options": test -Dtest=DataversesIT,DatasetsIT,SwordIT,AdminIT,BuiltinUsersIT,UsersIT,UtilIT,ConfirmEmailIT,FileMetadataIT,FilesIT,SearchIT,InReviewWorkflowIT,HarvestingServerIT,MoveIT,MakeDataCountApiIT -Ddataverse.test.baseurl='http://phoenix.dataverse.org '
13:45 pdurbin
Currently, I have to remember to copy anything new from the run test suite script over to old Jenkins. So they stay in sync.
13:47 donsizemore
Base URL for tests: http://localhost.localdomain
13:47 pdurbin
hmm? :)
13:48 donsizemore
i'm passing DVAPIKEY: '{{ dataverse.usermgmtkey }}' DV_APIKEY: '{{ dataverse.usermgmtkey }}' DV_API_KEY: '{{ dataverse.usermgmtkey }}' as environment variables
13:48 pdurbin
The whole setup we have now is confusing. I'm sorry about that.
13:48 donsizemore
just to be sure i got them all
13:48 pdurbin
You'll need a burrito for sure.
13:48 pdurbin
So that the api test suite can make users to test with.
13:50 pdurbin
All (most?) of the stuff you need to successfully run the api test suite is described at http://guides.dataverse.org/en/4.13/developers/testing.html#getting-set-up-to-run-rest-assured-tests ... burrito key, the ability for anyone to create dataverses, etc.
13:50 donsizemore
i had college roommates that didn't puke as spectacularly as maven
13:50 pdurbin
"In short, Dataverse needs to be placed into an insecure mode that allows arbitrary users and datasets to be created and destroyed." :)
13:52 donsizemore
aha. i thought that curl line was already fired, but it's only in sampledata, which is disabled by default
13:52 pdurbin
This is all the stuff I do on phoenix to put it into an insecure mode and make it ready for running the api test suite: https://github.com/IQSS/dataverse/blob/v4.13/scripts/deploy/phoenix.dataverse.org/post
13:52 donsizemore
i was assuming it was already set
13:53 pdurbin
no, I drop the database on every phoenix run
13:53 pdurbin
so I have to run setup-all.sh again
13:53 donsizemore
i mean in ansible... but that case was off
13:53 pdurbin
oh, I don't know
13:53 donsizemore
i see what happened
13:53 pdurbin
I am a student of dataverse-ansible. :)
13:54 donsizemore
i can't even claim taalibe at this point
13:55 pdurbin
stefankasberger: wait. Your big reveal of PyDataverse during the community call is in... 2 hours?!? Do I have that right?
14:20 stefankasberger
@pdurbin: the call starts in 1h40, yes, but the reveal will not be that big. :)
14:21 stefankasberger
@pdurbin: and, yes, i saw your post from yesterday. thanks!
14:28 donsizemore joined #dataverse
14:29 donsizemore
@pdurbin woo woo! "[ERROR] Tests run: 19, Failures: 14, Errors: 0, Skipped: 0, Time elapsed: 35.746 s
14:31 pdurbin
donsizemore: sounds like progress
14:32 pdurbin
stefankasberger: ok but no one has posted anything to our mailing list yet.
14:32 pdurbin
Did a tweet go out at least?
14:32 pdurbin
I don't see a tweet.
14:32 pdurbin
stefankasberger: what should I say in an email to the google group?
14:36 stefankasberger
how long do i have time for presenting it? and how is the format usually? should i do a screenshare and show a bit around? for me a discussion afterwards would be most valuable, to get feedback and also to show how others can contribute.
14:38 pdurbin
The format is usually very informal. On most calls there is no agenda and we talk for maybe 20 minutes or so. We never go longer than an hour. So it's really up to you. Do you care what I write to the Google Group? I'll come up with something if you don't mind. :)
14:52 pdurbin
stefankasberger: this is what I just posted. I hope it looks ok: https://groups.google.com/d/msg/dataverse-community/op753PRmoUY/0E9rTnLOAQAJ
14:54 donsizemore joined #dataverse
14:58 Richard_Valdivia joined #dataverse
14:59 stefankasberger
Thats great, thanks.
15:00 pdurbin
phew :)
15:00 pdurbin
I also posted this: https://github.com/CenterForOpenScience/osf.io/pull/8939#issuecomment-490114059
15:02 Richard_Valdivia
Hello everyone!! Sorry if this is not a right channel to asking about that. IŽm trying to understand how add new metadata in our Dataverse instalation (version 4.4). After read the documentation, I could not yet understand how these works. I have to create a new TSV file?
15:04 pdurbin
Richard_Valdivia: hi! Yes, this is a good channel to ask in. Yes, you will probably need to create a TSV file. What kind of metadata do you want to add?
15:16 Richard_Valdivia
Hi pdurbin !! well, our team of librarians are still working on the metadata and I and my partner are studying the process for later (next few weeks) to do the deployment.
15:17 pdurbin
Cool. I'm wondering what field or discipline.
15:18 Richard_Valdivia
So there is no way to do the metadata creation through the Dataverse interface, correct?
15:18 pdurbin
That's correct. It's command line (API ) only right now. Sorry.
15:19 pdurbin
It sounds like you found our documentation. I wrote a lot of it and can probably answer questions about it. :)
15:20 Richard_Valdivia
About your question, at the moment we are working on an important project for us. This is a research on the missing person bones. We call it CAAF.
15:21 Richard_Valdivia
haha... sorry pdurbin... English is not my native language. Maybe I make some mistakes... :)
15:21 pdurbin
We all make mistakes. :)
15:23 Richard_Valdivia
Yes, you right. IŽm reading the docs. Just to know if I understand right. Fist I have to create a text file (TSV) with all fields. And use "curl" to import this customization to Dataverse?
15:23 Richard_Valdivia
This is correct?
15:23 pdurbin
Yes, that's correct.
15:25 pdurbin
Do you think other installations of Dataverse might be able to use your custom metadata block?
15:25 Richard_Valdivia
Hmmm ... I got it. I thought I had missed something because I found the whole process a bit complex. I was imagining that I was following a very complicated path and that I should have a shortcut.
15:26 pdurbin
I'm asking because sometimes a member of the Dataverse community will create a custom metadata block and make it available for other installations. For example, Harvard Dataverse has been looking at the custom metadata block for software created by Johns Hopkins.
15:26 pdurbin
Yes, the process is extremely complex, unfortunately.
15:27 Richard_Valdivia
Well pdurbin!! At this time I belive this is not useful for others. But this is our firt project. We call here a "opennig project"...
15:28 pdurbin
Ok. That's fine. Harvard Dataverse has a few custom metadata blocks that are not useful to other installations. I was just curious. :)
15:29 Richard_Valdivia
IŽm from Universidade Federal de São Paulo - Brazil. After the first implementation that this one on the "bones", we will have the creation for several disciplines and researches in UNIFESP. So I believe that in the future we will be able to contribute with our metadatas.
15:30 Richard_Valdivia
But your curiosity is important. I am working to go to Dataverse Meeting 2019 and to exchange ideas with other partners.
15:31 Richard_Valdivia
pdurbin!!! If I succeed in this implementation, can I contribute our experiences here?
15:31 pdurbin
It's a great place to exchange ideas. donsizemore and stefankasberger will be there. Not sure about other people who are here in this channel.
15:32 Richard_Valdivia
Contribute and asking more questions, of course... :)
15:34 pdurbin
Richard_Valdivia: right now the best way to contribute is to clone the Datavere git repo and edit doc/sphinx-guides/source/admin/metadatacustomization.rst to suggest improvements to the doc you're reading. Or you could open an issue if you want someone else to edit it.
15:35 pdurbin
Richard_Valdivia: also, we have a community call in a little less than half an hour. I'll be there and you can call in and ask questions about metadata blocks if you'd like: https://dataverse.org/community-calls
15:36 pdurbin
Which disciplines in UNIFESP do you want to support? Which disciplines would benefit from a custom metadata block?
16:04 pdurbin
We're about to start the community call: https://dataverse.org/community-calls
16:05 pdurbin
andrewSC bjonnh bricas_ donsizemore jri juancorr pmauduit Richard_Valdivia stefankasberger yoh ^^
16:13 pdurbin
stefankasberger: you're doing great. pyDataverse sounds awesome! :)
16:24 jri joined #dataverse
16:54 jri joined #dataverse
17:35 donsizemore joined #dataverse
17:43 donsizemore
@pdurbin knock knock, mr. institutional memory?
17:44 donsizemore
@pdurbin see, you've graduated from "mr. shib"
17:52 dataverse-user joined #dataverse
17:57 stefankasberger
thanks. am looking so forward to the release on PyPI. :)
17:59 Richard_Valdivia joined #dataverse
18:02 jri joined #dataverse
18:03 Richard_Valdivia
Hello! pdurbin!!! Unfortunately my network down and I could not see the rest of the previous messages. But according to your question, the disciplines that, in the future, we will include data are those related to the sciences of the sea, bioinformatics and data of the disciplines related with neurology.
18:20 pdurbin
Richard_Valdivia: those sound like interesting metadata blocks. Do you know if any of those disciplines have standards to base their metadata on? For example, social science uses a metadata standard called DDI. For astronomy we use VO. For life sciences, we use ISA-Tab.
18:22 pdurbin
donsizemore: I just re-watched Indiana Jones and the Last Crusade. Indy: "Can't you remember?" Henry: "I wrote it down so I wouldn't HAVE to remember!"
18:26 Richard_Valdivia
pdurbin: I do not have this answer at this time. IŽm new in actual project and IŽm software developer. But at the earliest opportunity I'll ask the librarians about it. I understand that this subject is interesting to every Dataverse community and and that would be a good contribution!
18:28 pdurbin
Yes, it would be a great contribution. I'm thinking that perhaps our metadata exper could help you and your librarians create a custom metadata block for bioinformatics, neurology, or (what I would call) marine biology. Let me go ask him what he thinks.
18:28 * pdurbin
walks down the hall
18:31 dataverse-user joined #dataverse
18:35 pdurbin
Hmm. He isn't at his desk.
18:36 pdurbin
Richard_Valdivia: my thought was that maybe he could work with you to create a custom metadata block that would be useful for multiple installations of Dataverse. Then, once you understand the process better, you could create the one for your bones. :)
18:37 pdurbin
Of course, it sounds like your priority is the bones so you may not like that idea. :)
19:07 donsizemore joined #dataverse
20:09 pdurbin
Richard_Valdivia: still around? I just talked to our metadata expert.
23:04 jri joined #dataverse
