IQSS logo

IRC log for #dataverse, 2020-01-02

Connect via chat.dataverse.org to discuss Dataverse (dataverse.org, an open source web application for sharing, citing, analyzing, and preserving research data) with users and developers.

| Channels | #dataverse index | Today | | Search | Google Search | Plain-Text | plain, newest first | summary

All times shown according to UTC.

Time S Nick Message
08:31 Benjamin_Peuch joined #dataverse
11:45 poikilotherm joined #dataverse
12:17 pdurbin joined #dataverse
12:17 pdurbin Happy New Year!
12:20 poikilotherm Happy New Year :-)
12:26 pdurbin I already emailed Benjamin_Peuch the good news that I'll be speaking at FOSDEM: https://fosdem.org/2020/schedule/speaker/philip_durbin/
12:26 pdurbin poikilotherm: have you decided if you're coming? :)
12:26 poikilotherm Nope, not yet
12:26 poikilotherm But good to know when you will have your talk
12:27 poikilotherm Most likely I will come on Sunday then ;-)
12:28 pdurbin Great!
12:30 pdurbin Now I just need to write the talk. :)
12:31 poikilotherm It's a lightning talk, so it should be just a few cards or 1-2 slides, wouldn't it?
12:31 poikilotherm s/wouldn't/shouldn't/
12:31 pdurbin Yeah, only 15 minutes.
12:33 pdurbin I put "science" in the title to try to draw people in. :)
12:34 poikilotherm Oh 15 Minutes?
12:34 poikilotherm Nice
12:35 poikilotherm Most lightning talks I know only offer 2, 5 or 10 minutes
12:35 pdurbin Yeah, I gave a 5 minute one at LibrePlanet (also on Dataverse): http://wiki.greptilian.com/talks/2017/libreplanet-dataverse-lightning-talk/
12:36 pdurbin In that case, I didn't have any slides at all. I just brought up the Dataverse home page. Unplanned.
13:33 Benjamin_Peuch Hello everybody. Happy New Year!
13:33 Benjamin_Peuch Yes indeed pdurbin, I got your email. Thanks a lot. This is great news :)
13:34 Benjamin_Peuch They messaged you after Christmas so they can't even say they were keeping the info as a gift
13:54 pdurbin Benjamin_Peuch: the weird thing is that they've changed the URL for my talk. So the one is my email is now a 404. Here's the new URL: https://fosdem.org/2020/schedule/event/dataverse/ . And you saw that I was wrong about my flight? I won't be landing until lunchtime.
13:57 Benjamin_Peuch Haven't they heard about persistent identifiers? :D
13:58 pdurbin apparently not! :)
13:58 Benjamin_Peuch Indeed, regarding your flight, this could mean that you would not be available as soon as we were hoping.
13:59 Benjamin_Peuch Assuming you'd be ready for a meeting soon after that very long trip.
14:01 pdurbin Well, the flight itself is less than 4 hours. Faster than flying to California. :)
14:03 pdurbin Benjamin_Peuch: have you heard of Kokob or Baogo? I was thinking about signing up for this dinner: https://doodle.com/poll/vdvvd5xdzd2ytu2n
14:07 Benjamin_Peuch Kokob is delicious! Fine Ethiopian cuisine
14:08 Benjamin_Peuch It's best to go there with somebody else so that you get to enjoy their injera menu
14:08 pdurbin mmm, injera
14:09 Benjamin_Peuch Ever tried it?
14:09 Benjamin_Peuch I've never heard about the Baogo
14:10 Benjamin_Peuch Asian burgers. That must be interesting
14:10 pdurbin Yeah, my daughter's friend was adopted and is from Ethiopia. Her parents have cooked for us (and travelled there several times). Plus there's a nice restaurant in Boston.
14:11 pdurbin Benjamin_Peuch: this is the other event I had on my radar from Jan 31 before you reached out to me: https://chaoss.community/chaosscon-2020-eu/
14:12 pdurbin But I'm still happen to come visit if that works. Too bad it would be kind of late on a Friday afternoon.
14:13 pdurbin happy*
14:13 pdurbin happy to come visit
14:23 Benjamin_Peuch Hu. I thought this was the Chaos Computer Club for a moment
14:24 Benjamin_Peuch Oh I see. Much of the crowd attending this event will also be present at the FOSDEM, likely
14:25 Benjamin_Peuch If you're still up for a meeting on Friday, we would really love to have you pdurbin
14:25 Benjamin_Peuch We should see what time is most convenient for you since you'll arrive later than planned and we don't want to rush you
14:31 pdurbin Benjamin_Peuch: well, should we pick a time? Maybe I'd just come over directly from the airport?
14:32 pdurbin I'll try to keep my timezones straight this time. :)
14:34 Benjamin_Peuch If that's fine by you then great!
14:35 pdurbin How do you recommend getting from the airport to the state archives?
14:39 Benjamin_Peuch By train is definitely a good plan: it can take you to Central Station (Gare centrale / Centraal Station), which is located right next to the State Archives' headquarters.
14:40 Benjamin_Peuch And sure enough there's a train station beneath the airport.
14:44 * pdurbin likes trains
14:46 Benjamin_Peuch Hopefully then this doesn't happen to you :D
14:46 Benjamin_Peuch https://youtu.be/vyItidbjxLM
14:49 pdurbin lol
14:51 pdurbin Benjamin_Peuch: any thoughts on any sort of "fringe" event? Would it make any sense to use the state archives as a venue?
14:54 Benjamin_Peuch Yes, I also wanted to come back to this point.
14:56 Benjamin_Peuch I believe you don't need to register to attend the FOSDEM? Indeed that would be a bit counter-intuitive with the general philosophy of the event
14:57 Benjamin_Peuch I don't know if we can expect many people to attend our fringe meeting if we ask for registrations? So, you know, we know for how many people we need to prepare some coffee
15:00 pdurbin This will be my first time to FOSDEM but you're right, from what I understand. It's completely free to attend. You just show up. So yes, it would be nice for a fringe event to be similar.
15:00 Benjamin_Peuch Ideally. But I'm not certain this could work here.
15:01 Benjamin_Peuch We are an old, cultural heritage institution, you see. :x
15:01 Benjamin_Peuch I just heard from my boss about this. She's not against the idea of a fringe meeting.
15:02 donsizemore joined #dataverse
15:02 pdurbin Great!
15:02 Benjamin_Peuch Oh I see that some fringe events are not free of charge anyway: https://fosdem.org/2020/fringe/
15:02 Benjamin_Peuch Anything goes pretty much.
15:03 Benjamin_Peuch We will have to cap the number of possible attendees, but I'm not expecting a huge crowd.
15:03 pdurbin Wow, from free to $1499.
15:03 Benjamin_Peuch (Then again, I'd love to be proved wrong.)
15:03 Benjamin_Peuch Eeyup.
15:03 pdurbin I didn't expect so many people for our group photo: https://twitter.com/philipdurbin/status/1206700431037980673 :)
15:04 pdurbin But I piggy-backed off our holiday party. :)
15:05 Benjamin_Peuch Haha, you're right in the middle!
15:05 Benjamin_Peuch What a lovely-looking dog, too. :)
15:05 Benjamin_Peuch Congratulations on your prize!
15:05 pdurbin That's Davis. He's also in a YouTube video from donsizemore
15:05 Benjamin_Peuch He's a chonky boy!
15:06 pdurbin Here's Davis: https://www.youtube.com/watch?v=y_1IdJ-nYY0
15:06 pdurbin Only a few of us are allowed to give him treats.
15:08 Benjamin_Peuch Sound policy. I can imagine the problems otherwise.
15:09 Benjamin_Peuch I'll come back to you with the details for the fringe event asap.
15:09 Benjamin_Peuch If I may change the subject, my colleague Youssef noticed that the "Terms" section of Dataverse seems to work quite differently from "Metadata".
15:09 Benjamin_Peuch Why is that?
15:10 pdurbin Different how?
15:10 pdurbin (not that I disagree)
15:10 Benjamin_Peuch It can't be modified via a tsv file like the "Metadata" section.
15:12 pdurbin You can't customize it, you're saying. Do you want different licenses? CC-BY or whatever? Or something else?
15:14 pdurbin What is your use case?
15:15 Benjamin_Peuch Indeed the possibility of offering different licenses would be great.
15:15 Benjamin_Peuch Making CC-BY the default choice for instance would be ideal, and not just for us I think, since this is a widely used license in the academic world.
15:16 Benjamin_Peuch (Whereas CC0 is pretty extreme and, I would assume, not fit for most researchers and research centers.)
15:16 Benjamin_Peuch One of our project partner, a demographer, also told us that he found the "Terms" section a little intimidating.
15:17 Benjamin_Peuch Because it's very long and it has so many fields, some of which are not always easy to distinguish.
15:17 Benjamin_Peuch I was planning to voice his concern as an issue on GitHub.
15:17 pdurbin Not just for you is right. There is a lot of interest in making Terms more flexible. I could point you to exisiting GitHub issues but it would probably be better for you to open a new issue with details about your use cases.
15:18 pdurbin Multiple new issues.
15:18 Benjamin_Peuch Really? We wouldn't want to spam the Dataverse GitHub.
15:20 pdurbin GitHub issues are how we gather feedback. And feedback is a gift.
15:21 pdurbin "Terms are long and intimidating" sounds fine to me. With a screenshot to remind eveyone how many fields there are.
15:24 Benjamin_Peuch I'll do just that then. Thanks :)
15:26 pdurbin thanks!
15:27 pdurbin We try to keep the "create dataset" page somewhat short. Because otherwise it's too much to read, too many fields. But then authors can click "edit metadata" and add extra fields if they want to.
15:27 Benjamin_Peuch We thought that was the logic behind the split in different metadata pages.
15:28 Benjamin_Peuch On the one hand, it means extra clicks, to later go fill in all the other metadata fields.
15:29 Benjamin_Peuch On the other hand, it's fewer risks of losing the information a user started punching in by inadvertently closing their browser for instance.
15:29 pdurbin Yeah, it's controlled by "displayoncreate" at http://guides.dataverse.org/en/4.18.1/admin/metadatacustomization.html
15:38 Benjamin_Peuch Thanks for the tip. :)
15:39 pdurbin Sure, you could also make a field required, if it's important. There's a UI for this.
15:40 Benjamin_Peuch Ah yes indeed, in the admin panel.
15:40 Benjamin_Peuch Being able to manipulate the Terms fields in that same panel would be handy.
15:41 pdurbin Absolutely. It sounds like you might hide some fields. So users aren't overwhelmed.
15:43 Benjamin_Peuch Yes, that's what we had in mind.
15:43 Benjamin_Peuch Another question if you don't mind.
15:43 pdurbin I don't mind. I'm just catching up on my email.
15:43 Benjamin_Peuch Haha good luck.
15:44 Benjamin_Peuch When I'm viewing a dataset's metadata, if I click on the Contact button at the top of the page, this will send an email to the address that the data depositor specified.
15:44 Benjamin_Peuch But if I click "Request Access" for files with restricted access, this will rather notify the Dataverse admin. Is that correct?
15:48 pdurbin Yes, that's exactly right.
15:48 pdurbin Is that weird? :)
15:49 Benjamin_Peuch Not necessarily.
15:50 Benjamin_Peuch But that means that the underlying philosophy is that one or more Dataverse administrators should handle the requests for access, and not (only) the dataset authors themselves?
15:51 Benjamin_Peuch I know that in some data archives, for specific (especially sensitive) datasets, permission from the authors but also from the data archive's director must be obtained.
15:52 pdurbin Hmm, in that case, I guess you could configure the dataset so that only the director can approve access. Also, please don't put sensitive data in Dataverse yet. We're working on this feature. :)
15:54 Benjamin_Peuch No worries. This wasn't the idea. :D
15:55 Benjamin_Peuch What we had in mind was rather outsourcing the handling of requests for access to the dataset authors entirely.
15:56 pdurbin But you still want the director to approve also? Even if the author says it's ok? Sorry if I'm not following you. My head is in a bunch of emails.
16:00 Benjamin_Peuch On the contrary, we would rather that only the person whose email address is referenced in the "Contact" section of the metadata be notified of requests for access.
16:00 Benjamin_Peuch That they alone (or along with other contact persons mentioned, since the field can be duplicated) receive email when someone clicks "Request Access".
16:00 Benjamin_Peuch Good luck with the emails. Thank you for your time. :)
16:02 pdurbin Hmm, maybe you could write a script to make sure that the contact emails match the emails of the people who can approve requests.
16:03 pdurbin Maybe this is obvious but the contact doesn't even have to have a Dataverse account. The person who can approve accounts does have to have an account.
16:06 Benjamin_Peuch Indeed. But likewise, it's more important for depositors to have an account than for people who want access to data, right?
16:07 Benjamin_Peuch Since requests for access can be sent without being logged in.
16:08 pdurbin Well, if you click "Request Access" when you aren't logged in, you are prompted to create an account, right?
16:09 pdurbin Yeah. "You need to Sign Up or Log In to request access to this file." In a popup.
16:09 pdurbin So you have to log in.
16:12 Benjamin_Peuch Hu, you're right. My bad.
16:13 pdurbin No worries. The account is important so that access can be revoked in the future, if necessary.
16:14 Benjamin_Peuch Indeed. It makes sense.
16:28 Benjamin_Peuch I must be on my way.
16:28 Benjamin_Peuch Thanks a lot for the help, as always!
16:29 Benjamin_Peuch91 joined #dataverse
16:29 Benjamin_Peuch91 I must be on my way. Thanks a lot for the help pdurbin, as always. :)
16:29 Benjamin_Peuch91 Until next time!

| Channels | #dataverse index | Today | | Search | Google Search | Plain-Text | plain, newest first | summary

Connect via chat.dataverse.org to discuss Dataverse (dataverse.org, an open source web application for sharing, citing, analyzing, and preserving research data) with users and developers.