IQSS logo

IRC log for #dataverse, 2020-01-27

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
07:44 jri joined #dataverse
08:02 pkiraly joined #dataverse
08:14 juancorr joined #dataverse
08:18 pkiraly @poikilotherm, Could you tell me again the name of the shell you use? (I forgot it)
08:22 jri joined #dataverse
08:59 stefankasberger joined #dataverse
09:01 poikilotherm Heipkiraly
09:02 poikilotherm I use "Tilix" as kinda window manager
09:02 poikilotherm Inside that I use zsh
09:02 poikilotherm Tweaked with https://github.com/Powerlevel9k/powerlevel9k
09:03 poikilotherm Hehe I just learned about powerlevel10k :-D
09:03 poikilotherm https://github.com/romkatv/powerlevel10k
09:19 pkiraly Thanks a lot!
09:19 pkiraly Thanks a lot, Oliver!
09:54 jri joined #dataverse
10:47 Benjamin_Peuch joined #dataverse
11:19 pdurbin joined #dataverse
11:21 pdurbin So how was #dataverse2020? :)
11:43 pdurbin Were any talks recorded?
11:59 poikilotherm pdurbin: no records AFAIK
12:05 pdurbin Any notes?
12:06 juancorr joined #dataverse
12:07 poikilotherm Yeah - sending via Twitter as I have no idea if they may be shared...
12:08 poikilotherm About that sociocracy thing: let me ask my colleague
12:08 poikilotherm BTW Slava is really interested in that topic
12:08 poikilotherm We chatted with Gustavo, him and others about it
12:09 pdurbin What sociocracy thing?
12:10 poikilotherm That blog article you asked for. Sorry, Twitter seems to have not send me a notification :-/
12:10 poikilotherm Just saw it when I send you the notes link
12:10 poikilotherm s/send/sent/
12:10 poikilotherm O.O
12:10 poikilotherm Forgive me
12:11 poikilotherm I'm so stupid
12:11 poikilotherm Tomatos on my eyes. All red over here.
12:11 poikilotherm facepalm
12:11 poikilotherm Sure, please send me your blog post link about SLOPI
12:12 poikilotherm Monday morning... Let's pretend this never happened
12:14 Benjamin_Peuch Did you kill somebody by mistake, poikilotherm? :o
12:14 poikilotherm Yeah, my brain.
12:14 pdurbin Maybe I haven't had enough coffee yet because I'm not following very well but sure, I'd love feeback on my new blog post about SLOPI! I'm trying to make it the canonical thing to read:
12:15 pdurbin Searchable Linkable Open Public Indexed (SLOPI) Communication or Why open source projects should avoid Slack: http://blog.greptilian.com/2020/01/25/slopi-communication/
12:16 pdurbin "2 minute read" :)
12:16 poikilotherm I love it. It's to the point with nice examples.
12:17 poikilotherm You could ask your kids to create a nice picture for it :-)
12:18 pdurbin My daughter helped me with the logo. :)
12:18 poikilotherm Nice!
12:20 pdurbin Actually, I was having trouble figuring something out in Inkscape and was about to google for the answer but she figured it out. She's 10.
12:28 poikilotherm https://twitter.com/poi_ki_lo_therm/status/1221771868803346433
12:31 yoh joined #dataverse
12:39 pdurbin poikilotherm: thanks!
12:40 poikilotherm It's such a nice and catchy term for it :-)
12:44 pkiraly I am trying to install the data-capture-module (https://github.com/sbgrid/data-capture-module) to Ubuntu 18.04. It is not clear which python version it requires. Is there anyone of you happens to know it?
13:01 pdurbin_m joined #dataverse
13:02 pdurbin_m pkiraly: I'm pretty sure Pete was on CentOS 6.
13:08 pkiraly It is not clear to me if CentOS 6 uses Python 2 or 3. I tried 3, but runs into errors. The pip installation worked for me only for Python 2.
13:16 pdurbin Hmm. I think he was using Python 2. He's quite responsive on GitHub. Please open an issue.
13:17 pkiraly pdurbin: OK, thanks!
13:18 pdurbin Sure. Please bear in mind that all the DCM stuff is pretty experimental. Even Pete's installation isn't on the map yet. But it's up and I think it's all still working.
13:21 pkiraly pdurbin: is there an alternative way for large files upload with rsync?
13:22 pkiraly From the documentation it is not clear at all how it work. My guess is that it check the file upload and when it is done, it triggers some Dataverse API, but maybe I am wrong.
13:24 pdurbin pkiraly: I mean, you could build your own DCM equivalent. You're right, at the end, once the files have been uploaded via rsync (or whatever), a Dataverse API is called to make Dataverse aware of the "package". A package looks like a single file in Dataverse but it represents a directory of files.
13:28 pdurbin Once we get Pete's installation on the map (or if we ask him nicely), we can link to it so you can see what a "package" looks like. :)
13:30 pkiraly pdurbin: thanks a lot!
13:32 donsizemore joined #dataverse
13:33 pkiraly The last commit of the project was submitted a year ago.
13:45 Benjamin_Peuch Very nice post, pdurbin.
13:45 Benjamin_Peuch I don't have enough experience to give much feedback, but I liked the example with the donuts. :)
14:03 donsizemore @pkiraly regarding PR #6083: you were able to pull the JSON representation of an unpublished Dataverse without an API token?
14:10 pdurbin Everyone loves donuts.
14:11 pdurbin Benjamin_Peuch: any sense yet of how many people will be there on Friday? Just you and me? :)
14:13 pdurbin poikilotherm: I took a quick look at https://www.sociocracyforall.org and it's coll that they're headquarted less than a 2 hour drive away but I'm not sure I completely understand what it has to do with SLOPI. Are you the expert on Sociocracy within the Dataverse community? Or did you hear about it from someone else?
14:15 Benjamin_Peuch pdurbin: My colleague Freya, who's a jurist, and my manager, Rolande Depoortere, a digital archivist, will be there. And from the outside: Dorothea Iglezakis, from Stuttgart University, and Patrick Vranckx, from Université catholique de Louvain.
14:15 pdurbin Oh! It'll be nice to meet Dorothea! She's working on a software metadata block.
14:16 MrK joined #dataverse
14:16 poikilotherm pdurbin: https://twitter.com/poi_ki_lo_therm/status/1215413129942794241 I'm definitely not an expert, just a hack/follower
14:20 pdurbin poikilotherm: thanks, I'm with Slava in that I've heard of Holacracy.
14:22 pdurbin I heard of Holacracy at https://wiki.snowdrift.coop/governance but now I see that Sociocracy is mentioned there as well but I don't remember if it was there before or not.
14:23 pdurbin pkiraly: Pete wrote me back and said he'll try to jump in to this channel later, hopefully not too late for you. Time zones. :)
14:23 poikilotherm Sociocracy and SLOPI IMHO are connected because you will want to make decisions done more transparently and inclusive. SLOPI is pretty good at that
14:27 pdurbin Sure. I did see the word "inclusive" in there. SLOPI is definitely inclusive.
14:28 pdurbin Speaking of SLOPI, any chance the notes from #dataverse2020 can be posted to the mailing list? :)
14:28 donsizemore does Dataverse have an official Code of Conduct? https://confcodeofconduct.com/
14:28 poikilotherm pdurbin Benjamin_Peuch I just asked my boss if I should visit you guys on Friday and not come to Brussels on Sunday then. I'll see what he says.
14:28 pdurbin donsizemore: nope, and GitHub dings us for not having one at https://github.com/IQSS/dataverse/community
14:29 Benjamin_Peuch That's great news, poikilotherm. :D
14:29 pdurbin Benjamin_Peuch: not it's not. I want poikilotherm to come to my talk. ;)
14:29 Benjamin_Peuch Oh right, your talk is on Sunday?
14:29 pdurbin yeah, 2pm
14:29 pdurbin 14:00
14:30 poikilotherm LOL
14:30 poikilotherm My boss said I should come on Friday :-D
14:30 poikilotherm Phil is your talk recorded / streamed?
14:30 pdurbin I practiced it yesterday. Feeling pretty good about it. I need to talk faster or cut out some content though. :)
14:30 pdurbin Yes, it should be recorded.
14:30 pdurbin I'm planning on putting it on https://iqss.github.io/dataverse-tv/
14:31 poikilotherm So I'd have a chance to still see it then :-)
14:31 poikilotherm I'd really love to see all of you guys.
14:31 pdurbin yeah, but no chance to heckle me :)
14:31 MrK pdurbin: Yeah, conferences have pretty tight schedule. We have the group photo for you :P
14:31 pdurbin MrK: where? Let's see a tweet. :)
14:32 MrK pdurbin: It's Stefan's photo, I can just link the pciture
14:32 poikilotherm pdurbin: LOL. No way I'd do that... I behave myself ;-)
14:33 pdurbin stefankasberger: pics or it didn't happen
14:34 poikilotherm I swear we did it. Asked that nice lady doing the conference organization to help us out :-)
14:34 pdurbin good
14:35 poikilotherm Benjamin_Peuch: it was Bruxels Midi, right?
14:38 pdurbin poikilotherm: "not far from Brussels Central Station" according to http://www.arch.be/index.php?l=en&m=news&r=agenda&e=fosdem-20-dataverse-fringe-meeting
14:38 Benjamin_Peuch poikilotherm: No: Gare centrale / Centraal Station.
14:38 Benjamin_Peuch It's not the same as Bruxelles-Midi / Brussel-Zuid.
14:38 pdurbin Benjamin_Peuch: oh, if you want, you could update that announcement to link to my talk on Sunday: https://fosdem.org/2020/schedule/event/dataverse/
14:38 Benjamin_Peuch (Name are in French and in Dutch because those are the official languages of the Brussels region.)
14:39 Benjamin_Peuch pdurbin: You're right. I will do that right away.
14:39 pdurbin thanks!!
14:40 pkiraly pdurbin: thnkas!
14:40 Benjamin_Peuch poikilotherm: I can forward to you some practical information I sent to Philip regarding mobility in Brussels.
14:40 poikilotherm Benjamin_Peuch: I could be either late or early. Have 2 connecting trains from Düren to Aachen, then Brussels Nord to Central and continue on foot.
14:40 Benjamin_Peuch Train tickets and all that.
14:41 pkiraly donsizemore: let me check. I did not tested all the API calls, just gave a format change.
14:41 Benjamin_Peuch It's okay if you arrive at a later point, poikilotherm. :)
14:41 Benjamin_Peuch The more, the merrier.
14:42 poikilotherm Meh. I'd be there by 14:00 the earliest. I think I'll just take the earlier trains then and enjoy some lunch. (Would be there by 11:34)
14:43 MrK pdurbin: https://drive.google.com/file/d/1kaaj8d-Jr9LtMofMsCf5dWKklhXLMki-/view?usp=sharing
14:43 Benjamin_Peuch And in case you don't like what you found to eat, there will be sandwiches, poikilotherm.
14:44 Benjamin_Peuch Open source wizard!
14:44 pdurbin MrK: thanks!!
14:45 poikilotherm Benjamin_Peuch: do you know what happens with my ticket when I miss a train in Brussels Nord?
14:45 poikilotherm It's just a 5 minute time slot for changing to the connecting train
14:47 pkiraly donsizemore: you are right, I made a mistake there.
14:47 pdurbin MrK pkiraly poikilotherm stefankasberger: https://twitter.com/philipdurbin/status/1221806715429163008 :)
14:50 poikilotherm Man DB is just crazy. From Brussels Zuid/Midi it's just a short walk to you. Instead they want me to get on a connecting train from Nord to Central. Pfff....
14:54 Benjamin_Peuch poikilotherm: For inland trips, you can take any train. You are not bound to a particular train.
14:54 Benjamin_Peuch So you can take the next one without fear. :)
14:55 poikilotherm Is this true also for german tickets by Deutsche Bahn?
14:56 pkiraly donsizemore: I added an extra commit to fix the problem.
15:01 pdurbin pkiraly: have you seen https://github.com/IQSS/dataverse-kubernetes/issues/68 ?
15:03 pkiraly pdurbin: thanks, I was not aware of it
15:04 pdurbin pkiraly: sure. You might find this (and all the Docker files) helpful as well: https://github.com/IQSS/dataverse/blob/v4.19/conf/docker-dcm/readme.md
15:06 MrK poikilotherm: Btw are you a solo developer in your facility working on Dataverse?
15:07 poikilotherm MrK: yes, just me :-)
15:09 pdurbin Notes from #dataverse2020 at https://drive.google.com/drive/folders/1tlP0mAaSf3S3NyQUOHkqU46kVwrJmWHJ via https://groups.google.com/d/msg/dataverse-community/HtAzybL5SiY/9xWjenACEQAJ thanks to pkiraly who mentioned posting them at https://twitter.com/kiru/status/1221810619265044482 (thanks!!!)
15:12 pdurbin Awesome that DataverseTV is in the notes. :)
15:12 pkiraly pdurbin We mentioned you several times.
15:12 pdurbin !
15:15 pdurbin poikilotherm: not many notes from your talk. :(
15:15 poikilotherm No one asked a question. Maybe it was just too much.
15:16 pdurbin You're from the future.
15:17 pdurbin Heh. "Tell us what do you think when you hear the word Dataverse"
15:18 pdurbin nice. "Portugal - One in production and another almost ready also"
15:19 pdurbin "Latvia - some institutions started to set up DV"
15:19 poikilotherm :-D
15:19 poikilotherm We'll get there
15:20 poikilotherm For now this was a first public announcement of the project
15:20 pdurbin Sure. It's advanced alien technology. Eventually we'll learn to use it.
15:21 poikilotherm Get people to know that this exists
15:21 poikilotherm In 2020 there will be at least 2 production instances running on it
15:21 poikilotherm Jülich DATA and DARIUS @ Stuttgart
15:21 poikilotherm Dunno if SSHOC project will also create production instances within EOSC this year.
15:21 Benjamin_Peuch poikilotherm: Contacted you in PMs about your travelling arrangements.
15:21 donsizemore @pdurbin there are three new dataversia around the corner in singapore
15:22 pdurbin donsizemore: yes! So I hear!
15:22 Benjamin_Peuch pdurbin: They're considering setting up a Dataverse instance at Université catholique de Louvain apparently.
15:23 Benjamin_Peuch The person responsible for this, Patrick Vranckx, will be attending on Friday.
15:23 pdurbin Benjamin_Peuch: yes, I remember the conversation at http://irclog.iq.harvard.edu/dataverse/2019-06-25#i_98617
15:24 pdurbin patrick33 :)
15:24 Benjamin_Peuch Oh I see. They've been going at it for longer than we thought then. :o
15:24 pdurbin :)
15:25 Benjamin_Peuch We have begun configuring Shibboleth at the State Archives so we can work with universities as an SP.
15:25 Benjamin_Peuch They apparently also considered Dataverse at KIK-IRPA - the Royal Institute for Artistic Heritage, but last we heard they were moving towards Invenio instead.
15:28 poikilotherm Benjamin_Peuch: PM on IRC?
15:28 pdurbin Sure, the thing Zenodo is built on. As long as they're publishing data, I'm happy. :)
15:29 Benjamin_Peuch poikilotherm: Private messages, here on Shout.
15:29 Benjamin_Peuch Haha, that's the spirit pdurbin.
15:29 Benjamin_Peuch (We know Dataverse is better tho, obviously.)
15:29 pdurbin Benjamin_Peuch: check this out: https://github.com/IQSS/dataverse.harvard.edu/issues/56
15:29 pdurbin who knows if we'll do it or not though :)
15:30 Benjamin_Peuch Do you want to harvest the data (and the metadata) or just the metadata?
15:30 pdurbin just the metadata to make it searchable
15:30 pdurbin discoverable
15:31 Benjamin_Peuch That's what I thought. No need to duplicate it. :)
15:32 Benjamin_Peuch Then again, now I'm wondering... When we say "harvest data", don't we generally mean "collect the data's metadata"?
15:32 pdurbin One of the ideas in the air is that people could come along and create a dataverse and then link some of these datasets into that dataverse they create. Basically, it would be a way for them to flag data as important or interesting. Even if they don't upload any original data.
15:32 Benjamin_Peuch That's what portals do in general.
15:32 Benjamin_Peuch pdurbin: Indeed, grow a large network of interconnected Dataverse and thus increase the chances of the metadata popping up in the queries of users.
15:33 pdurbin If you look at https://dataverse.harvard.edu/dataverse/harvard you can see that Harvard Dataverse already harvests ~63,000 datasets. Vs. ~35,000 original datasets.
15:33 Benjamin_Peuch Yes. That's a very impressive figure. :)
15:33 pdurbin 2 million is even more impressive :)
15:34 pdurbin But I wonder if the non-US gov data would be drowned out. Dunno.
15:34 pdurbin Anyway, it's not my idea. All I did was check if CKAN supports OAI-PMH or not. :)
15:34 pdurbin and it does, which is cool
15:35 pdurbin The conversation is here: https://twitter.com/philipdurbin/status/1220742303184089088
15:40 poikilotherm Benjamin_Peuch: those PMs don't seem to work. Feel free to reach out on Twitter or by mail
15:42 pdurbin poikilotherm: you and your Matrix. Or maybe Benjamin_Peuch would have to register with nickserv (what a pain).
15:48 poikilotherm Guys gotta run pick up kiddos
15:48 poikilotherm Cu
15:52 Benjamin_Peuch Sent you an email, poikilotherm.
16:18 pameyer joined #dataverse
16:19 pameyer @pkiraly - are you still on?
16:24 pameyer looks like I lose to time zones...
16:35 pkiraly pameyer, I had to run out now. I will check back later today
16:35 pdurbin pameyer: time zones. Thanks for stopping by! :)
16:36 pdurbin pameyer: am I allowed to link to your installation here even if it's not on the map?
16:38 pameyer pdurbin: might be better to link to the staging site?
16:38 pdurbin Oh! Where?
16:38 pameyer pkiraly: no problem, I saw https://github.com/sbgrid/data-capture-module/issues/36
16:39 pameyer https://dv.sbgrid.org/
16:39 pameyer pdurbin: how's your trip prep going?
16:40 pdurbin well, I feel more prepared for a talk than ever, having practiced it yesterday, but I'm still packing :) looks like it'll be in the 50s and 60s with rain
16:43 pdurbin and thanks for the link to your staging site
16:47 pameyer no problem - thought you had it already
16:47 pdurbin I remembered the other two. prod and future prod. :)
16:49 pdurbin seems like interest in DCM is heating up
16:54 pameyer it does - not sure if Thanh-Thanh is sorted or not though
16:56 pdurbin Not sure. She opened that issue shortly after Danny and Sonia visited.
16:59 pameyer the wisdome of a "system status" / "health check" API endpoint in the code came late to me
17:00 pdurbin :)
17:01 pdurbin pameyer: oh, did you see this DCM thread? https://groups.google.com/d/msg/dataverse-community/3TNkfmj75pY/MrOUbba1CAAJ
17:02 pameyer I hadn't
17:02 pdurbin I'm not sure if you met Michel but he was at last year's community meeting. Nice guy.
17:03 pameyer ... apparently "read the jenkins pipeline and find the test script" is not a great UX :(
17:03 pameyer don't think I met him; I missed most of last year's community meeting
17:04 pdurbin He also installed Geoconnect. He's quite brave.
17:06 pameyer that is impressive
17:06 pdurbin :)
17:07 pameyer I've forgotten the username, but if I'm remembering correctly there was someone trying to install the DCM in it's ansible only / don't install this unless you're pete days....
17:07 pameyer that was impressive
17:07 pdurbin Right, Rebecca Barros, I think.
17:08 pdurbin man, that was a while ago: http://irclog.iq.harvard.edu/dataverse/2017-10-30#i_59272
17:08 pdurbin DCM is getting old. ;)
17:09 pdurbin Hey, do we have a sequence diagram for DCM? Something like http://guides.dataverse.org/en/4.19/admin/make-data-count.html#architecture
17:09 pameyer I don't think there's a sequence diagram
17:09 pdurbin ok
17:10 pameyer I tend more towards scrap paper, for better or worse
17:10 pdurbin :)
17:10 pameyer but I'm pretty sure that there are links to some presentations with system architecture slides
17:10 pameyer cartoon arrows are kinda like a sequence diagram ... or not
17:11 pdurbin cool, if you find a diagram maybe attach it to a comment on https://github.com/IQSS/dataverse-kubernetes/issues/68
17:13 pameyer cool
17:15 pameyer not 100% sure about kubernetes networking and rsync thought - I remember being suprised by hostname header requirements to get networking going, and rsync protocol doesn't have that header
17:15 pameyer and poikilotherm's not _wrong_ about sha1 ; but I haven't seen a pre-image collision yet
17:16 pdurbin If anyone can get rsync working in kubernetes it's poikilotherm.
17:17 pdurbin SHA-1 is alive and well in git, at least. :)
17:25 poikilotherm pdurbin about git & sha256 things are moving... https://github.com/git/git/blob/master/Documentation/technical/hash-function-transition.txt
17:26 pdurbin Huh, news to me. Or I forgot. Thanks!
18:03 pdurbin poikilotherm: oh, you and others might be interested in this new podcast: https://www.pubhouse.net/2019/12/episode-1-lets-break-into-microprofile.html
18:58 donsizemore joined #dataverse
19:59 donsizemore @pdurbin knock knock?
20:33 pkiraly joined #dataverse
20:35 pkiraly pameyer: thanks for the answer at https://github.com/sbgrid/data-capture-module/issues/36#issuecomment-578836194
20:37 pkiraly pameyer: I have a general question regarding to the workflow: when the user upload files via rsync the file will remain on the dcm server, or it is copied to the Dataverse storage?
20:50 pameyer pkiraly: once the checksums get verified, they get moved to the shared DCM/Dataverse storage
20:51 pameyer if the files don't pass verification, they don't get moved
20:51 pkiraly Fine, thanks a lot!
21:28 pameyer no problem - happy to see people taking a look into it
21:57 pameyer donsizemore: I'm guessing that pdurbin is in transit
22:52 pdurbin pameyer: yeah, flight boarding soon. Red eye.
22:52 pdurbin I'll guess I'll disconnect for a while. See you all on the other side.
22:52 pdurbin left #dataverse

| 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.