IQSS logo

IRC log for #dataverse, 2018-07-19

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
00:52 jri joined #dataverse
03:52 jri joined #dataverse
06:30 juancorr joined #dataverse
06:57 jri joined #dataverse
07:59 jri_ joined #dataverse
11:15 jri joined #dataverse
13:22 bricas pdurbin: is there any way to simplify the login screen if there's only ever 1 institution that can be selected?
13:48 pdurbin bricas: unfortunately, no. Check out how this installation forked the code: https://dataverse.nl . Do you want to create a GitHub issue for this?
13:50 bricas yeah, i'll add an issue. thought i'd ask first. :)
13:50 bricas btw, is the master branch the latest stable release?
13:51 bricas (the branches may already be doc'ed somewhere, sorry if i've missed that)
13:51 pdurbin bricas: yep: http://guides.dataverse.org/en/4.9.1/developers/version-control.html#the-master-branch and don't worry about not knowing it's documented. That's my job. :)
13:51 pdurbin Your job is to ask. :)
13:51 bricas ah perfect. thanks.
13:53 pdurbin Sure. This installation also has a single installation in the dropdown: https://researchdata.ntu.edu.sg
13:55 pdurbin There might be others. Not sure.
14:29 bricas pdurbin: if i just delete the orcid row from authenticationproviderrow will that get rid of the orcid button?
14:29 pdurbin bricas: you should use the API
14:32 pdurbin Hmm, but maybe it isn't documented.
14:33 bricas ...was looking too :)
14:33 bricas it mentions the db table a few times (hence my first question)
14:33 pdurbin It should be similar to the POST at http://guides.dataverse.org/en/4.9.1/installation/oauth2.html but a DELETE
14:35 pdurbin bricas: here's the code: https://github.com/IQSS/dataverse/blob/v4.9.1/src/main/java/edu/harvard/iq/dataverse/api/Admin.java#L254
14:36 bricas roger dodger
14:36 pdurbin You could do a GET on the id number before and after.
14:36 pdurbin To make sure you got the right one.
14:36 pameyer joined #dataverse
14:36 bricas got it.
14:36 bricas worked as described.
14:38 pdurbin bricas: great! Do you mind creating an issue to document this?
14:38 bricas sure thing.
14:38 bricas i have issues, y'all.
14:38 pdurbin bricas: oh, wait. I just found it.
14:39 pdurbin http://guides.dataverse.org/en/4.9.1/api/native-api.html#delete-an-authentication-provider
14:39 pdurbin bricas: but if you feel like any of those docs are off, please feel free to open an issue.
14:40 pdurbin bricas: or open a issue to link from other pages to those docs or something
14:40 bricas i think maybe pointing to the api vs talking about table rows may be better
14:40 pdurbin yeah, absolutely
14:41 pdurbin And I've been thinking about moving a lot of those "admin" docs out of the native api page and into specific sections where they are much more relevant. The OAuth2 page in this case.
14:41 pdurbin Well....
14:41 pameyer yeah, I tend to treat going into the db as "I probably shouldn't be doing this" terratory
14:41 pdurbin maybe the Admin guide.
14:41 pdurbin because shib and builtin
14:41 pdurbin I don't know
14:41 pdurbin yeah, good way to treat it
14:43 bricas direct db interaction is usually "hic sunt dracones" stuff :)
14:44 pdurbin yeah, tread lightly
14:56 pdurbin bricas: thanks for creating the "Simplified login for single remote auth endpoint" issue: https://github.com/IQSS/dataverse/issues/4869
14:56 pdurbin and the "Config docs should mention the API and not DB tables" issue: https://github.com/IQSS/dataverse/issues/4870
14:57 pdurbin and for offering to work on the "Documentation enhancement for Apache config for mod_shib on Ubuntu" issue: https://github.com/IQSS/dataverse/issues/3213
14:57 bricas <thumbs up emoji>
14:57 pdurbin Have you see the new "Community Dev" column at https://waffle.io/IQSS/dataverse ?
14:57 donsizemore joined #dataverse
14:58 bricas i've not. i've never seen waffle.io at all actually
14:59 pdurbin we recently put a waffle badge at the bottom of https://github.com/IQSS/dataverse
15:00 pdurbin 14 in the "Community Dev" column. Do you want me to assign you to any of these issues and drag it over into that column? No pressure. :)
15:01 bricas heh. if i'm being honest, i have 60 other things i'm supposed to be getting done this summer, so i'll have to take a pass. but if they sit idle long enough, i may re-address them when i get some time.
15:02 pdurbin Ok, no worries. :)
15:02 pdurbin I predict they will sit idle. :)
15:03 bricas i'm not a java guy, so any real coding tasks are out of my wheelhouse -- but documentation i can do.
15:04 pdurbin Sure. All contributions are appreciated, of course.
15:13 pdurbin Including bug reports. :)
15:23 bricas of course. :)
15:26 donsizemore joined #dataverse
15:39 pameyer48 joined #dataverse
16:56 donsizemore joined #dataverse
17:51 bricas pdurbin: basic PR done for the shib apache 2.4 docs
17:53 pdurbin bricas: huh. I forgot about that "auth" branch. Thanks!
17:53 bricas np!
17:55 pdurbin jeez, it's a 404
17:56 pdurbin bricas: I take it you found that shib.md file more helpful than http://guides.dataverse.org/en/4.9.1/installation/shibboleth.html ?
17:59 bricas i can't honestly remember what i did at the time... i was just searching for shib and found those busted urls
18:00 pdurbin Cool. I moved it to QA. Thanks for the pull request!
18:02 bricas no prob. my original ticket was ubuntu "specific" (just because that's what we use) - but i'm sure the issue is really just apache version specific
18:03 pdurbin Ah, ok. I'm always glad to hear that Dataverse works on Ubuntu, by the way. Especially the shib stuff, which I don't believe is supported by the shib folks. They seem to favor redhat.
18:04 pdurbin And so do we, to be honest.
18:11 donsizemore joined #dataverse
18:12 donsizemore hey @pdurbin speaking of shib, want me to start an issue to track migration to shib 3? it's out and our ID MGMT head is on vacay, but once he's back I may see if I can grab him for a consultation
18:13 pdurbin donsizemore: huh, I had no idea is was out. Sure! Any idea how long shib 2 will be supported?
18:13 pdurbin fingers crossed for "forever" :)
18:14 donsizemore the question will be coordinating the use of R&S / federation with the upgrade to 3.0
18:15 pdurbin yeah that still needs to work
18:15 pameyer48 is the version of shib that dataverse ships with the latest shib 2?
18:16 pdurbin pameyer48: we just say to do a yum install from the shib yum repo
18:16 pdurbin or whatever bricas is doing on ubuntu :)
18:19 pameyer48 right - but there's shib-related javascript that ships with dataverse
18:20 pdurbin oh, yes. buh, whatever that thing is called
18:21 pdurbin This thing but for me it didn't come in a box: https://www.shibboleth.net/products/embedded-discovery-service/
18:21 pameyer48 the postive control for anything checking "don't use eval in javascript on arbritrary input"
18:22 pdurbin bricas: so really https://github.com/IQSS/dataverse/issues/4869 could be filed against EDS
18:22 pdurbin but we forked it so we kind of own it :/
18:23 donsizemore @pamayer48 yeah, that'll be interesting when somebody wants to stand up a new shib.
18:27 pameyer48 @donsizemore it was on my rador more for paranoia than new installs; but that's a good point
18:27 pameyer48 s/rador/radar/g
18:35 pdurbin donsizemore: speaking of upgrades, yesterday in our bi-weekly backlog grooming meeting I wanted to ask the team to estimate https://github.com/IQSS/dataverse/issues/4172 about payara 5 but didn't get the opportunity
18:41 bricas we intend to dockerize our install at some point, once we get the tuits. we just happen to use ubuntu for our regular WMs is all.
18:42 pdurbin WMs or VMs?
18:42 bricas VMs
18:42 bricas sorry.
18:42 pdurbin phew
18:42 bricas one thing i like in slack is i can edit messages so i look smart(ish)
18:43 pdurbin yeah IRC is like SMS
18:48 pameyer48 @bricas we're on a mix of centos and ubuntu (centos for dataverse thought)
18:49 pdurbin my home server is ubuntu
18:50 bricas we run alpine on (almost?) all of our dockerized things
18:51 bricas ubuntu otherwise
18:51 pdurbin ah, based on BusyBox
18:53 pameyer48 in my hands, trying to dockerize stuff into alpine can be ... interesting
18:53 pameyer48 starting from it's simpler
18:56 bricas our needs have been relatively simple, so it's been great
19:10 pdurbin bricas: dockerize your install of Dataverse? Have you seen https://github.com/IQSS/dataverse/issues/4665 ?
19:23 cthao38 joined #dataverse
19:24 cthao38 left #dataverse
19:24 cthao98 joined #dataverse
19:24 cthao98 Hi, is anyone familiar with running Solr on Kubernetes containers? On the slave's solrconfig.xml, I specified the masterURL but solr returns this error `Master at  is not available. Index fetch failed by exception: java.lang.IllegalArgumentException: Illegal character in scheme name at index 0: 172.12.0.6:8983/solr/#/collection1/r​eplication?command=indexversion&wt=j​avabin&qt=%2Freplication&version=2`
19:25 cthao98 I asked the same question on solr irc but no response so far
19:25 cthao98 :/
19:27 pdurbin cthao98: I'm not but does a master/slave setup work for you outside of kubernetes?
19:27 cthao98 I haven't tried that yet, I don't know if solr would let me run 2 instances on one machine?
19:28 pdurbin I think you can just run them on different ports.
19:28 pdurbin Back in Solr 4 days there was an example using Zookeeper. I don't know if they still use that for Solr 7.
19:29 cthao98 I think the goal is to avoid zk so we can minimize the complexity of the concept
19:30 pdurbin ok
19:30 pdurbin I don't know zookeeper anyway. :)
19:30 pdurbin I mean, I tried it once.
19:30 pdurbin And I talked to a guy who uses it with Solr.
19:31 cthao98 :)
19:31 cthao98 It seems like solr irc is oddly inactive
19:31 pdurbin cthao98: you might want to try the solr mailing list. I'm on it but I never read it unless I've asked something and am waiting for a reply.
19:31 pdurbin The creator of Solr replied to me once, which was great. :)
19:32 cthao98 I just sent an email to subscribe, hoping to get some results
19:33 pameyer48 @cthao98 you can definately run multiple solr instances on the same box
19:33 pameyer48 either running normally with non-default ports, or with multiple individual containers and different port mappings.
19:35 cthao98 I'm trying it on my computer and the second solr instance is complaining that collection1 is locked
19:36 pameyer48 different directories and different ports?
19:36 pdurbin cthao98: are you following a guide? solr docs?
19:36 donsizemore joined #dataverse
19:37 cthao98 I am not because I didn't find any useful guide
19:38 cthao98 everything on the solr websites wants me to use solrcloud with zookeeper
19:40 pdurbin sounds right
19:40 pameyer48 huh - looks like it's either sharding or zookeeper
19:41 pdurbin sounds the same as solr 4 days
19:41 cthao98 I'm doing replication
19:41 cthao98 hence master-slave
19:41 pdurbin cthao98: have you looked at https://github.com/dudash/openshift-docker-solr yet?
19:43 cthao98 I don't think it uses the replicationhandler though
19:53 cthao30 joined #dataverse
20:03 rucha-oza joined #dataverse
20:06 rucha-oza we have an online journal for publishing research work. we want to provide DOI to the research articles published in our online journal. is harvard dataverse account can be used to generate DOI?
20:07 pameyer48 hi rucha-oza: if I'm understanding you correctly, I don't think so.  DOIs generated for datasets inside harvard dataverse will have landing pages that resolve to harvard dataverse
20:08 pameyer48 to issue your own DOIs for your journals, your journal would probably need an account with crossref or datacite
20:09 rucha-oza we may not mind where it is finally resolved if there is link to their article published on our journal website
20:11 rucha-oza i have looked at the datacite website. they are telling to contact (or register with) their registered members to get DOI
20:12 pameyer48 https://datacite.org/become.html looks like they've got an application
20:12 pameyer48 again, I could be wrong - but I don't think that you'd want a harvard dataverse account to issue DOIs for articles published throught your journal
20:13 pameyer48 for datasets cited by articles in your journal, definately - but those would be different DOIs
20:13 rucha-oza i have also checked crossref. but their fees is higher for us as our revenue is not much high.
20:15 rucha-oza has dataverse any kind of fee?
20:18 pameyer48 the software is open source; but having an installation connected to the handle or doi system does have costs
20:18 pameyer48 orcid does as well, not sure about shibboleth
20:19 pameyer48 for using somebody else's dataverse installation, it depends on which one.  some are restricted to institutions, others are open.
20:20 pameyer48 but I'd strongly encourage you to check with the support or curation
20:20 pameyer48 folks for another installation first
20:23 rucha-oza u said "handle or doi system does have costs" - is there any page where i can find the pricing of services? i gone through many pages but found no information about pricing
20:26 pameyer48 for pricing, I think your best bet would be to contact them and ask.  I don't know all the factors that go into subscription prices, but do know that it can vary depending on circumstances (lower costs for lower number of identifiers issued, etc)
20:27 pameyer48 I *believe* the costs for handle identifiers are lower than DOIs, but this is (somewhat obviously) not an area where I'm the best person to answer questions
20:28 rucha-oza thank you very much for all these information and your time. have a nice day
20:29 pameyer48 you're welcome - good luck
20:30 rucha-oza left #dataverse
20:39 dataverse-user joined #dataverse
20:39 dataverse-user 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.