Time
S
Nick
Message
00:13 garnett joined #dataverse
00:16 garnett joined #dataverse
02:34 bmckinney joined #dataverse
02:50 bmckinney joined #dataverse
04:17 djbrooke joined #dataverse
04:26 axfelix joined #dataverse
04:31 axfelix joined #dataverse
04:32 garnett joined #dataverse
05:25 axfelix joined #dataverse
09:16 jri joined #dataverse
11:47 bmckinney joined #dataverse
13:19 djbrooke joined #dataverse
13:40 djbrooke joined #dataverse
13:44 pdurbin
https://trello.com/c/xiTCfbfd/23-how-to-report-security-issues
13:44 pdurbin
https://trello.com/c/Ajqp6DVF/22-list-public-servers-demo-beta-etc
13:44 pdurbin
https://trello.com/c/OD000m0f/27-embed-google-group-on-community-page
13:44 pdurbin
djbrooke: I found a third card as well ^^
13:46 djbrooke
Thanks pdurbin
13:47 pdurbin
djbrooke: oh, also there's https://opsview.hmdc.harvard.edu/viewport/www
13:47 pameyer joined #dataverse
13:53 pdurbin
djbrooke: and I just thought of https://github.com/IQSS/dataverse.org/issues/60
13:59 djbrooke joined #dataverse
14:07 djbrooke joined #dataverse
14:54 garnett joined #dataverse
14:58 djbrooke_ joined #dataverse
14:59 djbrook__ joined #dataverse
15:09 djbrooke joined #dataverse
15:21 djbrooke joined #dataverse
15:24 pdurbin
djbrooke: this just in! A Dataverse installation has been added to the Research & Scholarship category! https://incommon.org/federation/info/all-entity-categories.html#SPs
15:26 pdurbin
bjonnh: can you please try to log into https://dataverse-test.irss.unc.edu with your credentials from UIC?
15:31 pdurbin
djbrooke: login now works for my buddy at MIT: https://botbot.me/freenode/positivepython/2016-06-10/?msg=67665755&page=1
15:33 pdurbin
djbrooke: as expected, HarvardKey still does not work. If I log in to https://dataverse-test.irss.unc.edu I get 'Problem with Identity Provider The SAML assertion for "givenName" was null. Please contact support.' because Harvard (by policy) doesn't release all the attributes Dataverse needs. To see what is being released (eppn only) after login you can go to
15:33 pdurbin
https://dataverse-test.irss.unc.edu/Shibboleth.sso/Session
15:40 djbrooke joined #dataverse
15:47 djbrooke joined #dataverse
16:03 bmckinney joined #dataverse
16:35 djbrooke joined #dataverse
16:41 djbrooke_ joined #dataverse
16:52 bmckinney joined #dataverse
17:03 bjonnh
pdurbin: Yay successss!
17:32 metamattj joined #dataverse
17:33 djbrooke joined #dataverse
17:39 djbrooke joined #dataverse
18:03 pdurbin
bjonnh: nice! So. Working from UNC, MIT, and UIC. Not working from Harvard or Emory.
18:18 bjonnh
great
18:18 djbrooke joined #dataverse
18:18 bjonnh
you have someone at Emory right?
18:26 pdurbin
bjonnh: someone from Emory said it didn't work. I just started a spreadsheet and put UIC down as a successful login: https://docs.google.com/spreadsheets/d/1fWHGamXetTQw3cpf6Tc4efmTvoC97lpEeTLvN7Mxz3g/edit?usp=sharing
18:28 pdurbin
bjonnh: thank you very much for testing!
18:29 pdurbin
overheard: "Oh, the joys of operating an SP in a federation." :)
18:37 pdurbin
bjonnh: ah, the Identity Provider (IdP) for the University of Illinois at Chicago has "research-and-scholarship" under "Entity Category". That's why it works. No Harvard there. No Emory. Here: https://www.incommon.org/federation/info/all-entity-categories.html
18:38 pdurbin
Only 57 of the 430 InCommon IdPs are part of the Research & Scholarship category. So it seems like in production we should filter down to just them. Well, except we want Harvard on the list.
18:38 pdurbin
djbrooke: is this making sense or should I swing by?
18:44 djbrooke
I think we're all set - just responded to the guy from InCommon
18:59 djbrooke joined #dataverse
19:23 djbrooke joined #dataverse
19:42 djbrooke joined #dataverse
20:03 bjonnh
great!
20:35 djbrooke_ joined #dataverse
20:38 djbrook__ joined #dataverse
20:41 djbrooke joined #dataverse
20:43 djbrooke_ joined #dataverse
21:20 bjonnh
I have a stupid question about dataverse
21:20 bjonnh
Let's say I use dataverse.org and publish a dataset on it
21:21 bjonnh
and at one point I convince my institution or a big national research organisation (that's actually the plan) to host their own instance
21:21 djbrooke joined #dataverse
21:21 bjonnh
is there a way to move data around while keepind the same IDs?
21:22 bjonnh
or would the plan be to convince the big org to use the dataverse.org directly
21:23 bjonnh
(contributing servers maybe?)
21:25 bjonnh
(the org is not one involved in dataverse yet)
21:26 pdurbin
bjonnh: http://dataverse.org is just the project site. To deposit data you'd have to scroll down and pick one of the Dataverse Repositories on the map. At http://dataverse.org/researchers we suggest deposting data into https://dataverse.harvard.edu
21:27 bjonnh
yeah sorry
21:27 bjonnh
that's the one I'm talking about
21:28 bjonnh
Is there some guidelines on what should go on harvard, when a new instance should be spin up
21:28 pdurbin
Well, all the datasets at https://dataverse.harvard.edu start with "doi:10.7910/DVN/" as the DOI (followed by a six character string).
21:28 bjonnh
etc
21:30 pdurbin
bjonnh: I think http://dataverse.org/institutions is pretty good at explaining when to host with Harvard and when to host on your own installation.
21:31 bjonnh
now I have even more questions ;)
21:31 pdurbin
Technically, pameyer's installation is also at Harvard but it will have different branding. :)
21:31 pdurbin
bjonnh: you can ask away and I'll check the log but I gotta get my kid to her cello recital soon.
21:32 bjonnh
1) If one instance has to be shut down due to funding issues, can the data be transfered to another instance? (updating DOIs etc)
21:32 bjonnh
ok sure
21:32 bjonnh
2) Who handles the DOI registration if I make my institution or the national organization host an instance?
21:33 bjonnh
do they have to pay for these?
21:33 pdurbin
for 1, please see http://dataverse.org/best-practices/harvard-dataverse-preservation-policy
21:34 pdurbin
for 2, you do: http://guides.dataverse.org/en/4.3.1/installation/prep.html#architecture-and-components
21:35 bjonnh
pdurbin: seems that for 1, what you sent me only covers what is deposited at harvard.
21:36 bjonnh
pdurbin: but lets say my institution decides to spin-up an instance, and then the broke and broken state of Illinois cut funding to UIC (which well, may happen in the future)
21:36 bjonnh
is there a way to transfer all of that to another instance for conservation?
21:37 pdurbin
sure, if the other instance starts paying for and hosting the DOIs
21:37 bjonnh
ok
21:38 pdurbin
hopefully that wouldn't happen very often
21:38 bjonnh
well I prefer to have these things already answered before launching that :p
21:38 pdurbin
sure. exit strategy :)
21:39 bjonnh
well our state didn't vote a budget for more than a year now…
21:39 bjonnh
so we are in a strange situation
21:40 pdurbin
our teachers have been working without a contract for quite a while
21:40 bjonnh
:s
21:41 pdurbin
not at Harvard, at my kids' school :)
21:41 bjonnh
is DOI the only (or highly recommended) applicable identifier?
21:41 bjonnh
(not from the technical point of view)
21:42 pdurbin
we're working on Handle support as well: https://github.com/IQSS/dataverse/pull/3146
21:43 bjonnh
do you know if NIH has a way to produce/handle DOIs?
21:44 bjonnh
I cannot find them on ezid or datacite lists
21:44 pdurbin
no idea but if you find out, please let me know
21:44 bjonnh
sure
21:44 bjonnh
I'll ask that right now to my contacts
21:44 pdurbin
cool
21:45 bjonnh
well as you guessed NIH would be part of this project
21:45 pdurbin
nice
21:45 bjonnh
at least part of it
21:46 pdurbin
make 'em pay for the DOIs :)
21:46 bjonnh
of course
21:46 bjonnh
I mean
21:46 bjonnh
that would be the plan to get them host the instance and provide DOIs
21:46 bjonnh
in term of sustainability it make sense
21:47 pdurbin
good to think about sustainability
21:47 pdurbin
the software is free but hosting is not
21:48 bjonnh
is there any map of the current harvard installation and estimation of cost?
21:51 pdurbin
disk is not cheap, it turns out
21:52 pdurbin
that's the impression I get anyway
21:52 pdurbin
I don't have access to any numbers on cost.
21:54 bjonnh
do you think that's numbers that could be accessible when I'll try to push that further?
21:54 pdurbin
bjonnh: you could ask at support
21:55 bjonnh
I'd like to have initial NIH backing first before doing any inquiry of that kind
21:56 pdurbin
meh, I wouldn't sweat it
21:57 pdurbin
they/we can answer other questions
21:57 pdurbin
or you could ask at https://groups.google.com/forum/#!forum/dataverse-community
21:57 bjonnh
first I'm inquiring about the DOI situation
21:57 bjonnh
NIST, DOE, NASA, NOAA, FWS, USGS all have the capability
21:57 bjonnh
so…
