Time
S
Nick
Message
01:01 axfelix joined #dataverse
02:51 clabruno joined #dataverse
03:33 shauna joined #dataverse
03:52 axfelix joined #dataverse
05:05 axfelix joined #dataverse
05:46 axfelix joined #dataverse
05:48 garnett joined #dataverse
06:42 agarnett joined #dataverse
09:05 bencomp joined #dataverse
10:38 bencomp1 joined #dataverse
16:01 axfelix joined #dataverse
16:06 sanchitgoyal01 joined #dataverse
16:39 sanchitgoyal01 joined #dataverse
16:59 razorsharph joined #dataverse
17:01 bencomp
pdurbin: did you make it to the office? :)
17:02 bencomp
https://twitter.com/philipdurbin/status/566214348243496960
17:02 pdurbin
are you referring to https://twitter.com/philipdurbin/status/566214348243496960 ? :)
17:02 pdurbin
jinx!
17:02 pdurbin
yeah, trying to finish this: Account/Shibboleth: Converting from local to non-local · Issue #796 · IQSS/dataverse - https://github.com/IQSS/dataverse/issues/796
17:02 bencomp
what do I say in response to jinx? :)
17:03 pdurbin
bencomp: you can't say anything! you're jinxed!
17:03 bencomp
d'oh
17:03 pdurbin
even my kindergartener knows that ;)
17:04 bencomp
we don't have that here :)
17:04 pdurbin
I figured.
17:04 bencomp
I'm finally writing that email to dvn-auth about the auth group
17:05 bencomp
do we/you expect that people (semi-)formally join the group?
17:06 bencomp
or do group coordinators report what any somewhat involved person has to say?
17:07 pdurbin
well
17:08 pdurbin
right now this is who I consider to be part of the group: dvn-auth Subscribers - https://lists.iq.harvard.edu/mailman/roster/dvn-auth
17:09 bencomp
okay, that makes sense :)
17:10 pdurbin
but we can change what it means to be part of the group
17:10 pdurbin
I'd like to get away from the term "dvn" anyway
17:10 bencomp
maybe except for the "groups will be formed at the [meeting] in June" (http://community.dataverse.org/community-groups/index.html)
17:11 pdurbin
right
17:12 pdurbin
we should invite people to join the new group. it's very clear who is part of the old group. will there be similar clarity on who is part of the new group?
17:12 bencomp
I put my role in the group in our CRIS today :) https://pure.knaw.nl/portal/en/persons/ben-companjen%280db8708d-9b7d-44f5-b960-521261d8b2f9%29.html
17:14 pdurbin
bencomp: pull request time. add a link to your name
17:14 bencomp
:D
17:16 bencomp
I feel a group membership comes with rights and responsibilities. e.g. the right to 'influence' development by discussing functional and technical issues
17:17 bencomp
and e.g. the responsibility to get involved in discussions?
17:18 pdurbin
ok
17:18 pdurbin
"you get a vote"
17:18 pdurbin
or something
17:18 pdurbin
I would love more discussion around auth.
17:18 bencomp
perhaps, although a CG reports to the Advisory Team, who advises development
17:19 bencomp
so what would getting a vote mean? :)
17:20 bencomp
just thinking out loud, byw
17:20 bencomp
s/byw/btw/
17:22 pdurbin
:)
17:22 pdurbin
oh I dunno
17:22 pdurbin
it's hardly a democracy
17:22 bencomp
pdurbin: I could write that we don't know the exact group definition yet
17:22 pdurbin
which is fine
17:22 pdurbin
the people who show up can define it :)
17:23 bencomp
agreed :)
17:24 bencomp
I'd better arrange my trip soon :)
17:25 pdurbin
heh
17:25 pdurbin
oh I mean show up virtually too
17:34 pdurbin
bencomp: does your IdP return an attribute for "faculty" or "staff" or "student" or whatever?
17:35 bencomp
urn:mace:dir:attribute-def:eduPersonAffiliation is used for that
17:38 bencomp
pdurbin: but not all source IdPs provide it, afaik
17:40 pdurbin
bencomp: Harvard uses eduPersonScopedAffiliation http://iam.harvard.edu/resources/saml-shibboleth-attributes
17:40 pdurbin
"One or more of the following values: faculty, staff, student, affiliate, and member."
17:42 bencomp
hmm
17:43 sanchitgoyal01 joined #dataverse
17:44 bencomp
pdurbin: I wondered if it's possible to configure a filter on the Shib side to pick one of X attributes that more or less convey the same information
17:45 bencomp
also: http://xkcd.com/927/
17:45 pdurbin
bencomp: right. this is the same question as eppn vs NameId. Can each installation of Dataverse control what attributes are sent to the Dataverse applications via xml config files in /etc/shibboleth?
17:45 bencomp
indeed
17:45 pdurbin
bencomp: is there someone you can ask?
17:46 bencomp
although I thought issue 1422 was more on the Dataverse application side :)
17:47 bencomp
I'm the last one in the office now, it's dinner time
17:47 bencomp
+ weekend
17:47 bencomp
:P
17:48 bencomp
I'll talk to Arnoud next week, but I'm not sure he knows everything about Shib/SAML
17:48 pdurbin
well, I don't have to do anything for https://github.com/IQSS/dataverse/issues/1422 if you can configure shibd to always send eppn to the Dataverse app
18:16 bencomp
pdurbin: as I'm using your name in this email, would you like to read it before I send it to the list? :)
18:16 bencomp
I'm quite happy about the tone and content
18:17 bencomp
I'm going home now, it's in your inbox and I'll send it later if you're happy with the contents
18:20 bencomp1 joined #dataverse
18:20 pdurbin
bencomp: in general, looks great. I feel like we should give people something to opt-in to though.
18:20 pdurbin
maybe we could give them all edit access to a Google Doc or something
18:22 bencomp1
right. let me think about that later :)
18:22 bencomp1
good night and weekend for now!
18:48 sanchitgoyal01 joined #dataverse
18:50 agarnett joined #dataverse
20:49 bencomp joined #dataverse
23:03 clabruno joined #dataverse
