Time
S
Nick
Message
07:41 poikilotherm joined #dataverse
08:15 jri joined #dataverse
09:11 MrK joined #dataverse
11:50 pdurbin
poikilotherm: mornin. Your oauth pull request was merged.
11:54 poikilotherm
Good morning pdurbin :-)
11:54 poikilotherm
Yeah, I saw that :-)
11:54 poikilotherm
Thanks for adding the diff link
11:56 pdurbin
Oh. Sure. Now we can diff from develop, of course. :)
11:57 poikilotherm
;-)
11:58 poikilotherm
I'm working on a poster today, so I wont do new commits
12:06 pdurbin
cool, have fun
13:08 jri joined #dataverse
13:16 jri joined #dataverse
13:28 donsizemore joined #dataverse
13:36 pdurbin
donsizemore: morning. Shib question for you, when you're ready.
13:39 pdurbin
It's about this "configuring shibboleth on AWS using ELB" thread: https://shibboleth.1660669.n2.nabble.com/configuring-shibboleth-on-AWS-using-ELB-td7644856.html
13:43 donsizemore
hit me
13:46 * poikilotherm
pokes donsizemore
13:47 poikilotherm
There you go! ;-)
14:07 pdurbin
donsizemore: the question is if our docs here are out of date: http://guides.dataverse.org/en/4.18.1/installation/shibboleth.html#backup-sp-cert-pem-and-sp-key-pem-files
14:07 pdurbin
From the nabble thread, I'm reacting to "The supported version of the SP generates two keypairs, one for signing, and one for encryption, neither one by that name anymore."
14:08 pdurbin
when I wrote those docs, I was testing with shib 2, but now I think the world has moved on to shib 3, which I've never used
14:33 donsizemore
i'm running shib3, which was an in-place upgrade from the initial 2.6 installation
14:35 donsizemore
i'll take a look and compare the docs
15:01 pdurbin
donsizemore: thanks!
15:01 pdurbin
ever since my shibtest box got owned I haven't kept up with shib stuff
15:15 stefankasberger joined #dataverse
15:21 stefankasberger
@pdurbin: Hi! :) Am coding right now a lot around pyDataverse, and already implemented a proof of concept for the OAISTree (formerly known as DVTree).
15:22 stefankasberger
Regarding it, I have a question: is it possible to set the "Enable access request" checkbox from the user backend via the API ? Cause we want to restrict the datafiles for users via the form.
15:22 pdurbin
orly
15:22 pdurbin
That's a great question. Lemme poke around in the API Guide and code.
15:26 pdurbin
stefankasberger: I think you're in luck: http://guides.dataverse.org/en/4.18.1/api/dataaccess.html#allow-access-requests
15:30 Youssef_Ouahalou joined #dataverse
15:33 stefankasberger
Great. Is this a new feature in 4.18.1, or does it also work in older ones?
15:35 pdurbin
It's not new. It looks like it was added in Dataverse 4.10 around this time last year in https://github.com/IQSS/dataverse/pull/5309
15:37 andrewSC joined #dataverse
16:01 pdurbin
donsizemore: I just gave Erik of OCR fame your email address
16:03 stefankasberger
Thanks!
16:05 pdurbin
stefankasberger: sure. Have you seen the OAISTree/DVtree structure we're using in https://github.com/IQSS/dataverse-sample-data ?
16:34 stefankasberger
Will have a look. Have now my basketball session. :) Unfortunately, the community call is not possible for me right now cause of that.
16:36 pdurbin
stefankasberger: oh. Well we moved it to 2 hours earlier, if that helps.
17:24 pdurbin
donsizemore: some Arabic text with Dataverse in it: http://scinews.ir/en/articles/89230.html :) Google translate of the title: Importance of Research Data Archives: An Essential Step in Improving the Quality of Articles and Data Sharing
22:00 donsizemore
@pdurbin I am happy to report that Leonid arrived at the same conclusion I did based on my limited testing.
22:01 pdurbin
Is this about Solr?
22:02 donsizemore
whoops, yes
22:03 pdurbin
we like your localhost fixes
22:03 pdurbin
will probably put a note in the next release notes
22:03 pdurbin
that you're welcome to review and edit, of course
22:21 donsizemore
@pdurbin oh, I wrote a little script today to generate TSV given a number of variables and observations
22:22 donsizemore
@pdurbin i'm finding that a smaller number of variables performs well no matter how many observations
22:22 donsizemore
@pdurbin but a larger number of variables really slows the ingest process. it's reading cell by cell regardless, no?
22:23 donsizemore
@pdurbin akio suggested memory management may be the factor
23:16 pdurbin
Hmm, well variables are put in the database and indexed. Observations are not. Bu yeah, interesting. Good to test this stuff.
