Time
S
Nick
Message
00:40 garnett joined #dataverse
00:52 axfelix joined #dataverse
08:01 bencomp joined #dataverse
12:52 rliebz joined #dataverse
12:56 pdurbin
rliebz: mornin'!
12:56 rliebz
pdurbin: Good morning
12:59 pdurbin
lots of activity yesterday! I haven't had a chance to respond to much of it
13:12 rliebz
pdurbin: Yeah, I was just taking a look at that
13:13 pdurbin
it's great!
13:13 pdurbin
so happy to see this taking off
13:14 rliebz
pdurbin: Just as a heads up, I'm only going to be with the Center for the next few weeks, so I won't be so active on the python client after that
13:15 pdurbin
rliebz: oh! well, I believe you are/were an intern... so not shocking... where to?
13:16 rliebz
pdurbin: Capital One up in northern Virginia
13:17 pdurbin
looks like HQ is pretty close to DC
13:18 rliebz
Yep.
13:19 pdurbin
rliebz: great news. congrats!
13:20 rliebz
pdurbin: Thanks, it should be interesting
13:24 pdurbin
rliebz: don't forget about us! ;)
15:22 pdurbin
bencomp: Merce and I just met to talk about user accounts and auth
15:22 pdurbin
I have some notes for you to review.
15:23 pdurbin
bencomp: and not just because you're listed here: http://community.dataverse.org/community-groups/auth.html :)
15:57 metamattj joined #dataverse
15:58 bencomp
pdurbin: I'm listening :)
16:05 pdurbin
bencomp: https://docs.google.com/document/d/157sw9gaFGwrb0EtCGGlLMxSg0cooHfKEvIsjcebPcA8/edit?usp=sharing
16:05 pdurbin
(2015-04-30 Dataverse User Accounts and Auth Meeting)
16:05 pdurbin
sort of a brain dump
16:05 pdurbin
not the best notes. sorry
16:10 bencomp
I like the idea of orcid as auth source, but that is not what you discussed, is it? having your orcid in your profile is what came up in my head the other day too
16:11 bencomp
I see I don't know all the issues :) https://github.com/IQSS/dataverse/issues/923 Allow users to specify their ORCID and/or ISNI ID in their user account #923
16:13 pdurbin
bencomp: ORCID definitely came up but there's lots to decide there
16:14 pdurbin
different (upcoming) meeting :)
16:16 bencomp
heh
17:07 rliebz joined #dataverse
17:43 astrofrog joined #dataverse
17:43 astrofrog
hello
17:43 astrofrog
I'm trying to access the main dataverse site using the API 4.0 but things are hanging
17:43 astrofrog
I tried accessing https://dataverse.harvard.edu/api/dataverses/:root/contents?key= <api_key_here>
17:43 pdurbin
astrofrog: I don't like the sound of that!
17:44 astrofrog
but it just hangs, waiting for a response from the server
17:45 pdurbin
astrofrog: one of our two web servers *is* having trouble. we're taking it out of the load balancer
17:45 astrofrog
ok thanks
17:46 astrofrog
let me know when to try again
17:46 pdurbin
astrofrog: I take it you're ready for production! done with apitest? :)
17:46 astrofrog
well ready for a real-life test
17:46 astrofrog
managed to get it to work all nicely with apitest
17:46 astrofrog
now need to be able to access more data
17:47 pdurbin
gotcha
17:47 pdurbin
astrofrog: it's all looking green right now: https://opsview.hmdc.harvard.edu/viewport/www
17:48 astrofrog
pdurbin - thanks, but seems to still be hanging
17:48 pdurbin
boooo
17:48 pdurbin
astrofrog: what if you try a non-root dataverse?
17:48 astrofrog
do you see that URL hanging too, with your key?
17:49 astrofrog
seems to work then
17:49 astrofrog
but in order to find the non-root dataverses I think I need to access the root one right?
17:49 pdurbin
hmm. I wonder if querying the root is an expensive operation
17:49 pdurbin
astrofrog: can you please open an issue about this? https://github.com/IQSS/dataverse/issues
17:49 astrofrog
what would be an efficient way (using the native API ) to get a list of all dataverses?
17:49 metamattj joined #dataverse
17:50 astrofrog
https://github.com/IQSS/dataverse/issues/2122
17:51 pdurbin
astrofrog: thanks!
17:51 pdurbin
as for a list of all, dataverses, maybe try something like this: https://apitest.dataverse.org/api/search?q=*&type=dataverse
17:51 pdurbin
and then iterate through the list: http://guides.dataverse.org/en/latest/api/search.html#iteration
17:52 astrofrog
ok thanks
17:55 skay joined #dataverse
17:55 skay joined #dataverse
17:55 astrofrog
pdurbin - one more quick question: on this page the fits datafile has type 'plain text': https://dataverse.harvard.edu/dataset.xhtml?persistentId=hdl:10904/10069
17:56 astrofrog
but on the demo site FITS files are recognized as such:
17:56 astrofrog
https://dataverse-demo.iq.harvard.edu/dataset.xhtml?persistentId=doi:10.5072/FK2/UNYMBJ&version=DRAFT
17:56 astrofrog
is that normal?
17:59 astrofrog
(just because I'm using the contentType from my application to determine which files I can use)
18:00 pdurbin
astrofrog: that second dataset is in draft so I can't see it
18:01 astrofrog
ah
18:01 astrofrog
ok let me fix that
18:02 astrofrog
ok check again
18:02 * pdurbin
looks at https://dataverse-demo.iq.harvard.edu/dataset.xhtml?persistentId=doi:10.5072/FK2/UNYMBJ
18:04 pdurbin
astrofrog: I see what you mean. What's interesting is that there are *some* fits files in production: https://dataverse.harvard.edu/?q=*&fq0=fileTypeGroupFacet%3A%22fits%22
18:05 astrofrog
strange
18:05 astrofrog
is it to do with when they were uploaded?
18:05 pdurbin
something else...
18:06 pdurbin
PerA_1120uBolo_F.fits says "Plain Text" and it also doesn't have an MD5: https://dataverse.harvard.edu/dataset.xhtml?persistentId=hdl:10904/10069
18:07 pdurbin
astrofrog: this doesn't feel like a code issue to me... can you please email support
18:07 astrofrog
yep, will do
18:08 pdurbin
astrofrog: awesome. thanks!
18:08 astrofrog
so going back to the list of dataverses - the search URL you provided returns a flat list of *all* dataverses - but is there a way to find just the 'top-level' ones, such as e.g. 'cfa'? I would like to provide a hiearchical visualization to the user so they don't just see everything mixed up? (I guess in principle if the root dataverse was working through the API that's what I'd get?)
18:09 * pdurbin
thinks
18:09 pdurbin
astrofrog: what you want is certainly very reasonable!
18:10 pdurbin
internally, I use a "subtree" value for search/solr but I don't think I expose it...
18:11 pdurbin
oh. right. it's searchable. not on this list: https://github.com/IQSS/dataverse/issues/2038
18:11 pdurbin
hmm
18:13 pdurbin
astrofrog: it sounds like the "contents" api is what you want... but it's not working, as you said
18:13 pdurbin
I wonder if this is even possible in the GUI .
18:14 pdurbin
astrofrog: this comment is related: https://github.com/IQSS/dataverse/issues/182#issuecomment-64119983
18:17 astrofrog
ah yes
18:17 astrofrog
cool thanks
18:17 pdurbin
astrofrog: do you think this belongs in the Search API or the Dataverses API?
18:18 astrofrog
dataverses API
18:18 pdurbin
ok
18:18 pdurbin
astrofrog: please open an issue :)
18:18 astrofrog
in a way I was able to construct this on the test server
18:18 astrofrog
because the root stuff worked there
18:18 pdurbin
ah, ok. so maybe the issue you created is enough
18:18 astrofrog
I could see the top-level dataverses and the dataverses contained within each
18:18 pdurbin
astrofrog: and you could tell which was a parent of which?
18:18 astrofrog
right, let's see what happens with that first
18:19 astrofrog
well for a given dataverse I could see dataverses inside it
18:19 astrofrog
I think
18:19 astrofrog
I'll have to double check
18:19 astrofrog
but I think the thing is that in the test server the hierarchy was simple anyway
18:19 astrofrog
hence why I need to test with the 'real' thing :)
18:19 pdurbin
yeah
18:20 astrofrog
gtg, but back later
18:30 pdurbin
cool, he opened a ticket about fits: https://help.hmdc.harvard.edu/Ticket/Display.html?id=215316
19:40 bencomp joined #dataverse
19:47 garnett joined #dataverse
19:54 rliebz
pdurbin: Is there any way to retrieve any sort of metadata about a user given that user's API token? Like the username?
19:57 pdurbin
rliebz: not that I know of
19:59 pdurbin
rliebz: can you tell me about your use case?
20:19 rliebz
pdurbin: Really it would be nice to be able to have some identifying features about a user that connects to the OSF. At the bare minimum, there's a lot we could do with a username
20:20 astrofrog joined #dataverse
20:20 rliebz
pdurbin: For example, on an OSF project, it could say "Authorized by Dataverse User {username}"
20:22 pdurbin
rliebz: I agree in general but I'm not sure where you'd show "authorized by"
20:26 rliebz
pdurbin: We also have an internal use case where we would like to have a unique ID per users that isn't the API key
20:26 rliebz
pdurbin: But for other addons, we show "authorized by" on the project settings page so that shared projects or users with multiple accounts can figure out who authorized what
20:28 rliebz
pdurbin: We also plan to allow multiple Dataverse accounts per OSF account (mostly to facilitate multiple Dataverse installations), and while I don't know why a user would have multiple Dataverse accounts on a single installation, this would let users distinguish between them
20:31 axfelix joined #dataverse
20:31 pdurbin
rliebz: this is related: https://github.com/IQSS/dataverse/issues/1445
20:31 pdurbin
rliebz: and sorry, but I have to run to pick up my kids from school
20:39 rliebz
pdurbin: No worries
20:46 astrofrog joined #dataverse
20:54 astrofrog joined #dataverse
20:58 astrofrog joined #dataverse
21:05 garnett joined #dataverse
21:14 garnett joined #dataverse
21:20 garnett joined #dataverse
21:26 astrofrog joined #dataverse
21:51 garnett joined #dataverse
21:59 astrofrog joined #dataverse
