Time
S
Nick
Message
07:54 juancorr joined #dataverse
08:17 poikilotherm joined #dataverse
08:19 poikilotherm
pdurbin: speaking of dataverse-sample-data: you do have that webhook on your list, don't you? ;-)
08:21 jri joined #dataverse
09:05 MrK joined #dataverse
11:31 pdurbin
poikilotherm: "just the push event"? "Send me everything"? Or individual events?
11:32 poikilotherm
Err...
11:32 poikilotherm
Let me check
11:32 poikilotherm
Or you give dataversebot more right. Could setup the hooks himself
11:32 poikilotherm
s/right/permissions/
11:32 pdurbin
How much permission are we talking about?
11:34 poikilotherm
Most certainly that would be admin, wouldn't it? Is there some option to grant rights to admin webhooks only?
11:35 pdurbin
I don't know. Can you just tell me what you want? :) Or create an issue in the sample data repo with what you want? :)
11:37 poikilotherm
Sure
11:41 poikilotherm
Did you see my reply to Felipe about LDAP, OIDC and ADFS?
11:42 pdurbin
I hadn't. Nice reply.
11:42 pdurbin
I still can't figure out what he wants. LDAP support? Microsoft support? He seems to be focused on LDAP so far.
11:43 poikilotherm
Sure.
11:43 poikilotherm
Lots of people still on those kind of services ;-)
11:43 pdurbin
yeah, I helped support openldap back in the day
11:44 poikilotherm
You can do pretty cool stuff with it, but the web world evolved for good
11:47 pdurbin
I had an auth issue at https://github.com/IQSS/dataverse-android/issues/1 for a while but ended up waiting for the Search API to not require an API token.
13:50 stefankasberger joined #dataverse
18:51 craig-willis joined #dataverse
19:14 craig-willis
Hello -- is it possible to get individual dataset metrics from harvard.dataverse.edu via the APIs? I'm trying the makeDataCount endpoints and seeing "No metrics available for dataset X for null for country code null".
19:16 craig-willis joined #dataverse
19:35 madunlap joined #dataverse
19:41 pdurbin
craig-willis: hi! You're talking about Make Data Count metrics. We haven't turned on MDC yet for Harvard Dataverse. We're tracking this at https://github.com/IQSS/dataverse.harvard.edu/issues/3 . However, the older metrics are available using the Metrics API : http://guides.dataverse.org/en/4.18.1/api/metrics.html
19:42 pdurbin
For example, you can get the total number of datasets from https://dataverse.harvard.edu/api/info/metrics/datasets
19:44 craig-willis
pdurbin: Thanks, that makes sense. I'm trying to get per-dataset download metrics for datasets in a particular journal. Perhaps scraping it from the dataset pages?
19:47 pdurbin
Woof, that might need to be a database query. First of all, are you talking about old-style, pre-MDC downloads or are you talking about MDCs downloads? MDC is much more strict about double clicks, etc.
19:53 craig-willis
For this case, I don't know that I care. I just wanted to pull whatever metrics I can for the AJPS journal datasets.
19:59 pdurbin
craig-willis: ok, would you mind creating an issue at https://github.com/IQSS/dataverse/issues about this? Again, I think it would have to be a database query. I don't think we have APIs in place to get the old-style, pre-MDC metrics out. And those are the only download metrics (rather than MDC) that are currently enabled on Harvard Dataverse.
20:01 craig-willis
pdurbin: thank you.
20:26 iqlogbot joined #dataverse
20:26 Topic for #dataverse is now Dataverse is open source research data repository software: http://dataverse.org | IRC Logs: http://irclog.iq.harvard.edu/dataverse/today | Who's who: https://docs.google.com/spreadsheets/d/16h3jv24usMGq18495C-JA-yNcQCKiKDa65MTraNDd7k/edit?usp=sharing
20:33 pdurbin
sure!
21:13 pameyer joined #dataverse
21:15 pameyer
pdurbin: if you haven't started biking the rain, got a moment to sanity check me on something?
21:15 pdurbin
pameyer: welcome back! We missed you. What have you got?
21:16 pameyer
somebody was asking me about using custom metadata blocks in a dataverse installation they didn't host. I'm prety sure that's admin-only, and they'd need to talk to the installation admins
21:16 pameyer
dev guide makes it look like it's still that way; but I've been out of the loop
21:17 pameyer
how have things been otherwise?
21:17 pdurbin
You're right. Out of the box you can only load custom metadata blocks from localhost. So you have to ssh in.
21:17 pdurbin
It's been that way since 4.0. No changes.
21:18 pdurbin
pameyer: hey are you any closer to putting your installation on the map? :)
21:18 pameyer
em-dataverse is soft-launching, but we've still got some stuff for sbdb
21:19 pdurbin
Is em the bch one?
21:19 pameyer
no - the local one for the em facility
21:19 pdurbin
orly
21:19 pameyer
it's the ldap / pipeline "deposits" all datasets one
21:20 pdurbin
Ok, I wonder if there's anything in my email about it.
21:20 pameyer
not sure - other than auth, it's pretty much stock out of the box
21:21 pdurbin
Ok, so they may not have emailed support.
21:23 pdurbin
I probably told you about the new map already.
21:23 pameyer
I think so
21:24 pdurbin
Ok. Other new stuff: OIDC, Binder support.
21:24 pameyer
Binder's been in the works for a while
21:24 pameyer
cool to see the OIDC stuff; even if I haven't dug into enough to know much about it
21:25 pdurbin
Yes! Finally the UI bit was merged today! For Binder. All thanks to xarthisius. Please see https://twitter.com/philipdurbin/status/1204472858413371392
21:26 pdurbin
We've been thinking of OIDC as a way to support LDAP. Is em-dataverse using ADFS for LDAP?
21:26 pameyer
and it looks like either I didn't do a paranoid monologue about "port-scan your install before opening it up", or it didn't make it to the dev guide :(
21:26 pameyer
nah - that one's using shib
21:27 pdurbin
LDAP support through shib? I'm confused. How is em-dataverse getting LDAP support?
21:27 pameyer
ldap -> ipa -> shib -> dataverse
21:27 pdurbin
Interesting. Should I ask them to contribute their shibboleth2.xml file?
21:28 pameyer
I ended up using ipsilon because the semi-official shibboleth ipa wasn't something I could figure out
21:28 pdurbin
Oh ho. You set it up. :)
21:28 pameyer
yeah - "them" is me for that one
21:28 pdurbin
nice
21:28 pameyer
it's the one that the draft version logging was for
21:29 pdurbin
ah
21:29 pdurbin
Well, the LDAP thread woke up this week if you're interested: https://groups.google.com/d/msg/dataverse-community/rCsRC4Vvqjw/tmDu1WZyAwAJ
21:29 pdurbin
It sounds like you might have an option 6. :)
21:30 pameyer
huh - ldap seems to keep popping up
21:30 pdurbin
yeah
21:30 pameyer
it might be worth a PR to document - but it also might be too much of a snake's nest to be easilly supportable
21:32 pdurbin
Yeah. And OIDC might be a little tricky to support too. Who knows.
21:32 pameyer
good point
21:32 pdurbin
I figure the community can support itself to some degree. :)
21:33 pameyer
;)
21:33 pameyer
and thanks for pointing Thanh-Thanh to https://github.com/sbgrid/data-capture-module/issues/36
21:34 pdurbin
sure, any thoughts on https://github.com/IQSS/dataverse-kubernetes/issues/68 ?
21:35 pameyer
my first thought is that I don't know if dataverse-kubernetes supports a shared filesystem
21:35 pameyer
but if it does, it _should_ be straightforward to convert the docker-dcm compose files in
21:36 pdurbin
I gather that with Kubernetes all things are possible, including shared file systems. But who knows.
21:36 pameyer
it might also result in a black hole of my docker worst practices :)
21:36 pdurbin
heh
21:36 pdurbin
It's not like your Dockerfiles get spread all over the world.
21:37 pdurbin
I don't know if Oliver knows about the shared file system thing.
21:37 pameyer
as long as my email isn't getting all the notifications :)
21:37 pdurbin
:)
21:37 pameyer
you can probably share storage between pods
21:38 pdurbin
Or have a pod that's an NFS server? I don't know what's easier.
21:38 pdurbin
Probably your idea, if you know what you're doing in Kubernetes.
21:38 pameyer
.... reminds me that I thought I saw some chatter (Oliver and Don, maybe?) about using jacoco agent for instrumenting the war file for coverage measurements
21:38 pdurbin
yeah
21:39 pameyer
I hit a weird error with a method too long exception when trying that
21:39 pameyer
it's why that PR was a script to unpack the war, instrument, and repack
21:39 pdurbin
hey, your script works, I tested it and add more docs
21:39 pameyer
great
21:40 pdurbin
Not sure if more docs are better docs.
21:40 pameyer
incremental improvement, right?
21:40 pdurbin
or at least longer :)
21:41 pameyer
and then the circle of life flows into "the docs are too big and complex, we need to simplfy them" ;)
21:41 pdurbin
heh
21:42 pdurbin
The universe is expanding and contracting and expanding again.
21:42 pameyer
exactly
21:42 pdurbin
Do you think this rain will let up?
21:44 pameyer
probably not before I get home ....
23:37 pameyer
anyhow - stay dry ;)
