Time
S
Nick
Message
01:10 bjonnh
pdurbin: was there a date decided for that whole-tale thing?
01:49 pdurbin
bjonnh: Tentative time: Wednesday October 24, 2018, 1pm PDT/4pm EST and 9am NZDT (via Google Meet)
02:49 jri joined #dataverse
03:49 jri joined #dataverse
07:01 poikilotherm joined #dataverse
07:25 jri joined #dataverse
09:35 jri_ joined #dataverse
09:39 jri joined #dataverse
10:09 jri joined #dataverse
11:19 donsizemore joined #dataverse
11:26 juancorr left #dataverse
11:52 jri joined #dataverse
12:04 jri joined #dataverse
12:39 jri joined #dataverse
12:50 cdsp-rmo joined #dataverse
12:53 jri joined #dataverse
13:06 donsizemore joined #dataverse
13:44 pdurbin
cdsp-rmo: hi! There was some conversation about your NESSTAR issue yesterday (#4593): http://irclog.iq.harvard.edu/dataverse/2018-10-04#i_74719
13:44 cdsp-rmo
thanks, gonna take a look at this
13:45 pdurbin
And I'm looking at your comment from an hour ago. :)
13:52 cdsp-rmo
long time no use migrate
13:52 cdsp-rmo
strange thing it "disappeared"
13:52 pdurbin
yeah
13:53 pdurbin
I can't tell if he (Don) was trying to import a NESSTAR dataset or not. Can you?
13:53 cdsp-rmo
if he wants to try importing a dataset
13:53 cdsp-rmo
he can use the import point
13:53 cdsp-rmo
of batch
13:54 pdurbin
he's here a lot. let's tell him when he arrives
13:54 cdsp-rmo
yp
13:54 cdsp-rmo
*yup
13:54 pdurbin
I'm going to walk down the hall and talk to our metadata guy about your comment if he's in. And get coffee.
13:55 cdsp-rmo
I understand now from when the import type "Import" disappeared
13:55 cdsp-rmo
I've talked to my "metadata" guys too :D
13:56 cdsp-rmo
depositor can be multiple, but logically, there is only one organization that deposit data so ... our datas will be changed, at least. Dunno if it will be the case for everyone, so I will have to handle the case with an undestandable error message
13:57 cdsp-rmo
and about countries, it's a problem on our side mostly, countries are standardized, so there is no reason we use "custom" countries
13:58 cdsp-rmo
(also, there is a thing I think I have to fix in the code: only the "name" of the country is used, but the country code should be used instead, more robust)
13:58 cdsp-rmo
(especially for translations)
14:09 Julian18 joined #dataverse
14:11 jri joined #dataverse
14:21 pameyer joined #dataverse
14:27 jri_ joined #dataverse
14:30 jri__ joined #dataverse
14:45 pameyer
andrewSC: for an unsolicited opinion of http://irclog.iq.harvard.edu/dataverse/2018-10-04#i_74636; one drawback of java is that it usually pulls you into the maven ecosystem
14:49 andrewSC
pameyer: mmmmmmmmmmmm
14:49 andrewSC
is that a bad thing?
14:49 andrewSC
iirc it's kinda the only tool of its kind wrt java right?
14:51 pameyer
maven seems to assume its always online; and fail in odd ways if it's not, or if there are issues with remote repositores.
14:53 pameyer
so it works - but in my opinion, it's information worth keeping in mind when thinking about languages / frameworks. the predominate build tool mixes several steps (dependency resolution, dependency installation, compiling) in a way that seems to have poor abstraction boundaries.
14:53 donsizemore joined #dataverse
14:55 pameyer
I *think* it's incremental compilation is also non-functional
15:04 andrewSC
pameyer: duly noted!
15:04 andrewSC
thanks for the input :)
15:04 cdsp-rmo
donsizemore : hey ! It seems you wanted to try ddi files with dataverse
15:05 donsizemore
@cdsp-rmo well, I thought I did. things got more interesting this morning
15:05 cdsp-rmo
ah ?
15:05 cdsp-rmo
there is an other api in batch that allows you to upload one ddi file if needed
15:06 cdsp-rmo
that should still be there
15:07 donsizemore
@cdsp-rmo it turns out what i'll be doing is pretending to harvest but with XML e-mailed as a zip file instead of via OAI-PMH
15:07 pameyer
donsizemore: solr unhappiness?
15:07 cdsp-rmo
ah
15:08 donsizemore
@pameyer um, i had to give limits.conf a few more cookies and ice cream to make solr happy
15:08 cdsp-rmo
pretending to harvest ?
15:08 donsizemore
@pameyer i'm just adding three lines and want to format it as a code block
15:08 pameyer
donsizemore: gotcha
15:08 donsizemore
@cdsp-rmo i have 2.5 codebook metadata from IPUMS, but they don't have a harvesting server
15:09 cdsp-rmo
ok
15:09 cdsp-rmo
so, you have xml ddi file, that's it ?
15:09 donsizemore
@cdsp-rmo so we want to harvest locally, point at the existing DOIs, and decide where we want Dataverse to tell everyone we "harvested from"
15:10 cdsp-rmo
ok, I think I get it
15:15 pdurbin
I don't get it but that probably doesn't matter. :)
16:04 jri joined #dataverse
16:22 cdsp-rmo
happy hour time ... happy week-end everybody ! ;)
16:50 donsizemore joined #dataverse
16:59 pdurbin
pameyer: hi. Remember the schema.org discussion?
17:54 donsizemore
@pdurbin thanks for the tidying-up! i use macdown which sometimes differs from GitHub in slight ways
17:55 pdurbin
yeah, it's rst vs md
17:55 * pdurbin
prefers md
17:56 pdurbin
Did we conclude anything about import? Are you all set?
18:01 donsizemore
@pdurbin as of now we're back to the drawing board
18:02 donsizemore
@pdurbin we were going to set up a local OAI server or possibly harvest all the IPUMS stuff from DataCite, but we won't get all the metadata
18:02 pdurbin
ok, thanks
18:09 donsizemore
@pdurbin actually, we might be hopping on the #5104 train, but Mandy and Thu-Mai want to treat the metadata as harvested rather than imported
18:09 pdurbin
ok
18:26 pameyer
pdurbin: which schema.org discussion?
18:30 pdurbin
pameyer: #4371 but check out all the excitement about being able to download files from Zenodo if all you know is the DOI: https://github.com/whole-tale/whole-tale/issues/35#issuecomment-427411937
18:32 pameyer
pdurbin: cool - even though my instictive reaction to data files from DOIs is still "what if it's not a http uri?"
18:32 pdurbin
Then you client had better know how to deal with ftp:// or rsync:// or whatever :)
18:32 pdurbin
your*
18:33 pameyer
globus:// ....
18:33 pameyer
and I prefer md over rst too - but it is what it is
18:33 pdurbin
sure, pick your poison
18:34 pameyer
what I remember from the recent 4371 discussion was dropping the file download urls from the output
18:34 pdurbin
sigh
18:35 pdurbin
if we drop the file download URLs from the output, how are the whole tales and binderhubs and code oceans of the world going to download files from Dataverse?
18:35 pameyer
if we don't, how long will glassfishes stay running?
18:36 pdurbin
Sigh
18:37 pameyer
I haven't measured, and could be wrong, but my intuition is that crawler level traffic to those APIs wouldn't be great for stability
18:37 pdurbin
I disagree with dropping file download URLs from the output.
18:37 pdurbin
I probably should have been more vocal about this.
18:38 pameyer
I agree that it would be good to have them there.
18:39 pameyer
If I'm remembering correctly, the reason for dropping them was concerns about load/stability. If those aren't problems, that might be an argument for putting them back.
18:39 pdurbin
Let's have the download URLs in by default and add an installation level boolean to take them out for installations for which Amazon bills or load are a concern.
18:59 pameyer
or see how hard you can hit them before glassfish has issues
19:00 pdurbin
Boolean first.
20:03 donsizemore
@pdurbin knock knock?
20:25 pdurbin
donsizemore: for you. 5 minutes. :)
20:30 pdurbin
I hope it can wait until Tuesday.
20:37 pdurbin
Have a good long weekend, everybody.
20:37 pdurbin left #dataverse
21:19 pameyer left #dataverse
23:05 iqlogbot joined #dataverse
23:05 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
