Time
S
Nick
Message
02:07 skay joined #dataverse
02:13 axfelix joined #dataverse
06:53 jri joined #dataverse
07:34 jri joined #dataverse
07:42 pvh_sa joined #dataverse
08:26 pvh_sa joined #dataverse
08:27 romainM
finally done : https://github.com/IQSS/dataverse/issues/3368
09:32 jri_ joined #dataverse
09:35 jri__ joined #dataverse
10:34 pdurbin
fantastic bug report
11:30 rmo joined #dataverse
12:11 donsizemore joined #dataverse
12:27 yoh joined #dataverse
12:41 donsizemore joined #dataverse
12:55 pdurbin
donsizemore: I hope my emails about https://github.com/IQSS/dataverse/pull/3000/files make sense. And good morning.
13:08 donsizemore joined #dataverse
13:37 pdurbin
romainM: great bug report! thanks!
13:41 romainM
pdurbin: np, what is good for you is good for us too ;)
13:42 pdurbin
we're in this together now! :)
13:42 djbrooke joined #dataverse
13:42 romainM
the issue was understandable ?
13:42 pdurbin
yep, seems like a legit bug
13:43 romainM
ok
13:43 pdurbin
all new issues automatically show up at the top of the Backlog column at https://waffle.io/IQSS/dataverse
13:44 romainM
saw that tool before, great one :)
13:45 pdurbin
djbrooke introduced it to us and I really like it
13:45 djbrooke joined #dataverse
13:59 pameyer joined #dataverse
14:17 djbrooke_ joined #dataverse
14:36 donsizemore joined #dataverse
14:45 djbrooke joined #dataverse
14:56 donsizemore
@pdurbin knock knock?
14:58 donsizemore
@pdurbin i'm cooking up some perl. best to go ahead and take the batteries out of the smoke detector
14:59 pdurbin
donsizemore: mmm, perl. My first real language. Standup in 1 minute though.
15:02 bricas_
pdurbin: heya! i finally did the upgrade from 4.4 to 4.5 on our production install. i assume i'll always get a pile of sql warnings when deploying a new .war? (tables already existing and such)
15:03 donsizemore
@bricas_ that's correct. as long as the error ends in "already exists" you're good to go
15:04 donsizemore
@bricas_ you caught the bit about adding the timer server line to your jvm-options?
15:04 bricas_
i did.
15:05 bricas_
everything seems to have gone well, actually.
15:06 bricas_
we have no actual data yet, so that helps. :)
15:13 djbrooke joined #dataverse
15:21 djbrooke joined #dataverse
15:27 donsizemore joined #dataverse
15:30 pdurbin
bricas_: yeah, all those warnings are scary if you actually read them. It's suboptimal that you see them.
15:31 pdurbin
donsizemore: oh, you're cooking up some Perl as a workaround until https://github.com/IQSS/dataverse/issues/3369 is addressed?
15:32 djbrooke joined #dataverse
15:34 donsizemore
@pdurbin just want to query the database for dataset IDs (1st with distribution dates, 2nd with production dates but no distribution dates) then make the appropro curl call.
15:35 donsizemore
@pdurbin and back in the 90s when I worked for UNC ATN, it was called back then, over the chalkboard in the meeting room was a sign that read "Think Outside the Box." undernearth it someone chalked, "we can't even FIND the box"
15:35 pdurbin
heh
15:35 donsizemore
productionDate is datasetfieldtype 42. distributionDate is datasetfieldtype 56. got that.
15:36 donsizemore
dvobject stores every *other* type of date one could need for each dataset.
15:36 donsizemore
i've dumped both schema and test uploads and can't find anything named productiondate or distributiondate, case insensitive. akio's tracing the code but i was hoping you'd know off the top of your head
15:38 donsizemore
s/test\ uploads/test\ database/g
15:40 pdurbin
donsizemore: try http://localhost:8080/api/admin//datasetfield/distributionDate and http://localhost:8080/api/admin//datasetfield/distributionDate
15:40 pdurbin
whoops, and http://localhost:8080/api/admin//datasetfield/productionDate
15:41 donsizemore
so i'm inserting a dataset ID in there? returns a 404 from 4.5
15:42 donsizemore
i can do a -X GET with no $id and it returns the entry from datasetfieldtype
15:42 pdurbin
donsizemore: sorry, I'm probably only adding to your confusion
15:42 pdurbin
I'm trying to help you understand where productionDate and distributionDate come from.
15:43 donsizemore
@pdurbin ah, ok. i found them in that table, i can see the API call modifying the citationdatedatasetfieldtype_id when I do a PUT or a DELETE
15:43 pdurbin
cool
15:43 pdurbin
donsizemore: so what do you need now?
15:44 donsizemore
@pdurbin I just can't find where they're stored in the database so I can test for the existence of either before choosing one or the other
15:44 donsizemore
Thu-Mai would prefer distributionDate if we have it, then productionDate, then publicationDate
15:44 donsizemore
but i can only find publicationdate in the database
15:45 donsizemore
i'm taking up too much of your time, i can open a help ticket
15:46 pdurbin
donsizemore: I hear you that publicationdate is part of dvobject: http://phoenix.dataverse.org/schemaspy/latest/tables/dvobject.html
15:47 pdurbin
but you won't find productionDate or distributionDate in that schemaspy output because they aren't hard-coded. they come from the citation metadata block
15:47 pdurbin
donsizemore: those curl commands really return a 404?
15:48 pdurbin
donsizemore: here's the output of those curl commands from my laptop: https://paste.fedoraproject.org/432127/72907147/
15:49 pdurbin
does that help?
15:51 donsizemore
@pdurbin ah, ok. i was passing the API endpoints a dataset ID to see what would happen =)
15:52 pdurbin
donsizemore: no more 404?
15:53 donsizemore
@pdurbin correct. so we'll have to pull JSON for each dataset?
15:54 pdurbin
well, you might be a step or two ahead of me
15:54 pdurbin
it sounds like you want a list of dataset ids
15:55 donsizemore
@pdurbin i can pull a list of dataset IDs, then I want to check for say productionDate IS NOT NULL
15:56 djbrooke joined #dataverse
15:57 djbrooke joined #dataverse
15:57 pdurbin
oh, I see
15:57 pdurbin
yes, that will be in the JSON output of each dataset
15:57 pdurbin
well, it will be absent for some (or most), it sounds like
15:59 donsizemore
@pdurbin correct, we want to test for distributionDate and use that if possible
15:59 donsizemore
@pdurbin then test for productionDate without distributionDate, runner-up
15:59 djbrooke_ joined #dataverse
15:59 donsizemore
@pdurbin then use publication date for Miss Congeniality
16:00 pdurbin
:)
16:01 pdurbin
I'm supposed to meet someone for lunch. I'll be back.
16:10 djbrooke joined #dataverse
17:17 djbrooke_ joined #dataverse
17:17 donsizemore joined #dataverse
17:19 donsizemore
@pdurbin meant to tell you to enjoy lunch! i'm back from the gym and am installing perl-JSON
17:19 pdurbin
donsizemore: Thanks. I guess I'm not sure if you're working on a workaround (calling the API ) or trying to help Akio make a pull request to add a setting.
17:21 donsizemore
@pdurbin akio's working on the pull request to add the setting. i'm working on a perl script to pull the unique dataset ids, pull their citation metadata json, then test for the presence of production/distribution date and issue the appropriate curl call. at least, that's where i was when i left for the gym
17:21 djbrooke joined #dataverse
17:24 pdurbin
going to the gym is good for clearing your mind
17:26 djbrooke joined #dataverse
17:40 djbrooke joined #dataverse
17:44 djbrooke joined #dataverse
17:45 donsizemore
@pdurbin woo woo, this got easier. Thu-Mai is only interested in the distribution date of the earliest version for each dataset, otherwise we go with pubdate
17:46 pdurbin
cool
17:50 djbrooke joined #dataverse
18:10 donsizemore
@pdurbin am I correct in seeing that the output of http://localhost:8080/api/datasets/$dataset/versions/1/metadata/citation won't contain "distributionDate" if one isn't set?
18:16 jri joined #dataverse
18:17 jri joined #dataverse
18:58 djbrooke joined #dataverse
19:02 donsizemore
@pdurbin so I think I'm correct, after I thought I was wrong. I have some perl to retroactively set /api/datasets/$dataset/citationdate to distributionDate for datasets boasting said. I'll contribute if you all promise not to make too much fun of it
19:08 djbrooke joined #dataverse
19:09 djbrooke joined #dataverse
19:09 djbrooke_ joined #dataverse
19:31 djbrooke joined #dataverse
20:18 axfelix joined #dataverse
20:40 pdurbin
heck our installer is written in Perl :)
21:27 djbrooke_ joined #dataverse
21:29 djbrooke_ joined #dataverse
21:31 djbrooke joined #dataverse
22:03 axfelix joined #dataverse
