Time
S
Nick
Message
07:42 jri joined #dataverse
08:45 jri_ joined #dataverse
09:23 pkiraly joined #dataverse
10:32 pkiraly
@poikilotherm, Hi
10:33 pkiraly
poikilotherm, I am thinking of improving the metadata block import. My idea is to base it on the Solr Schema API .
10:34 pkiraly
instead of files, which is your solution
10:35 pkiraly
It requires a change in Solr configuration, which enables the Schema API
10:36 pkiraly
and it also includes changes in Dataverse code base, to call Solr API directly
10:36 pkiraly
This way we can get rid of file management
10:38 pkiraly
But this kind of change would also makes your recent work on this (the xinclude based solution) outdated
12:41 pkiraly joined #dataverse
12:42 Youssef_Ouahalou joined #dataverse
13:43 pdurbin
pkiraly: hi. I think he's on vacation. Or moving. Or both. :)
14:25 donsizemore joined #dataverse
14:26 pkiraly
pdurbin, yes, it seems so. I will try next week as well
14:28 pkiraly
pdurbin, Is an issue a must have for a PR? In the last days I worked on some custom metadata block, and found that if the spreadsheet has an error, the API gives cryptic answer sometimes. I worked on the Java code, but hasn't yet created an issue.
14:35 pdurbin
pkiraly: it's fine if you just create a pull request
14:36 pdurbin
(but we do appreciate issues)
14:50 pkiraly
pdurbin, one more question: the files which test the rest API calls has a special in their name like RO, but I forget what is that exactly. Do you happen to remember?
14:50 Jim54 joined #dataverse
14:51 pdurbin
IT
14:51 pkiraly
thanks!
14:51 pdurbin
DatasetsIT, etc
14:51 pdurbin
IT stands for Integration Test
14:52 Jim54
@pdurbin - the community emails about a custom metadata block - seems like he's trying the metadatablocks.json file/api instead of the api to add a new tsv - is that a mistake?
14:53 pdurbin
Yeah, that's why I was trying to get him to verify that his block has been loaded.
16:11 donsizemore joined #dataverse
16:11 donsizemore
@pdurbin knock knock?
16:12 donsizemore
@pdurbin low-key freak-out over here
16:12 pdurbin
!!!
16:12 pdurbin
what are we freaking out about?
16:12 donsizemore
@pdurbin two externaltools. one is data explorer (file level). one is matthew's merge app (dataset level)
16:13 donsizemore
@pdurbin prior to today, the dataset merge button went to his webapp; the explore button to data explorer
16:13 donsizemore
@pdurbin but as of today, they both go to matthew's webapp _IF_ terms of use are enabled
16:13 donsizemore
@pdurbin matthew says the path is different when terms of use are enabled
16:14 donsizemore
@pdurbin and... we need terms of use enabled. matthew suspects a bug in 4.20, but any suggestions are welcome
16:15 pdurbin
Danny opened an issue about this a while back. I can't remember which tools were being scrambled. Probaby two file level tools.
16:16 donsizemore
we've changed the order of them in the externaltools table. matthew's dataset-level tool seems to take precedence no matter what
16:17 donsizemore
in our case it seems to choose the dataset-level tool at the file level
16:18 pdurbin
Definitely sounds like a bug. Can you please open an issue? I'm not sure what a workaround would be.
16:19 donsizemore
@pdurbin now that i go to the file landing page, things work.
16:21 pdurbin
ok
16:50 pkiraly
pdurbin, https://github.com/IQSS/dataverse/pull/6921 with integration test!
16:51 pdurbin
pkiraly: thanks!
17:46 donsizemore
@pdurbin so... thu-mai says advanced search is broken on demo.dataverse.org?
17:46 donsizemore
@pdurbin (it throws NPEs on our 4.20 test instance)
17:47 donsizemore
@pdurbin she takes it back. advanced search does work on demo.
17:49 pdurbin
phew
17:49 pdurbin
donsizemore: doing an "index all" might help your test instance
17:51 pdurbin
donsizemore: also, I'd like to pick your brain about Matthew's dataset level tool.
17:52 donsizemore
i've got the correct schema.xml and solrconfig.xml, cleared solr timestamps and did a 'continue' reindex
17:52 donsizemore
i'm spinning up a 4.20 instance in vagrant to poke at the dataset-level dealy (using wholetale)
17:53 donsizemore
our current issue is we can do an advanced search by title or subject, but not by variable name or label
17:53 pdurbin
weird, should work
17:55 pdurbin
I can't seem to force an NPE on Harvard Dataverse.
18:03 donsizemore
now we're down to variable labels
18:04 pdurbin
cool
18:04 pdurbin
time for sprint planning
18:12 jri joined #dataverse
18:19 jri joined #dataverse
18:42 jri joined #dataverse
18:51 jri joined #dataverse
19:15 jri joined #dataverse
19:25 donsizemore joined #dataverse
20:35 donsizemore joined #dataverse
