Time
S
Nick
Message
01:40 jri joined #dataverse
04:40 jri joined #dataverse
06:59 jri joined #dataverse
07:22 juancorr joined #dataverse
07:34 juancorr joined #dataverse
08:40 juancorr
@Jim__: semrush is our fan number 1. We have more visits from semrush than Google
08:40 juancorr
juan
11:44 pdurbin joined #dataverse
12:25 donsizemore joined #dataverse
12:26 donsizemore
mr @pdurbin sir may i pester you one more time?
12:27 pdurbin
of course!
12:29 donsizemore
@pdurbin only because you and kevin have a difference of opinion. see the bottom two responses of https://help.hmdc.harvard.edu/Ticket/Display.html?id=268461 ?
12:29 donsizemore
upgrading to 4.9.3+ would solve this, but we're not ready just yet
12:31 pdurbin
Any insight from the authenticateduserlookup table?
12:35 donsizemore
a select * from authenticateduserlookup where authenticateduser_id= returns her authenticateduser id, but her authenticationproviderid is 'builtin'
12:36 pdurbin
Is it possible that she has a single account in the system but there's a mismatch between the email address in the authenticateduser table and the email address on the builtinuser table?
12:37 donsizemore
i think so. kevin says just make the e-mail addresses match
12:38 pdurbin
You could. But in the ticket you say she wants two accounts.
12:38 pdurbin
Or maybe I'm misunderstanding.
12:39 donsizemore
i'll ask her if i can squish them. i wish i could just upgrade
12:40 pdurbin
If she only has one account in the system, making the email addresses match is the way to squish them. Yes.
12:41 pdurbin
Once she's all set she could always go create a second account if she really wants to.
12:44 pdurbin
Upgrading is the other way to fix this. No more potential for mismatched email addresses.
12:44 pdurbin
Yesterday I was saying upgrading wouldn't help because you were saying there's no entry at all in the builtinuser table. Or that was my understanding anyway.
12:45 pdurbin
But there she is, so you can just make the email addresses match.
12:47 donsizemore
@pdurbin there was no entry for her e-mail in the builtinuser table, i thought that was the unique identifier. as long as we're good i'm good =)
12:51 pdurbin
nope, the email is not the unique identifier
12:51 pdurbin
It's confusing. See https://github.com/IQSS/dataverse/issues/1151 for more confusion. :)
12:54 donsizemore
i was looking at that issue earlier this morning =)
13:03 pdurbin
:)
13:59 cdsp-rmo
pdurbin: sorry for the pull request #5223, it was meant to be a WIP :S
13:59 cdsp-rmo
but thanks for the review, I may check this points
14:00 cdsp-rmo
I have a question about a problem I have, related to translations
14:01 cdsp-rmo
http://image.noelshack.com/fichiers/2018/43/2/1540303281-screenshot-from-2018-10-23-16-01-03.png
14:02 cdsp-rmo
Here is the french section of user notifications
14:02 cdsp-rmo
and here the english one
14:02 cdsp-rmo
http://image.noelshack.com/fichiers/2018/43/2/1540303281-screenshot-from-2018-10-23-16-00-45.png
14:02 cdsp-rmo
on the englis one, you have the link for the permissions of the related dataset
14:02 cdsp-rmo
but not for french
14:03 cdsp-rmo
is it something already mentionned in a ticket ? Or should I open one ?
14:09 donsizemore joined #dataverse
14:10 juancorr joined #dataverse
14:53 pameyer joined #dataverse
15:08 pdurbin
cdsp-rmo: hi! Please open a new GitHub issue, especially since I'm getting 503 errors when clicking those images. Please attach them to the issue. Thanks!
15:17 andrewSC joined #dataverse
15:25 cdsp-rmo
ah
15:25 cdsp-rmo
ok
15:33 pdurbin
thanks!
15:34 cdsp-rmo
https://github.com/IQSS/dataverse/issues/5226
15:34 cdsp-rmo
should be better like this
15:34 cdsp-rmo
;)
16:02 pdurbin
cdsp-rmo: thanks!
16:02 pdurbin
The community call is about to start: https://dataverse.org/community-calls
16:14 andrewSC joined #dataverse
16:21 Jim_ joined #dataverse
16:24 jri joined #dataverse
16:26 jri joined #dataverse
17:20 donsizemore joined #dataverse
18:39 andrewSC
kinda on that struggle bus right now
18:40 andrewSC
how do i add a value for a custom metadata type?
18:40 andrewSC
s/for/to/
18:41 pdurbin
andrewSC: have you read this (new) page already? http://guides.dataverse.org/en/4.9.4/admin/metadatacustomization.html
18:41 andrewSC
i have but i'm still in a bit of a gray area?
18:41 andrewSC
https://gist.github.com/andrewSC/f775ff94a34e91c2a19e8a9cb51778a6
18:42 pdurbin
ah, you're adding a controlled vocabulary
18:42 andrewSC
i'm trying to figure out where exactly DV is storing those values and how to add one to that list?
18:42 andrewSC
ye
18:43 pdurbin
andrewSC: here's a recent example: https://github.com/IQSS/dataverse/pull/5161/files
18:43 andrewSC
mmmmmmmmmmmmmmm
18:44 andrewSC
so to make the change i would need to redeploy?
18:44 pdurbin
Well, here's how you re-run the curl command to load it up: https://github.com/IQSS/dataverse/issues/5151#issuecomment-428593280
18:45 andrewSC
ahhh gotcha
18:58 andrewSC
ahhh haaaa, i found the tsv used for the custom metadata from whoever set this up before me.. now i feel slightly better :)
19:00 pdurbin
phew
19:00 pdurbin
Is it the kind of thing you'd like to contribute back?
19:08 andrewSC
unfortunately not, the whole tsv is specific to some metadata we leverage internally
19:08 pdurbin
ah, no worries
19:09 pdurbin
The structural biology metadata block developed by pameyer might be of interest to other installations.
19:10 andrewSC
mmmmmmmmmmmm
19:11 pameyer
that block isn't in the dataverse repo, but I'm pretty sure it's a publicly visible google sheet somewhere
19:11 andrewSC
wrt to the curl command you mentioned earlier, will that still work if the file provided isn't specificly the citation.tsv?
19:11 pdurbin
pameyer: you might want to let your researchers use it in anger for a while before sharing it more broadly. Up to you. :)
19:11 pameyer
andrewSC: my usual practice for playing with metadata blocks is to wipe out everything on the dataverse side when I change the block; otherwise there's possible unexpected behavior
19:12 pameyer
pdurbin: good point
19:12 andrewSC
mmmmmmmmmmmmmmmmmmm
19:13 pdurbin
andrewSC: no matter what I say here, you should leave a comment at https://github.com/IQSS/dataverse/issues/5205 which cdsp-rmo opened about questions you have. We intend to document the curl commands and such better.
19:14 pdurbin
But pameyer has a good point. I usually get set up to drop my database frequently when iterating on changes to metadata blocks.
19:15 pdurbin
But to answer your question, the filename shouldn't matter.
19:15 pdurbin
A stock installation loads up 5 or six blocks. I forget.
19:16 pdurbin
six: https://github.com/IQSS/dataverse/blob/v4.9.4/scripts/api/setup-datasetfields.sh
19:22 pameyer
also, if you fork the citation block; the api tests will fail.
19:22 pameyer
which is legitamite; you're not supposed to fork the citation block
19:25 pdurbin
Yeah, for any of those six blocks we ship, you should make pull requests (like the one above) to improve them. Not recommended to fork.
19:32 pameyer
well, I'm pretty sure that you wouldn't accept a PR to remove all but 1 subject ;)
19:32 pdurbin
Nope. But that would make a good issue.
19:33 pdurbin
We had a little momentum on changing how subjects are handled but not anymore.
20:03 pdurbin
pameyer: I'm coming at you to ask about https://github.com/IQSS/dataverse/issues/4371#issuecomment-428687428
20:27 Guest54726
@andrew - fwiw https://docs.google.com/document/d/1idh4wvCcoxSo7rPkolOkTD5Wvn6PQO__uXqkf7eQujA/edit?usp=sharing has some info on what works/doesn't work when trying to update a tsv via the curl command
20:28 Guest54726
as pameyer says, best practice is to start fresh, but it is possible to update descriptions/add terms etc. (the section I'm talking about is mostly in the suggested text at the bottom).
20:29 Guest54726
and there are things that won't work...
20:31 Jim__ joined #dataverse
20:33 Jim__
BTW - with the recent updates merged into dev to add URIs for terms in the tsv files, the update instructions should include running curl to update the tsv with those additions...
20:34 Jim__
Is there a place to put things that should be in the release notes like that (for 4.9.10 I assume)
20:36 pdurbin
Jim__: Kevin usually has a running Google doc of stuff that will end up being in release notes. You might want to ask him if he has started one for the next release yet.
20:37 Jim__
OK - thanks
20:39 pdurbin
sure
21:44 jri joined #dataverse
