Time
S
Nick
Message
05:02 poikilotherm joined #dataverse
06:35 juancorr joined #dataverse
06:57 Slava1 joined #dataverse
08:25 stefankasberger joined #dataverse
08:52 Slava1 joined #dataverse
10:10 pdurbin
pmauduit: you might want to ask poikilotherm about https://github.com/IQSS/dataverse-docker/issues/25
10:10 poikilotherm
Morning guys
10:16 pdurbin
poikilotherm: it looks like someone added a "people" column to the new "dataverse meetups" spreadsheet linked from the bottom of https://dataverse.org/events and your name is there.
10:16 poikilotherm
Oh?
10:17 poikilotherm
Did you do so?
10:17 * poikilotherm
goes looking
10:19 poikilotherm
I left some comments :-)
10:26 poikilotherm
pdurbin can you point me to some issues or other places why you choosed to request scopes from google to read email, etc, but not from Github?
10:28 pdurbin
not me
10:28 poikilotherm
With you I meant IQSS :_D
10:28 poikilotherm
Sorry
10:28 pdurbin
I mean I didn't add the People column and I didn't add your name. :)
10:30 pdurbin
poikilotherm: you're saying we ask for email from Google but not GitHub? That seems a little... inconsistent.
10:30 poikilotherm
Right.
10:30 poikilotherm
I stumbled over this because I got an NPE during adding the scope
10:30 poikilotherm
And this is due to the Github impl not using any scope at all
10:32 poikilotherm
https://i.imgur.com/lNsrDoP.png
10:33 pdurbin
I see what you mean.
10:33 poikilotherm
I would change that to include scopes read:profile and user:email, so it will become this
10:33 poikilotherm
https://i.imgur.com/74CtqWI.png
10:33 poikilotherm
But if there are good reasons why you guys changed that, I will refactor
10:36 pdurbin
I suspect asking for email from GitHub will result in a better user experience. Right now people who sign up with GitHub probably have to type in their email address manually. I guess the upside of the current approach is Dataverse not learning about your GitHub email address if you don't want it to be revealed.
10:36 poikilotherm
Right.
10:38 poikilotherm
On Github there is no option to concent only to access the profile and not the mail
10:38 pdurbin
wow, 168 commits at https://github.com/IQSS/dataverse/pull/3539/commits
10:49 pdurbin
Yesterday I asked Michael about if scope needs to be saved to the database for ORCID login and he said he doesn't remember. I don't either. We can't always look to the ancients.
10:53 pdurbin
poikilotherm: I just added a comment to the "2015-04-30 Dataverse User Accounts and Auth Meeting" doc under "email address": https://docs.google.com/document/d/157sw9gaFGwrb0EtCGGlLMxSg0cooHfKEvIsjcebPcA8/edit?usp=sharing
10:56 poikilotherm
I will test with an empty scope for now
10:56 poikilotherm
And see if that works and results in the same
10:58 pdurbin
Lots of auth in the air. Lots of discussion about that pull request for Microsoft auth.
11:00 poikilotherm
I really hope I don't have to refactor that one too :-/
11:04 poikilotherm
Meh. Empty scope doesn't work - illegal.
11:04 poikilotherm
It has to be NO scope
11:05 pdurbin
Should IQSS upgrade ScribeJava for you as part of the Microsoft pull request?
11:05 poikilotherm
For me?
11:05 poikilotherm
You mean you would do the same thing as I am doing now?
11:06 pdurbin
Well, maybe not everything. But maybe we could get us to the latest version of ScribeJava.
11:06 poikilotherm
Because the refactoring is due to the upgrade, not because I just enjoy it ;-)
11:07 pdurbin
I'm just thinking that if we have to put OAuth stuff through QA anyway, maybe some devs at IQSS should hack on that Microsoft branch and update ScribeJava to the latest.
11:08 pdurbin
I'm already thinking we'll probabaly need to get in the docs and clean them up. I just left a comment about this on this issue you opened: https://github.com/IQSS/dataverse/issues/5665
11:13 pdurbin
Anyway, please think about it.
11:13 poikilotherm
Well, we could just work together on this
11:13 poikilotherm
You guys have push access to my branch
11:13 poikilotherm
And this is almost done
11:14 poikilotherm
Some minor tweaks necessary and some polishing
11:23 pdurbin
That's a good point. Here's an idea. What if you start with the Microsoft branch and create a new branch than contains your refactoring and ScribeJava upgrade. That way, we'd have some options when we decide which branch to test when it comes to QA.
11:51 pdurbin
I keep saying that branches are free. :)
12:11 donsizemore joined #dataverse
12:11 dataverse-user joined #dataverse
12:17 dataverse-user
Hi! we're having an issue in our Dataverse Repository, so I would really appreciate your help with this: We are trying to delete a dataset that it's not published yet, but the Repository says it can't be deleted. We are checking the log, and it shows this error:s3://urdataverse:16d5055d2b0-1ca22b98c0d1]] Do you know which can be the reason for this to happen?
12:34 donsizemore
@dataverse-user is there any more descriptive error?
12:43 pdurbin_m joined #dataverse
12:43 pdurbin_m
dataverse-user: if you can provide your server.log file it would be helpful. And welcome!
12:46 pdurbin_m
donsizemore: https://github.com/whole-tale/wt-design-docs/pull/100
12:46 pdurbin_m
xarthisius: ^^
13:09 donsizemore
@pdurbin eh? eh? https://github.com/IQSS/dataverse-ansible/blob/master/files/wholetale.json
13:24 pdurbin
donsizemore: much better. Thank you!
13:24 pdurbin
dataverse-user: still there? We'd like to help.
13:27 poikilotherm
pdurbin: do you have an ORCID sandbox api account lying around?
13:28 pdurbin
poikilotherm: let me look in our ticketing system. I assume it's expired by now.
13:29 poikilotherm
I do have Github and Google running, looks reliable. But I want to test ORCID, too, as ORCID is very special with custom data retrieval etc
13:29 pdurbin
donsizemore: can you please take a look at this license sitemap question: https://groups.google.com/d/msg/dataverse-community/RZb6SsuL17E/Aj1sqX7-AQAJ ?
13:30 pdurbin
poikilotherm: and ORCID uses XML instead of JSON , right?
13:31 poikilotherm
At least from what I see in the code, yes
13:31 poikilotherm
I didn't take a look at the ORCID docs yet
13:32 poikilotherm
(So I dunno if they switched to JSON )
13:32 pdurbin
poikilotherm: we have tech hours this afternoon. Do you want to call in?
13:32 donsizemore
@pdurbin oh don't worry, Dataverse UNC got dinged for more than the license =)
13:33 poikilotherm
That's going to be tricky :-( Need to look after kids as my wife is busy this afternoon
13:34 pdurbin
poikilotherm: ok, maybe some other time. I found some old ORCID sandbox creds at https://help.hmdc.harvard.edu/Ticket/Display.html?id=243207 but they're for an old server (shibtest) that is long gone. RIP. donsizemore might remember shibtest.
13:34 poikilotherm
:-D
13:34 poikilotherm
Ok then I need to ask ORCID for a new one
13:34 poikilotherm
Might take a few days
13:34 pdurbin
yeah
13:35 poikilotherm
They are a bit unresponsive...
13:35 pdurbin
hmm, sorry to hear that, I could try my contact if you need me to
13:35 pdurbin
if you come to PIDapolooza you'll meet ORCID people :)
13:36 pdurbin
donsizemore: do you want to reply on that license/sitemap thread? Or create an issue? :)
13:36 donsizemore
@pdurbin i wanted to better dig into what's going on first. i'm new to sitemaps as of friday
13:36 pdurbin
ok, that's fine, mostly I just wanted you to know you're not alone :)
13:37 donsizemore
yeah, i saw the traffic this morning. if i know google, though, it's a generic warning about a handful of errors that they're ignoring in the first place
13:37 donsizemore
i just wanted to make sure that's the case
13:37 pdurbin
ok
13:38 pdurbin
poikilotherm: what should I say about the microsoft auth pull request at standup?
13:38 poikilotherm
Hmm.
13:39 donsizemore
@pdurbin i think a login failure should be presented with white text on blue background
13:39 pdurbin
donsizemore: instead of red on green? :)
13:40 poikilotherm
I really don't know what the best approach.
13:40 poikilotherm
This is almost finished.
13:41 pdurbin
donsizemore: speaking of color blindness, I've been meaning point you toward this new feature of visbug: https://github.com/GoogleChromeLabs/ProjectVisBug/pull/381
13:41 donsizemore
which one was the "guru meditation" with the ASCII floppy disk?
13:41 donsizemore
(you're thinking color schemes, i'm thinking BSOD)
13:42 pdurbin
better that than the BOFH
13:43 poikilotherm
It would be awesome to have the Microsoft auth in the codebase if the community wants and needs it
13:43 pdurbin
poikilotherm: that's something I'm struggling a bit with. How much demand is there for support for logging into Dataverse with an Active Directory account. I *think* this would be a very nice and popular feature.
13:44 pdurbin
I don't think anyone cares if Dataverse supports live.com accounts and I'm not even sure if the pull request supports those.
13:44 poikilotherm
Well, please keep in mind that this will not work with an AD on premises
13:44 pdurbin
it won't?
13:46 pdurbin
are you sure?
13:48 poikilotherm
https://docs.microsoft.com/en-us/azure/architecture/reference-architectures/identity/azure-ad
13:48 poikilotherm
You can integrate
13:48 xarthisius
pdurbin: hi! Can you point me to docs describing datasetId vs datasetPid ?
13:48 poikilotherm
But you need to have an Azure AD for this
13:50 xarthisius
I got it
13:50 pdurbin
xarthisius: you're all set?
13:50 xarthisius
pdurbin: I thought so but http://guides.dataverse.org/en/latest/installation/external-tools.html describes only datasetId
13:50 xarthisius
what's datasetPid then?
13:51 pdurbin
xarthisius: lemme go build from "develop". One sec please.
13:51 xarthisius
sure, thanks!
13:51 * xarthisius
needs a cup of coffee
13:54 poikilotherm
pdurbin: I dunno if this is a feasible setup for people out there
13:55 poikilotherm
It might be a good idea to look into using Java EE auth methods
13:55 poikilotherm
And directly support things like LDAP
13:56 poikilotherm
Java EE 8 provides a whole new set of this stuff
13:56 poikilotherm
Get more standards based again
13:58 pdurbin
xarthisius: please see http://guides.dataverse.org/en/develop/api/external-tools.html ... I did a pretty serious rewrite and I'd love feedback on that page.
13:59 pdurbin
poikilotherm: yeah. And hour and 15 minutes until standup. Still not sure what I should say. But there's always tomorrow.
13:59 pdurbin
donsizemore: tomorrow is now today so I'd love to talk about burritos etc today.
14:03 donsizemore
@pdurbin FWIW our 4 complaints from the Goog all involve https://dataverse.unc.edu/dataset.xhtml?persistentId=doi:10.15139/S3/12438
14:03 donsizemore
@pdurbin yes i fixed my PUT/POST error and am testing granting auth on root in the task. will push when it completes successfully
14:03 pdurbin
donsizemore: sitemap complaints or dataset search validator tool complaints?
14:03 donsizemore
(but also i love burritos, so tell me more)
14:04 donsizemore
@pdurbin shrug. probably the latter but pulling from the former
14:06 pdurbin
donsizemore: granting auth on root is perfect! Thanks! I was going to say that what we *could* do instead is support the -insecure flag: https://github.com/IQSS/dataverse/blob/v4.16/scripts/api/setup-all.sh#L12 ... but now I see that I'm still calling grant-authusers-add-on-root at https://github.com/IQSS/dataverse/blob/v4.16/scripts/deploy/phoenix.dataverse.org/post#L10 anyway, so you've got the right
14:06 pdurbin
approach.
14:07 xarthisius
pdurbin: can queryParameters contain arbitrary values (e.g. can I pass "scope": "datasets" in there?)
14:09 pdurbin
xarthisius: I don't think so but I'm testing it.
14:12 pdurbin
xarthisius: nope. {"status":"ERROR","message":"Unknown reserved word: datasets"}. But please keep talking. Maybe there's another way to get what you want. You could offer two manifest files. One at the file level and one at the dataset level. Better advertising for Whole Tale that way. :)
14:14 xarthisius
pdurbin: I just wanted to simplify some logic in the thing that handles ext tool url, but I'll work around it
14:14 xarthisius
manifest is one thing, but I need to update the endpoint it goes to ;)
14:15 xarthisius
and I can't assume everyone will update DV to 4.16 so I need to be backward compatible
14:15 pdurbin
xarthisius: yes! You'll want to expect datasetPid=doi:10.7910/DVN/TJCLKP or whatever.
14:16 pdurbin
Yes, it would be nice to offer the Whole Tale file level manifest for installations of Dataverse that are still on Dataverse 4.16 or lower. But this is up to you, of course.
14:20 hblancoca joined #dataverse
14:20 poikilotherm
pdurbin you might take a look at https://javaee.github.io/security-spec for more information...
14:20 pdurbin
hmm, is there a newer ee4j or jakartaee equivalent?
14:21 poikilotherm
This should be in Jakarta, yes
14:22 pdurbin
https://jakarta.ee/specifications/security/ right?
14:22 dataverse-user
pdurbin: sorry I was out of the office..Thanks! I'll send the log file
14:22 poikilotherm
https://jakarta.ee/specifications/security/1.0/apidocs/
14:22 poikilotherm
Ay
14:22 poikilotherm
+e
14:22 pdurbin
dataverse-user: thanks! You can email it to support
14:23 pdurbin
poikilotherm: I'm a little lost in this spec but I'm not opposed to using it.
14:23 dataverse-user93 joined #dataverse
14:23 pdurbin
hblancoca: welcome!
14:24 hblancoca
hi pdurbin
14:24 poikilotherm
pdurbin: see also https://blog.payara.fish/using-jakarta-security-on-tomcat for more :-)
14:24 poikilotherm
Hi hblancoca
14:40 hblancoca
anyone knows if exist any way to replace doi for unpublished dataset? may be through database?
14:42 pdurbin
hblancoca: you want a different DOI for a dataset than the one in the database?
14:43 hblancoca
yes, because i upload dataset before config handle service
14:47 pdurbin
hblancoca: I really hate saying that you might need to hack on your database but it might be the easiest way. Please look at the dvobject table, the columns authority, identifier, and protocol: http://phoenix.dataverse.org/schemaspy/latest/tables/dvobject.html
14:47 pdurbin
(I assume you're running a relatively recent version of Dataverse. Long ago these were in the dataset table.)
14:50 hblancoca
pdurbin: we are in 4.9.4 dataverse version. the persistent id asigned was "doi:10.5072/FK2/2HYT3J"
14:51 pdurbin
ok 10.5072 is just for testing anyway: https://github.com/IQSS/dataverse/issues/5448
14:52 pdurbin
hblancoca: on second thought, instead of me telling you to hack on your database... can you please email support
14:53 donsizemore
@pdurbin try the api-test-suite now? someone was in my office and i put the change into master, but... try it?
14:57 pdurbin
donsizemore: can do. Thanks!
15:13 pdurbin
donsizemore: [ERROR] Tests run: 46, Failures: 29, Errors: 4, Skipped: 2
15:15 pdurbin
fewer than yesterday :)
15:15 pdurbin
standup time
15:17 donsizemore
@pdurbin p.s. i'm testing our current Dataverse install on CentOS 8
16:02 pdurbin
wat
16:02 pdurbin
that's awesome
16:04 pdurbin
andrewSC bjonnh bricas_ hblancoca juancorr pmauduit stefankasberger xarthisius the community call is about to start: https://dataverse.org/community-calls
17:33 donsizemore joined #dataverse
17:36 dataverse-user joined #dataverse
17:36 dataverse-user
I have a question about how to download datasets
17:37 dataverse-user
I am a grad student at UC Davis
17:37 dataverse-user
I am trying to download this data "Databanks International, 2018, "2016 Edition LINKS.tab", Cross-national time-series data archive"
17:37 dataverse-user
But can't figure out how to do that. Please help.
17:39 pdurbin
dataverse-user: hi! From here? https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/K2SSTK
17:39 donsizemore
@dataverse-user in the top-left of the file list, in the header, there's an unlabelled check-box all by itself
17:39 donsizemore
@dataverse-user check that box, then click "select all nn files"
17:39 donsizemore
whoops i'll let @pdurbin take over
17:39 pdurbin
donsizemore: heh. No worries. The data seems to be restricted.
17:40 donsizemore
information wants to be free.
17:40 pdurbin
And no "Request Access" button.
17:40 pdurbin
which you can turn off
17:41 pdurbin
donsizemore: under "Notes" it says "MIT affiliates should access this dataset by logging into Dataverse and selecting Massachusetts Institute of Technology."
17:41 dataverse-user
The guy who manages this data website told me it should be accessible through our library. But I cannot download it even through our library
17:42 pdurbin
dataverse-user: your best bet is probably the "Contact" button (next to "Share") to ask the dataset owner if you can have the data.
17:42 pdurbin
donsizemore and I are not the dataset owners. :)
17:43 pdurbin
Which data website are we talking about?
17:43 pdurbin
donsizemore: I'm making progress, by the way: Tests run: 23, Failures: 9, Errors: 0, Skipped: 0
17:44 donsizemore
glad to hear it!
17:44 dataverse-user
Okay. Thanks!
17:44 pdurbin
dataverse-user: you're welcome. Good luck!
17:44 pdurbin
donsizemore: I'm using `./run-test-suite.sh http://localhost:8080
17:45 pdurbin
instead of ./run-test-suite.sh siteUrl
17:45 pdurbin
I'm not even sure what siteUrl is
17:45 pdurbin
Can we change it to localhost?
17:45 donsizemore
it's a variable set in ansible. you may safely use localhost instead
17:46 pdurbin
Oh! I have control in the yaml?
17:46 dataverse-user left #dataverse
17:47 pdurbin
I think all I have in the yaml is an all or nothing boolean. Should I make a pull request to change it to localhost?
17:49 pdurbin
or do we want or need it to work with http://ec2-3-81-100-151.compute-1.amazonaws.com or whatever
17:49 pdurbin
sorry, thinking out loud
17:50 pdurbin
donsizemore: actually, next we probably need this: http://guides.dataverse.org/en/4.16/developers/testing.html#publish-root-dataverse
17:50 donsizemore
the idea was to make it work remotely
17:50 pdurbin
right, right
17:51 pdurbin
but the root dataverse needs to be published
17:51 donsizemore
i'll drop in the publish root dataverse task next
17:51 pdurbin
perfect, thanks!
17:57 pdurbin
[ERROR] Tests run: 103, Failures: 0, Errors: 4, Skipped: 4
17:57 pdurbin
that's after I publish the root dataverse (still using localhost) ^^. Progress.
17:59 pdurbin
trying it with the siteURl
18:02 pdurbin
[ERROR] Tests run: 46, Failures: 27, Errors: 4, Skipped: 2 when I do ./run-test-suite.sh http://ec2-3-81-100-151.compute-1.amazonaws.com
18:03 pdurbin
same when I do ./run-test-suite.sh http://ec2-3-81-100-151.compute-1.amazonaws.com:8080 (with the port)
18:09 pdurbin
Ah ha. { status:"error", message:"Endpoint available from localhost only. Please contact the dataverse administrator"}
18:12 donsizemore joined #dataverse
18:13 donsizemore
@pdurbin i'll hard-code it to localhost
18:13 pdurbin
donsizemore: no no, I'm good.
18:13 pdurbin
but!
18:13 donsizemore
@pdurbin assuming i can stay on UNC's wi-fi long enough to type anything
18:13 pdurbin
can you run setup.all with -insecure?
18:13 pdurbin
that'll open up the admin APIs
18:14 donsizemore
i can make it an option, but... we can realistically just hard-code the API tasks to localhost
18:14 pdurbin
I'm fine with whatever. Should we focus on getting them to pass on localhost first? Then worry about getting them passing over the wire?
18:16 donsizemore
p.s. Dataverse-on-CentOS 8 snafu 1: no R en EPEL8 just yet
18:16 pdurbin
bah
18:17 pdurbin
Can I open an issue for the ec2 script to accept a switch for centos 7 vs 8?
18:25 donsizemore
i'm imagining the underlying change will be transparent once all the packages are present
18:26 pdurbin
cool
18:27 pdurbin
maybe a switch for ubuntu too :)
18:27 pdurbin
but I'm getting ahead of myself
18:29 pdurbin
thanks for trying centos 8
18:29 pdurbin
I had a note to tell you it's coming out today. But as usual, you're way ahead of me.
18:34 pdurbin
hmm, this seems to fail quite reliably: mvn test -Dtest=DatasetsIT#testSequentialNumberAsIdentifierGenerationStyle
18:35 donsizemore
@pdurbin switching wholetale.json back Unknown reserved word: {datasetPid}
18:35 pdurbin
donsizemore: unknown on "release" (4.16) but not on "develop"
18:36 pdurbin
but yeah, xarthisius and were just talking about maybe keeping the file level Whole Tale button around
18:36 pdurbin
for Dataverse 4.16 and lower
18:37 pdurbin
so switching back sounds fine
18:37 donsizemore joined #dataverse
18:38 donsizemore
@pdurbin and the 4.17 release is some time next week...? :eyebrow wiggle:
18:39 pdurbin
donsizemore: next week or the week after according to https://groups.google.com/d/msg/dataverse-community/RbpmnLm8nv4/iYbpMKn2BwAJ
18:39 pdurbin
but there's at least one show stopper bug right now
18:40 donsizemore
uh oh...?
18:41 donsizemore
@pdurbin well, i'm just disabling wholetale in my group_vars for now ;)
18:41 pdurbin
well, we have a fix but we don't like the fix :)
18:42 pdurbin
that's fine. let's keep an eye on the next move by xarthisius :)
18:45 pdurbin
donsizemore: Caused by: org.postgresql.util.PSQLException: ERROR: function generateidentifierassequentialnumber() does not exist
18:48 pdurbin
donsizemore: there's a script we can download from http://guides.dataverse.org/en/4.16/installation/config.html#identifiergenerationstyle
18:48 pdurbin
this one: http://guides.dataverse.org/en/4.16/_downloads/createsequence.sql
18:48 pdurbin
It's also in the code, of course.
18:50 pdurbin
Am I making any sense? :)
18:52 pdurbin
donsizemore: here's where phoenix loads up the sql file: https://github.com/IQSS/dataverse/blob/v4.16/scripts/deploy/phoenix.dataverse.org/post#L6
18:57 pdurbin
uh oh
18:57 pdurbin
psql:/tmp/createsequence.sql :20: ERROR: role "dvnapp" does not exist
18:57 poikilotherm joined #dataverse
18:59 poikilotherm
Hey guys, anyone still around?
19:01 pdurbin
poikilotherm: tech hours is about to begin
19:01 poikilotherm
:-)
19:01 poikilotherm
Go for it
19:01 poikilotherm
Sry I couldn't make it to Community call - had to look after kids
19:02 poikilotherm
If you want me to call in now, I can grab my headphones
19:04 pdurbin
I can ask. The first topic is https://github.com/IQSS/dataverse/issues/6208 I think.
19:04 pdurbin
What do you want to talk about? :)
19:06 poikilotherm
I dunno if you guys want to talk about #5991/#5997/microsoft
19:06 pdurbin
It's not on the agenda.
19:06 poikilotherm
Ok
19:06 pdurbin
After 6208 I'm hoping to talk about getting off Glassfish 4.
19:07 poikilotherm
Oh that's another VERY interesting ho topic
19:07 poikilotherm
+t
19:07 poikilotherm
If you think I should join or do anything - just shout out. I'll be around
19:07 pdurbin
Your name came up already.
19:08 pdurbin
Did you break JSF?
19:08 poikilotherm
What?
19:08 pdurbin
Did you cause 6208?
19:08 * poikilotherm
scratches his head
19:08 poikilotherm
How would that have happened?
19:08 pdurbin
offline mode, maybe?
19:09 poikilotherm
O.O
19:10 poikilotherm
All of this hasn't been touched
19:10 poikilotherm
THat plugin is totally optional
19:10 poikilotherm
Not in use by default, et
19:11 poikilotherm
Recent stuff that got merged was about Endorsed Java API and Solr config
19:12 poikilotherm
So if you revert that Endorsed and it starts working again (which seems very unlikely), it was me
19:12 poikilotherm
Right?
19:14 poikilotherm
https://github.com/IQSS/dataverse/commits?author=poikilotherm
19:15 pdurbin
right
19:17 poikilotherm
And if it really is that Endorsed thing... OMG... No words.
19:20 poikilotherm
And to my defense: http://irclog.iq.harvard.edu/dataverse/2019-08-29#i_103686
19:20 poikilotherm
;-)
19:23 pdurbin
heh. "I dunno if this is necessary for Netbeans"
19:25 poikilotherm
I googled a lot about this before I created the PR and all I could find on this was about Netbeans missing it back in 2011/2012, but nothing related to JSF/...
19:27 pdurbin
talking Glassfish 4 now
19:29 poikilotherm
I'm curious about the findings Gustavo talked about... I would be very glad if we can share this with everyone :-)
19:33 donsizemore joined #dataverse
19:47 pdurbin
ok
20:05 pdurbin
done
20:05 pdurbin
donsizemore: so can we please run createsequence.sql ? We need it for the API test suite.
20:06 pdurbin
Unfortunately, there's a hardcoded "dvnapp" in there and you use "dvnuser".
20:08 donsizemore
i'll add that next
20:08 pdurbin
Thanks! Should we switch away from "dvnuser" for now?
20:08 pdurbin
In the yaml, I mean.
20:08 donsizemore
isn't it a group_var?
20:08 pdurbin
It is.
20:09 donsizemore
i can make dvnuser the default, but in ansible land i can edit createsequence.sql to use the specified database user
20:09 pdurbin
Oh, if it's easy, please go for it!
20:22 pdurbin
oooh a 500 error
20:23 pdurbin
donsizemore: well, I'm seeing similar results to what you were seeing in Docker. I'm on EC2. The API test suite is *close* to passing. I can taste it.
20:26 poikilotherm
pdurbin: woohoo I'm off the hook :-D
20:26 poikilotherm
https://github.com/IQSS/dataverse/issues/6208#issuecomment-534719877
20:28 pdurbin
poikilotherm: check out the latest comment.:)
20:28 poikilotherm
Phew...
20:29 pdurbin
:)
20:29 pdurbin
sorry to raise your anxiety level
20:29 poikilotherm
That's ok
20:29 poikilotherm
It would have worried me if it would have been those changes
20:29 poikilotherm
That would have indicated for some bug in Netbeans
20:30 poikilotherm
So now I'm just dancing like Elmo
20:30 poikilotherm
;_)
20:34 pdurbin
:)
21:02 pdurbin
donsizemore: now the community is using the spin up scripts: https://github.com/IQSS/dataverse/issues/6154#issuecomment-534709219 Awesome!
21:19 poikilotherm
Good night guys. Read you tomorrow.
