Time
S
Nick
Message
07:35 juancorr joined #dataverse
08:02 bjonnh joined #dataverse
08:13 juancorr
Hi @hblancoca . Do you have fixed the problem?. The link to the file works for mi.
08:16 jri joined #dataverse
08:48 jri_ joined #dataverse
09:47 Benjamin_Peuch joined #dataverse
10:05 Benjamin_Peuch
Hello pdurbin.
10:07 Benjamin_Peuch
I wanted to ask you something. Let's say we have cut off access to our Dataverse save for users who have an institutional log-in. But then we want to let someone in who doesn't have one. The feature that can help us in this situation is the API command for creating builtin users, is that correct?
12:02 pdurbin
Benjamin_Peuch: that does sound like something installations have done, yes. Maybe NTU in Singapore? I'm not sure.
12:13 Benjamin_Peuch
Oh, another Dataverse to look into!
12:13 Benjamin_Peuch
I'll ask them. Thanks for the tip, pdurbin. :)
12:21 pdurbin
Benjamin_Peuch: sure, you can see folks from NTU (and UVa) in this thread, which is the one I was thinking about: https://groups.google.com/d/msg/dataverse-community/EXApk46Jm2I/D-McdTFWAgAJ
12:39 Benjamin_Peuch
Indeed it seems to be the same situation. Thanks!
12:40 GitterIntegratio joined #dataverse
12:40 icarito[m] joined #dataverse
12:40 poikilotherm joined #dataverse
12:56 Benjamin_Peuch
I was also wondering if someone could help me with UNFs.
12:57 Benjamin_Peuch
I read (here: https://rdrr.io/cran/UNF/f/README.md) that CRAN is "a network of ftp and web servers around the world" and that packages must be "submitted" to it.
12:57 Benjamin_Peuch
Does this mean that, if I want to produce a UNF without using Dataverse, I cannot do this locally? I have to send an online request?
12:57 Benjamin_Peuch
Sorry if this isn't the right place to ask this.
13:15 pdurbin_m joined #dataverse
13:17 pdurbin_m
Benjamin_Peuch: you can calculate a UNF locally if you want. You don't need Dataverse. Also CRAN is like npm or pypi or CPAN or Maven Central or Ruby gems. It's for R packages.
13:43 Benjamin_Peuch
Okay. I thought it had to be possible. Thanks again, pdurbin!
13:47 donsizemore joined #dataverse
14:06 Youssef_Ouahalou joined #dataverse
14:06 Youssef_Ouahalou
hello ☺
14:14 Benjamin_Peuch
Hi, fellow colleague
14:16 Youssef_Ouahalou
hahahah hi
14:31 Youssef_Ouahalou
I have a little problem with this command:curl -H "X-Dataverse-key: $API_TOKEN" -X POST $SERVER_URL/api/dataverses/$DV_ALIAS/datasets/:import?pid=$PERSISTENT_IDENTIFIER&release=yes --upload-file dataset.json
14:32 Youssef_Ouahalou
from this: http://guides.dataverse.org/en/4.18/api/native-api.html?highlight=already%20doi#create-dataset-command
14:46 pdurbin
Youssef_Ouahalou: you're trying to import a dataset that already has a DOI?
14:47 Youssef_Ouahalou
yes
14:48 pdurbin
We built that feature for an installation that was already minting DOIs with their homegrown system before they migrated to Dataverse. So they're using the same DOI namespace (authority) as before. It should work.
14:52 Youssef_Ouahalou
because when I type the command I have this as a result: -bash: --upload-file: command not found
14:52 Youssef_Ouahalou
{"status":"ERROR","message":"Error parsing Json: Invalid token=EOF at (line no=1, column no=0, offset=-1). Expected tokens are: [CURLYOPEN, SQUAREOPEN]"}
14:56 donsizemore
@pdurbin knock knock?
14:59 Benjamin_Peuch
We've looked into the error Youssef_Ouahalou mentions and we couldn't make head nor tail of it.
14:59 Benjamin_Peuch
The expected tokens bit seems to be about a syntax error, but the input JSON file is quite clean.
15:00 Benjamin_Peuch
There doesn't seem to be missing brackets of any kind (if that's what is meant in the error message).
15:00 pdurbin
Youssef_Ouahalou: I think the & is confusing bash. Can you please try putting the entire URL in double quotes?
15:01 pdurbin
If this works, I'd call it a bug in the documentation.
15:02 pdurbin
donsizemore: good morning. I have a meeting basically now but have a couple minutes. How can I help?
15:03 donsizemore
if i submit a pull request to update usage examples for say citation fieldtype, do you want me to do a block of examples, or make granular changes?
15:04 pdurbin
Examples sound nice.
15:04 pdurbin
But whatever is easier for you.
15:07 donsizemore
examples definitely, but by "block" i meant multiple endpoints (or just one per PR)
15:30 Benjamin_Peuch
Youssef_Ouahalou told me about how you found a solution to the problem, pdurbin. So it was about escaping the ampersand with quotation marks?
15:31 Benjamin_Peuch
Do you want us to open an issue on GitHub about this?
15:38 pdurbin
Benjamin_Peuch: yes, please!
15:39 pdurbin
donsizemore: and each endpoint has a different date field? Sorry, I don't think I've had enough coffee. Maybe make the first PR and I'll be able to follow more easily.
15:47 Benjamin_Peuch
Hey poikilotherm: when you have the time, could you add the following three labels -- [Feature: User Guide] [Small] [Type: Bug] -- to this issue -- https://github.com/IQSS/dataverse/issues/6563 -- please?
15:47 Benjamin_Peuch
There you go, pdurbin.
15:48 Benjamin_Peuch
I must go. I'll be back on Monday. Have a nice weekend, everybody!
15:54 donsizemore
@pdurbin no i was going to submit a PR specifically to update docs on one endpoint, but there are a bunch of endpoints on the Native API page. do you want a PR for each endpoint update, or do you want me to test a batch and provide updated usage examples in batches?
15:56 pdurbin
Oh, you're talking about endpoints other than the "switch the date used in a citation" endpoint?
15:58 pdurbin
You're basically talking about putting https://github.com/IQSS/dataverse/issues/6083 through QA? Is that right?
16:14 donsizemore
ooh ooh let me take a look at those
16:14 donsizemore
(our rising graduate student, BTW, is probably taking another job elsewhere =( )
16:43 pdurbin
bummer!
17:52 jri joined #dataverse
18:05 ilya joined #dataverse
18:08 ilya
Hi, I am adding some datasets with large files to Harvard Dataverse and I've encountered this bug: https://github.com/IQSS/dataverse/issues/6535 . I was wondering if anyone had a chance to have a look at it.
18:11 pdurbin
ilya: you're uploading from Dropbox?
18:12 ilya
Yes.
18:14 pdurbin
Hmm. Well, it sounds like you've definitely confirmed the bug, which is helpful. Please feel free to leave a comment about file sizes or whatever details you think might help us resolve it some day. It isn't in our current sprint and we haven't given it a size/effort estimate yet. I can suggest some workarounds if that would help.
18:16 ilya
I see. I am using some workarounds already - including help from Dataverse support but using Dropbox would make it a lot easier. There is a more information including sample files for testing in the ticket.
18:20 pdurbin
Oh, in https://help.hmdc.harvard.edu/Ticket/Display.html?id=285188 . I see what you mean.
18:21 ilya
Right. Well, I just wanted to check is there was an ETA on the fix. Thanks!
18:21 pdurbin
Well, it hasn't even been estimated yet. Have you seen our project board?
18:22 pdurbin
It might help give you a sense of which issues have been prioritized, etc.
18:22 ilya
No, I did not know there was one :)
18:23 pdurbin
Here's our project board: https://github.com/orgs/IQSS/projects/2 . At standup every day we talk about QA first, then code review, then "team dev".
18:25 pdurbin
I just finished reading through the RT ticket. It seems like the support and curation team has been helping you get your files into place.
18:27 ilya
I see, thanks. Yes, the support team has been very helpful.
18:28 pdurbin
There's a new pull request to do direct upload to S3 (where Harvard Dataverse stores its files). I wonder if it could be adapted to our Dropbox upload somehow.
18:32 pdurbin
This one: https://github.com/IQSS/dataverse/pull/6490
18:36 pdurbin
I just asked the developer in a comment.
18:50 ilya
Direct to S3 uploads are neat but it's probably not going to help in this case because downloads from Dropbox must pass through the backend before heading up to S3.
18:52 pdurbin
Sure, but I guess I'm hoping the backend can hand off the stream or something. I don't know. I'm grasping at straws. :) You're pretty committed to using Dropbox? We obviously should try to fix it but I'm not sure when we'll get to it.
18:57 ilya
I've uploaded smaller files using the browser, the support team was uploading some of the larger files for me and I was using cURL to upload the files via the API which has worked pretty well so far. Dropbox would make it more convenient but I think I should be able to manage without it.
18:58 ilya
Thanks for all the info!
18:58 pdurbin
Sure, have you seen the command line DVUploader?
18:59 pdurbin
This: http://guides.dataverse.org/en/4.19/user/dataset-management.html#command-line-dvuploader
19:00 ilya
Cool, it supports parallel uploads. I'll try that, thanks!
19:00 pdurbin
Sure! And thanks for reporting the Dropbox bug. It's been years since I've used Dropbox.
20:11 donsizemore
@pdurbin one more Q?
20:16 pdurbin
hit me
20:17 donsizemore
on the Native API page, older examples are a mix of `` (Dataverse orange quoted text) and code-block:: bash (grey boxen)
20:17 donsizemore
it looks like the newer Metrics stuff is all `` so ima keep the multi-line examples as code blocks but convert all the singles to `` ?
20:17 donsizemore
the quoted text may not be Dataverse orange... I see a limited range of colors
20:18 pdurbin
hmm, let me look at what i was doing
20:18 pdurbin
donsizemore: can you make them look like this? http://guides.dataverse.org/en/4.19/api/native-api.html#add-file-api
20:19 donsizemore
so, code blocks
20:19 pdurbin
yeah
20:19 donsizemore
i guessed wrong and had started converting. i'll redo them.
20:21 pdurbin
Yeah, it looks like I ultimately went with .. code-block:: bash
20:22 donsizemore
the Dataset Metrics section ^^ still uses ``
20:23 donsizemore
I can do it either way, I just see a mix and would like to know what y'all prefer. sounds like code-block
20:23 pdurbin
The problem is everyone does their own thing. And I'm sure I switch between things.
20:24 pdurbin
But yeah, I'm thinking code-block if that's ok with you.
20:24 donsizemore
i just assumed the dataset-metrics section was newer so that was the current thinking. but i can make it all code-block, i was going to update-and-test all the examples in the "Dataverses" section and submit a PR
20:26 pdurbin
Sounds fine. Sorry so confusing.
21:08 ilya joined #dataverse
21:14 ilya
I am getting {"status":"ERROR","message":"There was an error when trying to add the new file. This file size (2.8 GB ) exceeds the size limit of 2.5 GB."}. Would Dropbox uploads have the same limit?
21:25 ilya joined #dataverse
21:29 ilya joined #dataverse
21:31 donsizemore joined #dataverse
21:32 donsizemore
@pdurbin I'm fine either way (but it makes me want a style guide)
21:45 poikilotherm
pdurbin donsizemore I've been using Code-Block all over in cloud guide
22:28 pdurbin
donsizemore: yeah
22:28 pdurbin
poikilotherm: cool
22:33 donsizemore
closed my eyes and submitted a PR for the Dataverses stanza
