Time
S
Nick
Message
06:33 Virgile joined #dataverse
08:22 Virgile joined #dataverse
11:50 donsizemore joined #dataverse
13:53 pameyer joined #dataverse
13:56 Thierry joined #dataverse
13:57 Thierry
Hia ll
13:57 pameyer
hi Thierry
14:00 Thierry joined #dataverse
14:00 Thierry
May I ask a question about DOI management in Dataverse here, or is it not he place to do so?
14:02 pameyer
there might not be someone here who could give you a definative answer, but feel free to ask
14:05 Thierry
All right, thanks anyway! :) I have a dataverse install with some datasets on it, with the default (10.572/FK2/ stuff). Since recently, I can reserve real DOIs. I'd like to delete current "fake" DOIs from some datasets so that I can reserve "real" ones. BUT, it seems that the API to delete those 10.572/FK2 won't free those datasets from their DOIS because (error 404) indeed they do not exist within datacite
14:06 Thierry
So My question would be, is there somewhere in the database itself where I can act to delete those DOIS? I Have Dataverse 5.1.1
14:09 pameyer
there's definately a database table storing DOIs - but I don't think deleting them there would be the best way to solve your problem
14:10 pameyer
I'm pretty sure the database structure for datasets requires the DOI (or other PID) to be present; so I don't think there's way to "delete a dataset's DOI without deleting the dataset"
14:10 Thierry
Yeah I'm afraid this would be a litllte more complicated than that :p
14:10 pameyer
:)
14:11 pameyer
if I'm understanding you right; you've got datasets with the FAKE DOI provider / test DOI prefix that you want to switch to dataset / real DOI prefix
14:12 pameyer
two ways I could think of to do that - use the API to export the dataset as dataverse json, fiddle the json appropriately, and use the API to re-create them (once the dataverse install has been configured with the live doi info)
14:13 pameyer
second would be to update the database, manually overwrite the fake DOIs with ones under control of the new account, and then call one of the re-register APIs
14:13 Thierry
Ahhh thanks for the tip, I hadn't tought of this way of dong this. And yes, my problem is just exactly what you understood
14:14 pameyer
I'd also recommend waiting to see if pdurbin or donsizemore have ideas
14:15 Thierry
Yes, I will. But yet thanks for that, exporting to json/ reimporting will be my next move
14:15 pameyer
good luck :)
14:15 pameyer
... and on general principles, I'd recommend checking that you've got a current backup first
14:16 Thierry
I did :) I have Dataverse installed on a container, that is backed-up
14:17 pameyer
great :)
14:22 pdurbin joined #dataverse
14:29 pdurbin
Thierry: you might be able to update the dvobject table. However, I'd recommend emailing support
14:32 Thierry
I will, thanks pdurbin
14:35 pdurbin
Sure. The thing I wonder about is getting the metadata to DataCite. I thought we had an API to re-send it but I can't find it. At least it will be updated when you edit the dataset and republish.
14:36 pameyer
I thought there was a re-send / re-register API too; but it's not one I've used
14:37 Thierry
I'm wondering if I set "authority" to "" and identifier to "", will I be able to reserve a DOI from the GUI ? I still am a lil afraid to test event on one dataset cause it could make everithing blow away ^^
14:38 Thierry
in the dvobject table that is
14:39 pameyer
if you're thinking of the dataverse UI, I'd expect that to go badly (but it's not something I've tried)
14:41 pdurbin
Probably the saftest thing to do is to create a new dataset with the old information, using the export JSON technique that pameyer mentioned above.
14:42 Thierry
Yes I see. I'll try to het json file as suggests pameyer, then update database and resubmit the dataset. I already successfully configured the DOIs for new datasets, that correctly register using my real authentification and the correct shoulder
14:42 pdurbin
Then you would delete (destroy) the old dataset that doesn't have a real DOI.
14:43 Thierry
I'll try that, and let you know if I'm successfull in case anyone else needs to do that.
14:46 pdurbin
Thierry: cool, good luck.
16:02 poikilotherm
pdurbin you know what drives me mad? At least 3 different styles to push in settings to the database...
16:03 pdurbin
Well, only one style for the "setting" table, right?
16:03 poikilotherm
Oh it's a mix
16:04 poikilotherm
There is an API call for the lang stuff, which is also saved as distinct fields to the database
16:04 poikilotherm
It's setting/name/lang/langname
16:04 poikilotherm
That's fine
16:04 poikilotherm
Then you can push in JSON as a compound value for a single name
16:05 poikilotherm
And then there are compound values made up by adding :name:compoundname with a single value
16:05 pdurbin
huh
16:07 poikilotherm
Style 3 used by https://guides.dataverse.org/en/latest/installation/config.html#id184
16:07 pdurbin
:TabularIngestSizeLimit:Rdata
16:07 poikilotherm
Style 2 e.g. by https://guides.dataverse.org/en/latest/installation/config.html#maxfileuploadsizeinbytes
16:07 pdurbin
yeah
16:08 pdurbin
yeah
16:08 poikilotherm
And then there are comma separated lists
16:08 pdurbin
Ok, you're right. These are all to the setting table.
16:09 poikilotherm
Yeah
16:09 poikilotherm
That things is way to flexible
16:10 poikilotherm
Some absurd ways to retrieve config in the code can be spotted... I know, it's historically grown. But man, a aligned approach would be great... Not making 1000 different styles and ways to do things....
16:10 pdurbin
Clever developers use it for evil.
16:11 poikilotherm
It's even a huge thing to do a data migration for this if this should be fixed to an aligned style...
16:12 pdurbin
Yeah. Herding cats.
16:15 poikilotherm
https://scontent-ham3-1.cdnins[…]a1&_nc_sid=83d603
16:16 * poikilotherm
uploaded an image: (5952KiB) < https://matrix.org/_matrix/media/r0/download/bertuch.name/krvgLzCPzdyAHzuHRvFQDhWc/ezgif-2-2aeee682021a.gif >
16:16 pdurbin
heh, a classic
16:18 poikilotherm
OMG I just found another dirty hack
16:18 pdurbin
#4?
16:18 poikilotherm
https://guides.dataverse.org/en/latest/installation/config.html#archiversettings lets you use an arbitrary setting name.
16:19 pameyer
isn't "be liberal in what you accept" a recommended unix-ish design practice?
16:19 poikilotherm
And that is simply using the name you provide to lookup sth in the database
16:20 pdurbin
I forgot about that one.
16:21 poikilotherm
https://homepage.cs.uri.edu/~thenry/resources/unix_art/ch01s06.html
16:21 poikilotherm
pameyer: rule 12:
16:21 poikilotherm
“Be liberal in what you accept, and conservative in what you send”. Postel was speaking of network service programs, but the underlying idea is more general. Well-designed programs cooperate with other programs by making as much sense as they can from ill-formed inputs; they either fail noisily or pass strictly clean and correct data to the next program in the chain.
16:22 poikilotherm
But there also is a warning
16:22 poikilotherm
The original HTML documents recommended “be generous in what you accept”, and it has bedeviled us ever since because each browser accepts a different superset of the specifications. It is the specifications that should be generous, not their interpretation.
16:23 poikilotherm
Yet - there is no formal specification. Or it's "do WTF you want if you need a setting"
16:23 pameyer
so are dataverse settings the specification, or the interpretation? ;)
16:25 poikilotherm
But this is the rule of repair. IMHO it's violating Rule of Robutstness, Rule of Extensibility, Rule of Simplicity, Rule of Composition, Rule of Modularity, Rule of Least Surprise and maybe also Rule of Economy (now that one want's to align)
16:27 poikilotherm
Anyway folks, I'm outta here for today. Too much frustration, work to do at construction site. Read y'all...
17:10 donsizemore joined #dataverse
17:30 pdurbin
Construction is great for getting frustration out. Second only to demolition.
17:48 dataverse-user joined #dataverse
17:48 dataverse-user
hi chat
17:48 dataverse-user
i had a question about aws deployment
17:48 dataverse-user
anyone in this chat familiar with this?
17:58 dataverse-user left #dataverse
18:11 pdurbin
Whoops. Missed 'em.
18:25 pameyer
I wonder if the proliferation of chat apps has changed folks expectations of latency
18:42 pdurbin
IRC is for the patient.
18:50 pameyer
it's virtue we should probably be cultivating these days
18:52 pdurbin
Waiting for the vaccine is for the patient.
18:53 pameyer
definately
19:48 Alexandre joined #dataverse
19:48 Alexandre
Hi guys. How are you ?
19:49 Guest86960
I'm deploying dataverse-kubernetes from https://github.com/gdcc/dataverse-kubernetes
19:49 pdurbin
Guest86960: hi! Good. Yourself?
19:49 Guest86960
good too, pdurbin
19:50 Guest86960
so, I'm deploying dataverse-kubernetes from https://github.com/gdcc/dataverse-kubernetes ... We are using Handle.net instead of DOI
19:51 Guest86960
in init_2_conf_payara.sh file i'm tryng to "run" the asadmin commands to configure Handle, however without success...
19:51 Guest86960
I've tried
19:51 Guest86960
echo "create-jvm-options '-Ddataverse.handlenet.admcredfile=/hs/svr_1/admpriv.bin'" >> ${DV_POSTBOOT}
19:51 Guest86960
echo "create-jvm-options '-Ddataverse.handlenet.index=300'" >> ${DV_POSTBOOT}
19:52 Guest86960
can anyone help me please ?
19:53 pdurbin
Well, if you are manually running asadmin commands, I would suggest looking at this: https://guides.dataverse.org/en/5.3/installation/config.html#persistent-identifiers-and-publishing-datasets
19:54 pdurbin
And actually, configuring Handle takes a mix of asadmin commands and database settings.
19:54 pdurbin
(Documented in that page above.)
19:55 pdurbin
I have to admit that I've never configured Handle myself.
19:56 pdurbin
And poikilotherm who created that kubernetes code, using DOIs instead of Handles.
19:58 pdurbin
If it helps, it looks like there was a question about Handle in https://github.com/gdcc/dataverse-kubernetes/issues/217
19:59 Guest86960
Thanks, pdurbin I'll take a look! :-)
20:00 pdurbin
sure :)
20:12 poikilotherm
Heh. Looks like it's about time to get my hot pot stirred and serve the dish after it has cooked a few months, as pameyer said...
20:18 pdurbin
Some like it in the pot nine days old.
20:28 pameyer
yeah; I'll a little behind on things
20:39 poikilotherm
pdurbin congrats on getting the deactivate user PR merged
20:39 poikilotherm
That also cooked slowly so it'll be great
20:43 pdurbin
Ha! Well, I hope so. At least this means that we'll soon release 5.4.
20:45 pameyer
crock pot coding ;)
20:58 pdurbin left #dataverse
