Time
S
Nick
Message
01:24 jri joined #dataverse
01:39 jri joined #dataverse
02:05 jri joined #dataverse
06:47 juancorr joined #dataverse
07:06 jri_ joined #dataverse
11:17 yoh joined #dataverse
11:31 donsizemore joined #dataverse
12:26 stefankasberger joined #dataverse
12:26 stefankasberger
hi!
12:26 stefankasberger
@poikilotherm: do you have time in the next hour for a call?
12:27 poikilotherm
Hi stefankasberger
12:27 poikilotherm
Sure!
12:28 juancorr joined #dataverse
12:30 stefankasberger
now?
12:30 poikilotherm
https://webconf.fz-juelich.de/b/ber-dtz-xbg
12:31 poikilotherm
Anyone else interested in pyDataverse & dvcli is welcome to join
13:05 poikilotherm
Oh no... You're gone?
13:05 poikilotherm
Baby hit the wire?
13:20 stefankasberger joined #dataverse
14:26 poikilotherm
stefankasberger still around?
14:26 poikilotherm
Or did your little toddler run with your cable? ;-)
14:34 pdurbin joined #dataverse
14:35 pdurbin
Did you have the meeting?
14:35 pdurbin
I don't think that baby can walk yet. :)
14:40 poikilotherm
pdurbin: yeah we had
14:41 poikilotherm
You never know... Maybe it was the noise that made jitter on the wire :-D
14:41 pdurbin
gremlins
15:55 stefankasberger
:)
16:29 pameyer joined #dataverse
16:38 poikilotherm
pdurbin please help me. Is it now clear enough what my issue is about and what I want to do? https://github.com/IQSS/dataverse/issues/7077#issuecomment-664499714
16:39 poikilotherm
Maybe I got jggautier wrong due to my language restrictions
16:54 pdurbin
poikilotherm: this issue looks like a deep topic.
16:54 poikilotherm
But why?
16:54 poikilotherm
It's not that complicated, right?
16:54 poikilotherm
And I provided a tl;dr
16:56 pdurbin
Maybe you and Julian could have a quick call. You're both on Slack.
16:58 pameyer
@poikilotherm you're using DataCite schema and "not that complicated" in the same context? ;)
16:58 poikilotherm
pameyer: OK OK, but it's not Rocket Science
16:59 pdurbin
is it data science?
16:59 poikilotherm
Flying to the moon involves much trickier stuff :-D
16:59 poikilotherm
Yes pdurbin it is :-D
16:59 poikilotherm
We have bibliometricians here in our library
16:59 poikilotherm
They definitely enjoy this stuff
17:06 pameyer
not my area of expertise, but it seems like starting with an API to set the free text for resourceTypeGeneral on a draft version, and using it on publish, would be a place to start
17:06 poikilotherm
Why a free text for resourceTypeGeneral?
17:07 poikilotherm
That would lead to invalid XML , because the attribute has a CV, the value of the XML node is free text
17:07 pameyer
because that side-steps the controlled vocabulary stuff - lower complexity to start with
17:07 pameyer
... and because they don't have "X-ray diffraction" in their list of controlled values
17:09 pameyer
ex - doi:10.15785/SBGRID/1 was produced w\ valid datacite xml
17:14 poikilotherm
http https://ez.datacite.org/id/doi:10.15785/SBGRID/1
17:14 poikilotherm
<resourceType resourceTypeGeneral="Dataset">X-Ray Diffraction</resourceType>
17:14 poikilotherm
Again: the free text is the value of the node, not the value of the attribute ;-)
17:15 poikilotherm
For now I would start with the CV and leave the implementation of the free text as an exercise for later.
17:15 pameyer
ok - maybe I missunderstood what you're trying to do
17:15 pameyer
didn't read the whole issue
17:16 pameyer
had interpreted the intent as "get more useful info into doi metadata" (vs "implement the spec")
17:16 poikilotherm
To quote my issue: "tl;dr: Dataverse should offer a UI component to select the general dataset type with a CV based on DataCite/Dublin Core. The selected type should be used for metadata registration at DataCite."
17:17 pameyer
yup - what user problem is that directed towards?
17:18 poikilotherm
My particular use case is about registring research software as software, not dataset
17:18 poikilotherm
But there will be more use cases in the future
17:19 pameyer
hadn't known that software was in the spec
17:19 poikilotherm
For example your dataset you send
17:19 poikilotherm
Shouldn't that be of type image or similar?
17:19 poikilotherm
It looks like a 3D object
17:20 pameyer
that one's a series of 2d images covering rotational series of 3d reciprocal space
17:20 pameyer
but if I call it "image", somebody's going to try to open it with gimp/photoshop/previewer and wonder why adsc/mar format crashes it
17:21 pameyer
so calling it "image" means that somebody using the metadata might try to put it into their computer vision system
17:21 pameyer
short version - arguably could fit the "image" slot, but would fit poorly and (in my judgement) make things more annoying than necessary
17:23 poikilotherm
Sure. Whatever fits your use case. I not familiar enough with your use cases :-)
17:24 poikilotherm
It's likely this will be different for us.
17:24 poikilotherm
But like I wrote in the issue: this should be optional and default to plain simple "DataSet".
17:24 poikilotherm
Anywhere this makes sense to use, people can start using it in their dataverses
17:25 pameyer
yeah - use cases are the complicated part sometimes ;)
18:11 donsizemore joined #dataverse
19:27 pameyer
pdurbin: btw, thanks for the feedback on glassfish-setup / as-setup
19:30 pdurbin
pameyer: sure! Hope it still works for you.
19:30 pdurbin
I think we tried to keep it backward compatible.
19:31 pameyer
I _think_ I need to specify the db name - so backward compatable enough
19:32 pameyer
haven't tested it yet - we had a server move fri, and I only noticed when I went to setup a dev environment on a different system and it didn't work
19:32 pdurbin
:(
19:33 pdurbin
linux, mac or windows?
19:34 pameyer
linux - wouldn't matter too much with the docker deployment though
19:35 pdurbin
yeah
19:36 pdurbin
I think Mike is still on Glassfish 4. It's a pain to get set up again.
19:37 pdurbin
poikilotherm: look at what you've done: https://github.com/IQSS/dataverse/pull/7086#issuecomment-664592010
22:01 jri joined #dataverse
22:42 poikilotherm
pdurbin looks like I triggered some cases of code not doing proper error handling to me. when things go poop, maybe it should be fixed... ;-)
22:53 poikilotherm
I left a comment in the new issue
22:54 poikilotherm
Maybe we can avoid more pooping by adding a handler for the base class. It should match, but I need to verify.
