Time
S
Nick
Message
02:25 andrewSC joined #dataverse
02:28 jri joined #dataverse
02:37 andrewSC joined #dataverse
04:59 andrewSC joined #dataverse
05:28 jri joined #dataverse
06:48 poikilotherm joined #dataverse
07:18 jri joined #dataverse
08:13 stefankasberger joined #dataverse
08:31 stefankasberger joined #dataverse
10:36 stefankasberger joined #dataverse
11:57 poikilotherm
Morning pdurbin :-)
12:04 jri joined #dataverse
12:08 pdurbin
morning!
12:40 poikilotherm
Hope I did not bother you guys in https://github.com/IQSS/dataverse/issues/5862#issuecomment-494131600
12:58 pdurbin_m joined #dataverse
12:59 pdurbin_m
poikilotherm: not at all! Please feel free to spin off that idea into a dedicated issue. I'm just suggesting a quick fix to get new installations working again. :)
13:03 poikilotherm
Ok cool - wasn't sure if I had been to brisk
13:18 jri joined #dataverse
14:01 pmauduit_ joined #dataverse
14:03 pdurbin
nope, you're good
14:03 pdurbin
what you're suggesting relates a bit to https://github.com/IQSS/dataverse/pull/5858 poikilotherm
14:05 poikilotherm
Indeed. But just a tiny bit ;-)
14:05 pdurbin
poikilotherm: oh and I haven't made a pull request for the quick fix yet
14:06 poikilotherm
Huh?
14:06 poikilotherm
I thought that was part of what kcondon merged?
14:07 pdurbin
It's confusing.
14:07 pdurbin
We merged a flyway script that adds a column.
14:08 pdurbin
now we need a new pull request to fix that line to add "IF NOT EXISTS" I think
14:08 poikilotherm
Meh. If it is avoidable, we shouldn't do that.......
14:09 pdurbin
it would be a quick fix :)
14:09 pdurbin
we want to keep "develop" release ready
14:09 poikilotherm
Kinda hotfix and refactor later
14:09 pdurbin
yeah
14:09 pdurbin
small chunks
14:10 pdurbin
keep develop release ready
14:10 poikilotherm
Yeah
14:10 pdurbin
refactor later
14:10 pdurbin
that's my suggestion anyway
14:10 poikilotherm
I digged through the internet looking for options how to validate the integrity of the schema...
14:10 poikilotherm
dug
14:11 poikilotherm
It's much easier with Hibernate :-/
14:11 pdurbin
hmm
14:11 pdurbin
not part of jpa?
14:12 poikilotherm
It is a competitor to EclipseLink
14:12 donsizemore joined #dataverse
14:12 pdurbin
sure, an implementation
14:12 poikilotherm
Yeah.
14:12 pdurbin
but shouldn't checking integrity be part of the spec, part of JPA?
14:12 poikilotherm
Both EclipseLink and Hibernet implement JPA
14:12 poikilotherm
So in theory it should be exchangeable
14:13 pdurbin
donsizemore: hi! We're talking about the deployment bug you found. Would you like to make a pull request? I found some "IF NOT EXISTS" syntax we could add.
14:13 poikilotherm
Seems like it is a feature which is more sophisticated within HIbernate :-/
14:13 pdurbin
maybe JPA doesn't specify or require that feature. maybe it should
14:14 donsizemore
@pdurbin i'm happy to. sorry, i forgot i had a dentist appt. this morning and i'm just now getting in
14:14 pdurbin
donsizemore: thanks!
14:23 man joined #dataverse
14:23 man
hi
14:25 pdurbin
hi
14:33 pdurbin
andrewSC bjonnh candy` donsizemore icarito[m] jri larsks pmauduit_ poikilotherm rigelk sivoais: 90 minutes until the community call where Michael will be sharing some ideas for using ActivityPub. His writeup is linked from https://dataverse.org/community-calls
14:33 poikilotherm
Meh. Would love to join, but need to look after my son...
14:34 rigelk
thanks pdurbin - I will be there
14:35 pdurbin
my daughter's class is a block away at a Harvard museum. I'm going to join them for lunch in half an hour but I'll be back in time for the community call
15:00 amahdy joined #dataverse
15:04 amahdy
Hi Everyone, I am a newbie and I am trying to upgrade from 4.7 to 4.10.1 and I am stuck after upgrading from v4.8.5 to v4.8.6. there is multiple errors but I got to know that there is an issue when I tried to create or open a dataset i get a 500 internal server error.
15:04 amahdy
any help will be appreciated.
15:06 donsizemore
@amahdy hello =) are you sure you applied the SQL script? (and what are the first errors you encounter?)
15:13 stefankasberger joined #dataverse
15:14 amahdy
@donsizemore Thanks, I found that I applied the wrong SQL script. I will give it a try.
15:29 donsizemore joined #dataverse
15:30 donsizemore
@amahdy i always shut myself in a room to upgrade dataverse, and for production upgrades i get somebody to babysit the checklist with me. it's so easy to skip a step or do something out of order.
15:33 amahdy
@donsizemore I am still doing it on a clone from the production and I am preparing a step by step doc for me for the production upgrade.
15:34 rigelk
pdurbin: trying to join the gotomeeting, I realise none of my browsers is accepted by the web page (trying with the latest version of Chrome and FF). Sounds like I won't be able to attend the call.
15:59 pdurbin
community call starting soon: https://dataverse.org/community-calls
16:00 pdurbin
rigelk: hmm, and the only phone number we have listed is for the US.
16:00 pdurbin
rigelk: I'm asking. Hold on.
16:02 pdurbin
amahdy: hi! You're welcome to join our community call if you'd like.
16:03 pdurbin
rigelk: we're going to start in 2-3 minutes. We used to have international phone numbers for calling it but this changed when we switched to gotomeeting.
16:03 pdurbin
calling in*
16:04 pdurbin
rigelk: what if you try an app? I just found this: https://www.gotomeeting.com/features/mac-windows-linux
16:06 pdurbin
rigelk: also, you can add any questions to the notes doc: https://docs.google.com/document/d/1n_7aqRKOx9jWLnC3L-k_MDRyxs2DIRf2KFtXO-nd688/edit?usp=sharing
16:15 rigelk
okay, it doesn't seem like I can hear any audio on the app. I'll stay on the document and here.
16:15 rigelk
sorry I can't figure out how to join the call :/
16:16 pdurbin
rigelk: sorry, I'm taking notes if that helps.
16:16 pdurbin
in the doc above
16:16 rigelk
I'm on it. Thanks!
16:29 pdurbin
sure
16:30 pdurbin
rigelk: I just gave you a shout out as one of the PeerTube developers :)
16:30 rigelk
pdurbin: thanks. And again I'm sorry I could not attend :/
16:38 donsizemore
@pdurbin knock knock?
16:46 pdurbin
rigelk: not your fault. Thanks for participating!
16:46 pdurbin
donsizemore: talk to me
16:46 donsizemore
@pdurbin may i run a dumb question by you?
16:47 donsizemore
(as if i have any other kind)
16:51 pdurbin
I love dumb questions.
16:51 pdurbin
You should have seen my daughter's math homework last night. The older daughter, not the one at the museum.
16:52 donsizemore
@pdurbin higgs boson?
16:52 pdurbin
oh and is amahdy all set? Sorry to interrupt
16:52 donsizemore
i think so. he missed the SQL script in 4.8.6. (i've never done that, no, never)
16:52 pdurbin
heh, ok
16:52 pdurbin
please go on
16:53 donsizemore
so, i wrote a little script to query a copy of our postgres db, pull out storageidentifier and checksumvalue, then call irods' iquest to compare MD5s against the copy-of-a-copy on our off-site irods instance
16:53 pdurbin
I don't know what iquest is but ok.
16:54 donsizemore
right now our production DB has nearly 32,000 datafile entries with no checksum
16:54 donsizemore
there's a datafile integrity endpoint which will fix missing originalfilesizes, and a dataset integrity endpoint to recalculate the UNF
16:54 donsizemore
i can't find a way to trigger dataverse to verify checksums and/or compute missing checksums once the entry is already in the DB
16:55 donsizemore
didn't know if you remembered this off the top of your head, or whether i should open an issue, or whether i should write a script (tho i really don't want me modifying our production DB )
16:56 pdurbin
I'm looking at the code. There's stuff like fixmissingunf and fixmissingoriginaltypes and fixmissingoriginalsizes. None of these are what you want?
16:57 donsizemore
what i want is populateEmptyChecksum (and archivists would want verifyChecksum)
16:58 pdurbin
those would be nice to have. I actually forgot about some of those "fix" endpoints until Leonid reminded me yesterday. maybe there's more "fix" stuff in here I've forgotten about :)
16:59 pdurbin
donsizemore: at the very least I think you should open an issue to better document what you need. Or creating the checksum api endpoints you want if they don't exist.
16:59 donsizemore
i would love endpoints. will open an issue. though i'll ask jon's permission before i tell the world we have no checksums ;)
17:00 pdurbin
use a fake a name
17:00 pdurbin
create a new github account
17:00 pdurbin
I won't tell
17:00 pdurbin
"asking for a friend"
17:03 pdurbin
time for our staff recognition cake thing, then tech hours
17:33 Richard_Valdivia joined #dataverse
17:37 amahdy
@pdurbin yes thanks my issue was resolved.
17:46 donsizemore joined #dataverse
18:04 pdurbin
great!
18:05 pdurbin
donsizemore: I was wrong. I still have another hour until tech hours. Let me look a little more for those endpoints.
18:16 donsizemore
i just updated the issue. akio found one in the code, but it does them all brute-force
18:18 donsizemore
also i tested the if not exists PR on a branch in EC2, there were no unexpected SEVEREs in server.log after deploy
18:35 pdurbin
right, this issue: https://github.com/IQSS/dataverse/issues/5867 ... I was just looking at it. brute force? what's the endpoint called? I'm a little all over the place right now
18:37 donsizemore
api/admin/updateHashValues/
18:40 pdurbin
wow, sure enough
18:42 donsizemore
@pdurbin also, i just tested the documentation's create-dataverse JSON against 4.14 and it succeeded
18:43 donsizemore
@pdurbin the error in #5865 makes me suspect invalid JSON format
18:48 pdurbin
donsizemore: do you want to make a pull request to switch to the non-@ version? If you know what I mean. And a full curl command.
18:49 pdurbin
switch to --upload-file I mean
18:49 pdurbin
it's way more clear and you can use tab completion
18:49 donsizemore
that's what I did. if you think that will fix Dataverse creation I'm happy to make a PR
18:52 pdurbin
yes please!
18:58 donsizemore
ugh, CLI git makes me tired.
19:08 pdurbin
it does that to everybody
19:14 pdurbin
donsizemore: I just put a little review in but let me know if I should hack on your create dataverse doc pull request.
19:35 poikilotherm joined #dataverse
19:57 poikilotherm
*ping*
19:57 poikilotherm
Anyone still around? pdurbin?
19:57 poikilotherm
I just opened https://github.com/IQSS/dataverse/issues/5871 open for discussion
19:59 poikilotherm
Go stampede and hide from the migration monster...
20:10 poikilotherm
Ah, I see - tech hours
20:10 poikilotherm
@donsizmore @pdurbin: I use git on cmdline any time I need to do sth with it... ;-)
20:11 poikilotherm
(exception proves the rule?)
20:25 donsizemore joined #dataverse
20:25 donsizemore
@poikilotherm I use it exclusively on the command-line... which is why some tasks make me tired
20:26 donsizemore
@pdurbin thank you for notes on #5870... will update
20:28 pdurbin
donsizemore: thanks!
20:29 pdurbin
poikilotherm: I'm here for 2 more minutes before I need to get out of here to get my kid to cello. The new issue looks great! Thanks!
20:29 pdurbin
I talked to kcondon for a while about the pull request donsizemore made. The if not exists or whatever. And I'm sure sekmiller will be interested in that new issue too.
