Time
S
Nick
Message
07:57 poikilotherm joined #dataverse
08:08 jri joined #dataverse
08:11 jri joined #dataverse
09:47 jri_ joined #dataverse
10:57 nils`` joined #dataverse
11:34 andrewSC joined #dataverse
11:38 Youssef_Ouahalou joined #dataverse
12:11 MrK joined #dataverse
12:21 Youssef_Ouahalou
@pdurbin Sorry I did not see your message, you mean like this:https://www.google.com/search?q=Markdown+support&rlz=1C1GCEU_frBE870BE870&source=lnms&tbm=isch&sa=X&ved=0ahUKEwjE-ov2oPblAhVLYlAKHWbLDSwQ_AUIEigB&biw=1920&bih=969#imgrc=eCwApMaAW5nkkM:
12:22 Youssef_Ouahalou
Is it possible to introduce it?
12:32 pdurbin
Markdown support? Sure, if someone makes a pull request. :)
13:51 donsizemore joined #dataverse
14:09 stefankasberger joined #dataverse
15:01 donsizemore
@pdurbin Jim wants to show you his Hacktoberfest t-shirt!
15:09 pdurbin
donsizemore: let's see it. Also, if your ears were burning, Danny just gave you a shout out in the community call, which just started: https://dataverse.org/community-calls
15:12 jri joined #dataverse
15:14 pdurbin
All are welcome to join in, of course. Tania is doing a demo of https://validately.com
15:24 jri_ joined #dataverse
16:20 jri joined #dataverse
17:28 jri_ joined #dataverse
18:16 GeraFp joined #dataverse
18:19 GeraFp joined #dataverse
18:45 jri joined #dataverse
19:01 GeraFp
Hi all!
19:30 pdurbin
GeraFp: hi! Did you see someone was excited about the new Micosoft login?
19:30 pdurbin
this: http://irclog.iq.harvard.edu/dataverse/2019-11-18#i_112714
19:48 GeraFp
pdurbin: I did not see :( but this is a great news :D.
19:49 pdurbin
:)
19:49 pdurbin
GeraFp: is there anything we can help you with?
19:56 GeraFp
Yes, I have a question with the SelectOneMenu implementation in Dataverse. I need yo use the ValucheChangeListener attribute with SelectOneMenu but this method is not called :(, I see your code in DatasetPage.java and Dataset.xhtml and replicate some fragments but it's same result.
19:59 pdurbin
GeraFp: hmm, are you working on a specific issue? A specific bug fix? A specific (new) feature?
19:59 GeraFp
I see the Primefaces documentation and run the samples but when i implemented in dataverse not work.
20:02 Jim96 joined #dataverse
20:04 GeraFp
Yes, in CIMMYT are work in a implementation for autocomplete fields with Controlled Vacabularies. Currently, We work in the autocomplete fields task in the dataset form, the user selects a Controlled Vocabulary in a dropdown list and the terms are load in a list for complete other fields.
20:04 Jim96
(proudly wearing his Hacktoberfest shirt and admiring his Hacktoberfest stickers) - Hey @pdurbin, do you know if anyone has set of MDC chron jobs? Trying to walk TDL through that and looking for examples to start from.
20:04 pdurbin
GeraFp: very cool. Is there a number for that issue in GitHub?
20:05 pdurbin
Jim96: I think there's a crob job around here somewhere. Next community meeting let's get a pic of you and Oliver wearing your shirts together.
20:07 GeraFp
pdurbin: Yes, it's this https://github.com/IQSS/dataverse/issues/6154
20:12 pdurbin
GeraFp: thanks!
20:12 pdurbin
Jim96: I don't think we have a cron job yet. I think that's what this issue is about: https://github.com/IQSS/dataverse.harvard.edu/issues/3
20:12 pdurbin
donsizemore: how close are you to enabling Make Data Count for Odum?
20:13 Jim96
OK - I'll just have TDL go ahead and we'll submit them.
20:13 pdurbin
Jim96: thank you!!
20:14 pdurbin
donsizemore: thanks for https://github.com/IQSS/dataverse/issues/5510#issuecomment-555691096 ... can you please edit that comment to indicate which version you're running? It will make the line numbers more helpful.
20:14 pdurbin
GeraFp: I'm not very good with JSF but I'm calling for reinforcements.
20:16 Jim96
Sure - FWIW, I think my next MDC step is to do something to add pre-MDC counts into the MDC display somehow (TBD ) and to make sure files don't show more counts than the dataset etc. If anyone has ideas, let me know.
20:17 GeraFp
pdurbin: Oh!, thanks. I have some weeks with this issue and I haven't more ideas.
20:18 pdurbin
Jim96: well, maybe not ideas but there's a script somewhere a contributor wrote to try to populate the new MDC table with data from apache logs or something. Let me go look for the conversation at least. I never did get the script.
20:20 pdurbin
Jim96: I'm thinking of this conversation with Jerry19: http://irclog.iq.harvard.edu/dataverse/2019-09-09#i_105111 "Are you saying you wrote a script to take old Apache logs and supplement them with enough information from Dataverse to put them into the format that Counter Processor requires?" Are you thinking along those lines?
20:20 pdurbin
GeraFp: weeks! Oh no! Next time please reach out sooner. :)
20:24 donsizemore
@pdurbin I'm enabling Make Data Count for Odum?
20:25 Jim96
I was thinking of just using the internal db entries and, assuming those aren't enough info to populate MDC entries, just adjust the display to show, for example, "10K downloads, 4K with MDC info".
20:26 Jim96
I hadn't considered mining the apache logs to search for the missing parts, which is a nice idea - have to see if QDR/TDL really have those far enough back.
20:34 pdurbin
donsizemore: why not! :)
20:35 donsizemore
i just do what our archivists say
20:35 pdurbin
Jim96: oh, but wouldn't that be a misreprentation of those numbers, potentially? You'd want to be careful about the wording.
20:36 pdurbin
When we first started working on the UI I sort of assumed we'd watch to expose the old number if the user wanted to see it.
20:36 Jim96
Definitely need to be careful, but what concerns you?
20:36 pdurbin
Well, the whole point of MDC is that all kinds of filtering is happening. The old number is always going to be higher for that reason.
20:37 pdurbin
I guess I'm just trying to say that if we show a number it should be clear if it's from the old way or the MDC way.
20:38 pdurbin
GeraFp: I think the best thing for you to do would be to leave a comment on the issue you're working on. You can just say "JSF trouble" and we'll know what you're talking about. :)
20:39 pdurbin
Jim96: obviously, I was a little confused by what you meant by "pre-MDC counts". You mean the old, classic way of counting downloads. Which still happens in the background even when MDC is enabled. :)
20:40 Jim96
Right. It definitely can't collapse to one number, but starting at 0 for a Dataset that's been out for a while isn't appealing either.
20:40 pdurbin
Right! Dude, where'z my downloads?
20:40 pdurbin
Show me the old number!
20:40 pdurbin
My funding depends on it!
20:40 Jim96
Yeah - that's what I meant - I'd have to stop counting the unfiltered ones as of the MDC logging starting (do that now with the display off!)
20:41 pdurbin
Jim96: here's a thought. Would you want to expose the old number via API ?
20:41 Jim96
Ah - the rock start lifestyle of having the most downloads...
20:42 pdurbin
It's harder to game the system with MDC in place. :)
20:43 Jim96
I think something has to be in the display but no reason it could't be API as well. For TDL an API may help since they have some external metrics going.
20:43 Jim96
What - they're stopping botnets?
20:43 GeraFp
pdurbin: Yes, It's a great idea. I comment my question on the issue in a few moments. Thanks for your help :D
20:43 pdurbin
Well, MDC tries at least, to distinguish between machine and person downloads.
20:44 pdurbin
scripts vs clicks
20:44 pdurbin
I forget what terminology they use.
20:44 pdurbin
MDC also has the concept of not counting the same download with a given period of time.
20:44 pdurbin
I'm sure you could still game it, but they're trying. :)
20:45 Jim96
right - they remove robots and then (I think) watch the client used. But they report both people and script driven ones (since workflow scripts are legit).
20:46 pdurbin
yeah
20:46 Jim96
I agree - better than real signals (and any fraud) being drowned in robot hits etc.
20:46 pdurbin
I have a suspicion that everyone is going to like the old larger download numbers better. :)
20:47 pdurbin
GeraFp: thanks! Sorry I can't help more. We also have a dataverse-dev mailing list you can use to reach developers: https://groups.google.com/forum/#!forum/dataverse-dev
20:47 Jim96
Has anyone thought about exploding the one MDC number to let you see - by time, by country, manual versus automatic, etc.? I know the db has all that info, but haven't hear of anything on the display side yet.
20:48 pdurbin
Jim96: yes! Stuff I drew on my whiteboard anyway. Lots of rich data in there. But I don't think anyone has it top of mind. Have you seen the nice viz DataONE does with MDC metrics?
20:49 Jim96
A while back - and I saw that you had some discussion about possibly adapting that.
20:49 pdurbin
Yeah. Mostly I was just trying to show what's possible, based on the data.
20:50 pdurbin
Based on what MDC compels you to collect. By country, by month, etc.
20:51 pdurbin
Actually, I feel like there some API endpoints missing for getting the data out.
20:51 pdurbin
I remember thinking, "This is great if you know what's in your database already."
20:52 pdurbin
I'm looking at http://guides.dataverse.org/en/4.18/api/native-api.html#dataset-metrics-api
20:53 pdurbin
Oh, I know what I'm thinking, I think. I think I know what I'm thinking.
20:54 pdurbin
I don't think we do a great job of exposing these MDC metrics for the entire installation. I think we do ok showing MDC metrics for a particular dataset.
20:54 pdurbin
But do we do a good job of answering questions like "How many citations across this installation of Dataverse?" I don't think so.
20:56 Jim96
Yeah - that's an important number/thing to explore.
20:59 pdurbin
Maybe it could be added to the Metrics API . No mention of citations at http://guides.dataverse.org/en/4.18/api/metrics.html
21:04 pdurbin
"How many total citations?" "How many datasets have been cited?" "Which datasets have been cited?" etc.
21:04 pdurbin
GeraFp: I approved your request to join dataverse-dev. Thanks.
21:07 GeraFp
Thanks to you pdurbin!
21:22 pdurbin
GeraFp: sure. At a high level, I have your problem a lot. You're saying you expect a click to trigger a method in Java and it isn't working?
21:39 GeraFp
pdurbin: Yes, the Java Method in the attribute valueChangeListener on selectOneMenu is not called.
21:39 pdurbin
:(
21:39 pdurbin
T-T
21:40 pdurbin
GeraFp: have you pushed a branch to GitHub?
21:57 GeraFp
Yes, it's pushed in a development branch from CIMMYT in GitHub. It's here https://github.com/CIMMYT/Research-Data-Repository/commit/ceee339ffd10115289af373071d04c9796fe84ac
22:01 GeraFp
The files with the implementation are the vocabulary-test.xhtml and the manage bean is VocabularyTestPage.java. The files are separated for testing and develop.
