Time
S
Nick
Message
03:41 bjonnh
pdurbin: I'll not be able to look at that before tuesday
11:16 mdehollander joined #dataverse
11:45 pdurbin
bjonnh: no worries, thanks
11:46 pdurbin
mdehollander: thanks for opening https://help.hmdc.harvard.edu/Ticket/Display.html?id=253177 . Any news? If you switch to a tagged release do you still see slowness when uploading a PDF? The blue bar: https://i.imgur.com/9ohZLjV.png
12:14 mdehollander joined #dataverse
12:15 mdehollander
@pdurbin, I send a response today by mail. I can't access page you send, I don't have a login. After moving to a tagged release I see a general problem that I can't upload files bigger than 1 Mb. Should I make a github issue as well?
12:23 mdehollander
I just noticed that it is an issue on our side I think. I installed it on our test server and to be able to reach it through the firewall we used a proxy redirect using nginx. If try the vm on my laptop without nginx the upload works.
12:30 mdehollander
when I increase the client_max_body_size in the nginx config all uploads are working again :) Great!
13:11 rebecabarros joined #dataverse
13:21 pdurbin
mdehollander: oh! So you fixed it? :)
13:22 rebecabarros
Hey pdurbin, me again haha. Just a quick question. I noticed yesterday that every time I've clicked to explore a file with TwoRavens the count of Downloads of the dataset is increased. This is the expected behavior? I thought that the count it was only for proper download of the dataset as a whole.
13:23 pdurbin
rebecabarros: that's a good question. Did you happen to see if anyone else has reported this in GitHub issues? I don't know.
13:23 mdehollander
@Pdurbin, yes it is fixed
13:24 pdurbin
mdehollander: hooray! Good job. I use Apache rather than nginx. Maybe you could contribute something to our Installation Guide.
13:25 pdurbin
rebecabarros: I didn't find anything about TwoRavens and download count. The only one I found was this one: Download: Do not increment file download count if file is not downloaded, throws error. · Issue #3114 · IQSS/dataverse - https://github.com/IQSS/dataverse/issues/3114
13:27 pdurbin
rebecabarros: it might be best to ask at https://groups.google.com/forum/#!forum/dataverse-community because I don't remember if that was a design decision or not. You could even ask at https://groups.google.com/forum/#!forum/dataverse-dev
13:30 rebecabarros
pdurbin: ok, I will do that. Thanks
13:31 pdurbin
It sounds like you don't think it should count.
13:32 pdurbin
I haven't really looked into the download count logic. I thought I heard that it depends on if the dataset is a draft or published. I'm not sure.
13:38 rebecabarros
https://github.com/IQSS/dataverse/issues/3572 - This could be also related.
13:39 pdurbin
yeah, and https://github.com/IQSS/dataverse/issues/2911 and https://github.com/IQSS/dataverse/issues/3331
13:40 pdurbin
We struggle a bit with how to group issues together. I wonder if it would make sense to flag them all with "Feature: Download Count". A new GitHub label I mean.
13:41 pdurbin
There's already quite a long list at https://github.com/IQSS/dataverse/labels
13:41 rebecabarros
Yes, if the dataset is a draft the count is not increased. For me it sounds strange to count when the explore option is used but since this option makes the metadata download, could be that the reason.
13:43 pdurbin
It also downloads the file if you click "Estimate" to run a model. This is because Zelig needs the file to run the model. It might even download the file before you click "Estimate". I don't know.
13:56 axfelix joined #dataverse
13:56 rebecabarros
That could explain. You think I should ask in the community group to see other people opinion about it?
13:58 pdurbin
You would reach a lot more people that way.
14:09 rebecabarros
pdurbin: About the ingest that I mentioned yesterday, that I took almost 30 minutes, remember? I try again today and I noticed in the server log that apparently the process is done in a few seconds (https://pastebin.com/JvL1kiFZ) but I don't know why in the dataverse interface it takes all this time to stop showing the 'ingest process' and update the page with the tabular version of the file.
14:10 pdurbin
hmm, so you see "Tabular data successfully ingested; DataTable with 334 variables produced" right away
14:11 pdurbin
rebecabarros: you *could* try interacting with Dataverse via API . Upload the file via API, then download data via API. Just a thought.
14:16 rebecabarros
pdurbin: Yes, is an option. I will try that to see how it goes. Thanks
14:33 axfelix joined #dataverse
18:08 axfelix joined #dataverse
19:03 axfelix joined #dataverse
20:17 pdurbin left #dataverse
