Community Call Tomorrow (10/10)

47 views
Skip to first unread message

danny...@g.harvard.edu

unread,
Oct 9, 2017, 7:31:18 PM10/9/17
to Dataverse Users Community
Hi everyone,

Please join us for the Community Call tomorrow (10/10) at Noon EDT! Info:


See you soon,

Danny

Philip Durbin

unread,
Oct 10, 2017, 1:12:48 PM10/10/17
to dataverse...@googlegroups.com
Great call everyone. Here are my notes: https://docs.google.com/document/d/1VlZeMgJy2s1FZ0TNwAoaqpMrl2DixXX3rlZDI9n2Qvg/edit?usp=sharing

2017-10-10 Dataverse Community Call

Agenda

* 4.8 Release
* Community Questions
 
Attendees

* Danny Brooke (IQSS)
* Gustavo Durand (IQSS)
* Phil Durbin (IQSS)
* Julian Gautier (IQSS)
* Pete (HMS)
* Courtney (TDL)
* Sherry Lake (UVA)
* Amber Leahey (SP)
* Kevin Worthington (SP)

Notes

* 4.8 is now available on Github
   * https://dataverse.org/blog/dataverse-48-release
   * Expect a 4.8.1 release with performance enhancements.
      * https://github.com/IQSS/dataverse/issues/4173
   * (Amber) How can we test new 4.8 features such as large file upload?
      * (Danny) Large file upload is an either/or for HTTP vs. rsync. You can't use both currently.
      * (Pete) I plan to improve the documentation.
      * (Amber) Portage in Canada might be interested in big data support via Globus.
      * (Pete) We did look at Globus for file uploads but decided against it because it doesn't support client side checksums.
      * (Sherry) Harvard Dataverse is running 4.8 but not rsync, right?
         * (Danny) Correct.
      * (Courtney) We are interested in having both HTTP and rsync. Do we know what release this will land ini?
         * (Danny) It's not on our roadmap yet. Stay tuned.
      * (Gustavo) Per https://github.com/IQSS/dataverse/releases/tag/v4.8 there is a change in behavior having to do with dataset locking. The logic is based on permissions (but it’s easier in practice to describe with roles). As of 4.8, when you send a dataset into review, the dataset is locked and can't be edited by curators. The curator has to send the dataset back to the author even for minor edits like typos.
         * See this issue: Curator should be able to edit a Dataset that is "In Review" - https://github.com/IQSS/dataverse/issues/4139
            * Curator is representative here of a role with both “EditDataset” and “PublishDataset” permissions.
         * (Sherry) Roles and permissions are confusing.
            * (Gustavo) There are recent improvements in documentation in 4.8. See https://github.com/IQSS/dataverse/issues/2653
         * (Danny) Do we agree that the person who submitted the dataset for review shouldn't be able to continuing to edit the dataset while it's in review?
            * (All) Yes.
         * (Danny) There's author the journal use case to consider.
         * (Gustavo) Some day we'll build a GUI to allow the curator to send a note back to the author, perhaps as part of https://github.com/IQSS/dataverse/issues/3702
* Community Questions
   * (Julian) Derek and I have been figuring out a process to determine the usage of various features within Dataverse. We now have access to a copy of the Harvard Dataverse database and have been writing SQL queries, shared in https://github.com/IQSS/dataverse/issues/4169
      * (Kevin) How does this relate to miniverse? We may be interested in running it some day. Downloads by dataverse would be interesting.
         * (Julian) Perhaps this could lead to improvements in miniverse.
      * (Amber) We're interested in an institutional view of usage (where each institution is its own dataverse within an installation).
      * (Courtney) We also have a need for an institutional view. We have set up miniverse.
      * (Phil) It might be easier to maintain the code within Dataverse itself.
      * (Kevin) We'd rather have the metrics built into Dataverse especially if miniverse isn't well maintained.
      * (Pete) Are we talking about two types of usages? Software vs users.
         * (Danny) We want product development to be informed by metrics.
      * (Sherry) There's also Altmetrics and similar that can take into consideration citations and views outside of Dataverse. The "coming soon" text is gone.
      * (Courtney) We don't have the greatest usage statistics coming out of the system, which is what we're asked for the most.
      * (Pete) One more vote of agreement for having metrics built into Dataverse.
   * (Kevin) Where are we with the integration of Data Explorer as part of Dataverse? What sort of workflow should I be following to get my code contribution incorporated?
      * (Danny) How to explore data is one of the big questions in the current effort on sensitive data. It will be an opportunity to discuss with the community how data is explored in general.
      * (Amber) I know the Australian Data Archive is coming to visit at the end of October. Maybe we could have a community call where we demo the latest version of Data Explorer. It now works with cross tabulation and lets you tag variables with weights.
      * (Danny) Can we access Data Explorer publicly?
         * (Kevin and Amber) Yes. Here's a link: https://github.com/scholarsportal/Dataverse-Data-Explorer
   * (Courtney) Thanks for the help with the robots.txt code, Pete.
   * (Kevin) We're working on the integration of spatial data. Do you know if you're going to be changing much with re-integrating with WorldMap?
      * (Danny) https://github.com/IQSS/dataverse/issues/4164 talks about changes we need to make in Dataverse to work with the new version of WorldMap. I'd recommend holding off until this work is done.
   * (Sherry) We're working on our new fancy/dancy home page.
   * (Amber) File organization at the dataset level. When you have many files, the scrolling is very tedious. It doesn't encourage people with more than 20 files to retain structure to their dataset. Is there anything in the roadmap to help support structure at the file level?
      * (Danny) The design team is looking at it. Work on DataTags may be a good opportunity to review how useful we can make file tags general. See https://trello.com/c/4hTOKlx2/45-file-tags
      * (Amber) I thought I saw a mockup that does a good job of summarizing the tagged files. It's nice to be able to tell what types of files they are. It would also be great to have a hierarchy of files.
         * (Phil) Is the mockup in https://github.com/IQSS/dataverse/issues/3404 ? yes
         * (Amber) Yes. This mockup about Dataset-level landing pages is in #3404: https://cloud.githubusercontent.com/assets/687227/19237328/521c5f0a-8eca-11e6-86f5-376da9464b25.jpg
         * (Amber) There's also a mockup in https://github.com/IQSS/dataverse/issues/2465 about File-level landing pages at https://cloud.githubusercontent.com/assets/687227/18326321/2c3972ee-7513-11e6-94bb-1146f6da2826.jpeg
   * (Phil) Thanks for feedback on the Dataverse Community News.
   * (Danny) Thanks to Sherry and Pete for helping Venki with his customization questions.

--
You received this message because you are subscribed to the Google Groups "Dataverse Users Community" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-community+unsub...@googlegroups.com.
To post to this group, send email to dataverse-community@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/dataverse-community/585094c6-98b1-4500-8dcf-f7fce60597e2%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.



--
Reply all
Reply to author
Forward
0 new messages