(index page)
UC3’s Development Culture
UC3 takes an iterative deployment approach based on Agile principles. Our development is driven by user-centered design by which product (service) managers gather the needs of our stakeholders and translate them into prioritized use cases. By placing the needs of stakeholders and users at the center of our analysis and design, we develop services and collections that are impactful, useful and used.
We implement agile principles by way of scrum practices, with features that are prioritized, scheduled to be developed and deployed within time-boxed sprints. These activities are carried out via the product manager and the development team via ceremonies. Ceremonies are held to iteratively plan our current sprint, define our backlog, and release “done” product increments. Furthermore, we are reflecting on the process to continuously improve how we do our work. Having ceremonies allows us to stay focused on the tasks at hand, minimize distractions and facilitate communication across the team.
UC3 aims for regularly scheduled deployments for releases, resulting in release processes that are routine and deliberate as opposed to crisis-driven. Releases based on minimum viable products allows us to fail fast — make quick decisions and reroute, when necessary. Releases are comprised of versioning, release notes and updated documentation for new features and support. Integration with other services within and external to CDL are designed utilizing best practices to ensure robust interoperability.
Our focus is on motivation, community, and trust as opposed to structure and control. We strive for aligned autonomy, to give teams the ability to solve problems on their own. Teams decide what to build, how to build it and work together while doing it.
We provide communication via our roadmaps, blogs, and wikis to stay in sync with each other both internally and externally. Check out our team here!
Dash: 2017 in Review
The goal for Dash in 2017 was to build out features that would make Dash a desirable place to publish data. While we continue to work with the research community to find incentives to publish data generally, the small team of us working on Dash wanted to take a moment to thank everyone who published data this year.
In 2017 we worked in two week sprint intervals to release 26 features and instances (not including fixes).
In 2018 we have one major focus: integrate into researcher workflows to make publishing data a more common practice.
To do so we will be working with the community to:
- Release a read only API for download of datasets
- Release a Submission REST API for publishing and versioning datasets
- Implement ‘Make Data Count’- standardized data usage metrics for researchers to have standard views, downloads, and citations of published data
- Integrate with publishers (i.e. submit data to Dash while submitting an article to UC Press)
- Integrate with online lab notebooks (i.e. right click and submit data after analysis with accompanied metadata from Jupyter notebooks)
- Talk to as many labs and researchers as possible to educate on data publishing and better understand incentives and needs
Follow along with our Github and Twitter and please get in touch with us if you have ideas or experiences to share for making data publishing a more common practice in the research environment.
On the right track(s) – DCC release draws nigh
blog post by Sarah Jones
Eurostar from Flickr by red hand records CC-BY-ND
Preliminary DMPRoadmap out to test
We’ve made a major breakthrough this month, getting a preliminary version of the DMPRoadmap code out to test on DMPonline, DMPTuuli and DMPMelbourne. This has taken longer than expected but there’s a lot to look forward to in the new code. The first major difference users will notice is that the tool is now lightning quick. This is thanks to major refactoring to optimise the code and improve performance and scalability. We have also reworked the plan creation wizard, added multi-lingual support, ORCID authentication for user profiles, on/off switches for guidance, and improved admin controls to allow organisations to upload their own logos and assign admin rights within their institutions. We will run a test period for the next 1-2 weeks and then move this into production for DCC-hosted services.
Work also continues on additional features needed to enable the DMPTool team to migrate to the DMPRoadmap codebase. This includes additional enhancements to existing features, adding a statistics dashboard, email notifications dashboard, enabling a public DMP library, template export, creating plans and templates from existing ones, and flagging “test” plans (see the Roadmap to MVP on the wiki to track our progress). We anticipate this work will be finished in August and the DMPTool will migrate over the summer. When we issue the full release we’ll also provide a migration path and documentation so those running instances of DMPonline can join us in the DMPRoadmap collaboration.
Machine-actionable DMPs
Stephanie and Sarah are also continuing to gather requirements for machine-actionable DMPs. Sarah ran a DMP workshop in Milan last month where we considered what tools and systems need to connect with DMPs in an institutional context, and Stephanie has been working with Purdue University and UCSD to map out the institutional landscape. The goal is to produce maps/diagrams for two specific institutions and extend the exercise to others to capture more details about practices, workflows, and systems. All the slides and exercise from the DMP workshop in Milan are on the Zenodo RDM community collection, and we’ll be sharing a write-up of our institutional mapping in due course. I’m keen to replicate the exercise Stephanie has been doing with some UK unis, so if you want to get involved, drop me a line. We have also been discussing potential pilot projects with the NSF and Wellcome Trust, and have seen the DMP standards and publishing working groups proposed at the last RDA plenary host their initial calls. Case statements will be out for comment soon – stay tuned for more!
We have also been discussing DMP services with the University of Queensland in Australia who are doing some great work in this area, and will be speaking with BioSharing later this month about connecting up so we can start to trial some of our machine-actionable DMP plans.
The travelling roadshow
Our extended network has also been helping us to disseminate DMPRoadmap news. Sophie Hou of NCAR (National Center for Atmospheric Research) took our DMP poster to the USGS Community for Data Integration meeting (Denver, CO 16–19 May) and Sherry Lake will display it next at the Dataverse community meeting (Cambridge, MA 14-16 June). We’re starting an inclusive sisterhood of the travelling maDMPs poster. Display the poster, take a picture, and go into the Hall of Fame! Robin Rice and Josh Finnell have also been part of the street team taking flyers to various conferences on our behalf. If you would like a publicity pack, Stephanie will send out stateside and Sarah will share through the UK and Europe. Just email us your contact details and we’ll send you materials. The next events we’ll be at are the Jisc Research Data Network in York, the EUDAT and CODATA summer schools, the DataONE Users Group and Earth Science Information Partners meetings (Bloomington, IN), the American Library Association Annual Conference (Chicago, IL), and the Ecological Society of America meeting (Portland, OR) . Catch up with us there!
