attendees: Julia Trimmer (Duke) Eric Meeks (UCSF), Steven McCauley (Brown)
Let’s talk about what institutions need most on the VIVO Road Map. Here’s the link to the tweet about the Road Map poster at the conference (twitter.com/VIVOcollab/status/893487651713679362).
What looks most helpful?
Steven likes the TPF endpoint (nice to have although their linked data is not being consumed), and the improved ontology editor. It would be great to have libraries updated: Jena 3 would be good, as would upgrading from SOLR 4 to 6.
What about the micro-service architecture further out on the road map? This is kind of a vague term and can mean different things. But it would be great to have a better API for SOLR or another search index.
Currently, the VIVO components are very tightly coupled and “monolithic” so it’s hard to make changes without understanding all of the architecture. Steven thinks it would be great if the architecture would be less tightly coupled so that different universities could share components of the project.
Brown has developed a new frontend for their VIVO and it’s going live next week as a beta. It’s a standalone rails app that consumes JSON from SOLR. It’s VIVO agnostic, so you don’t have to know VIVO to work on the front end. They will be sharing it with the community and it’ll be on GitHub in its own repository.
Julia asked, is your new front-end an example of micro-service architecture? Steven said that micro-service is usually more component-ized and more API driven.
Creating a version of VIVO with a micro-service architecture would be a huge endeavor. Eric pointed out that it’s a tough sell as well, since typically, no new features come out of such an effort but it takes a great deal of work and the end result looks the same even if it’s a lot better. The business side of the team just needs to be patient about adding new features because it’ll be a lot easier to do with the new architecture.
Julia thinks that the community would benefit from any initiative that gets developers excited and in which more institutions would be willing to collaborate. It would be ideal to start small and see if the excitement and commitment grows and continues.
Steven said that making the front end easier to develop is a very high priority.
The hard part of what Brown just did with their new front end is using SOLR: managing what goes into the SOLR index from within VIVO is not straightforward. You really need to know VIVO to do it. It’s necessary to update a text file in order to manage the data in the SOLR index. It would be great if VIVO offered a web interface that would make it easier to manage that process.
Julia mentioned that she talked about VIVO to two CIOs at local universities and asked what they saw as the biggest barrier to adopting VIVO. One said that it’s too hard to install and load data in. VIVO should have some kind of templates that would make it easier to load data in different formats.
The other CIO said that basically, VIVO’s front-end needs updating. It’s not modern enough and it needs better features.
Eric pointed out that Profiles is very easy to spin up and requires a spreadsheet with only five columns (sometimes even this is difficult to get from a university!) But it allows people to play with the system, while also enabling them to experiment with more complex features if they’d like.
We wish the best to Brown next week with their new front-end!
