How to Start a Portal

Key considerations:

• Depending on the project need, portals can be driven dynamically from Views or statically from Tables. When choosing consider:

  • Files are only available to bulk download through Views. (Bulk download feature not exposed in the web UI in Tables).

  • There is potentially a performance hit in Views (in dispute; has not been fully tested) if there are varying access permissions across the set of files and many facets.

• External content that the portal renders within the site must be hyperlinked "https". Links out can be http or https.

• Removing facets can improve performance (maybe?). If you set a facet in a Table or View, it will be displayed in the portal.

• Sharing settings on Tables

  • Make Public - Anyone on the web, Can View. To surface Tables in the portal that are in project with restricted sharing settings, Tables must be explicitly whitelisted {document process to whitelist} by governance and someone on the leadership team. This is required any time the schema changes or values are added.

  • A change to ACL could disrupt visibility although, whitelist status is still maintained in the back-end.

Jira

Portal development coordination is tracked in the Community Portals Jira project. Create an appropriate Epic within your own project (for the computational oncology, systems biology, or neurodegenerative research teams, this should be in the SCIENCE project) to track the overall project. It is the job of the respective portal product manager to triage issues to the engineering team (Alina, Michael and Jay) appropriately. ‘backlog’ label is required to be visible to engineering team and product manager.

Ljubomir Bradic (Unlicensed) is responsible for triaging and assigning design related issues, and Ryan Luce (Unlicensed) for portal engineering components.

What are the first steps?

1. Fill out the 1-pager for product and design and submit a Jira in the PORTALS project for review.

2. Open a Jira to task the design team with the creation of relevant wire frames. Designers will implement logo and color scheme. Depending on the level of customization or needs for external parties, a complete wireframe may not be required. This is at the discretion of the design team, so don’t plan on not having one unless confirmed with Ljubomir Bradic (Unlicensed) .

3. A Synapse Project drives the backend of the portal. Make a project (e.g. PEC).

4. Open a Jira issue with PORTALS to (e.g. PEC) create required infrastructure. This creates an empty AWS bucket.

5. Implement design in the staging site by opening a Jira issue with PORTALS (e.g. PEC)

6. {TODO:determine what is required to make these connections} Create Table/View architecture

   1. All data View

      1. If you want to surface data entities in the portal,

   2. Publications

      1. Set schema

      2. Tool to scrape Pubmed for relevant publications using grant or pubmed id as input

   3. People

      1. Set schema

   4. Study

      1. Set schema - note "hint text" is tracked in Confluence page and must be manually set by web developer

   5. General notes about table schemas and column types

      1. Simple markdown (links, bullets, bold, etc) is supported in underlying tables, but you will need to specify which columns have markdown in the schema

      2. if you need to link to multiple synapse IDs, separate the list with commas

Use these questions to guide conversation with the design team about the homepage:

1. Why/goals/value- what is the focus of this portal

2. Relationship to consortium

3. Fit in ecosystem

4. Type of resources available

5. Diversity within type/amount within type

6. Main actions- e.g. find data/tools/people

7. Progress and future-- living repo/freshness of data

8. FAQ/help

Testing on Portals

1. Scenarios to test on staging when updating portals