RequirementsList

1. Access Control

• Access control (groups/attributes/collaborations) - publish provenance but only the bits that I want other people to see; sharing attributes: edit, execute, read; decide who you want to see your work and who you don't.
• Priority: high.
• Related heading: 23
• Comment: using predefined groups and user-definable groups? Publishing defaults to the user only with 'world' as the superset of all groups.

2. Annotation for Everything

• Easy annotator - easy to use annotation tool for objects in portal - annotation of any object in the portal and linking them - audit trail that can be annotated by the group.
• Priority: ?
• Related heading: 6, 11, 12, 14
• Comment: does this imply an ontology or, a least, a controlled vocabulary? What about free text annotation?

3. Automated Process Logging

• Automated capture of the whole lab process - logs: workflow audit, lab books.
• Priority: high.
• Related heading: 17, 23
• Comment: with filters to select different event types?

4. Personalisation

• Customisable GUI, project logos, colour scheme.
• Priority: ?
• Related heading:
• Comment: needs clarification. Does this come with the portlet container?

5. Certification of Content

• Certification of objects/content – domain packs: kite marked – certified workflows – certified services with annotation.
• Priority: ?
• Related heading: 2, 10, 11
• Comment: our impression is that this is critical if workflow re-use is to flourish beyond the local trust relationships. Who would do the certifying?

6. Content Seeding

• Content seeding (community building) – seeding.
• Priority: ?
• Related heading: 7, 8, 10, 15
• Comment: mygrid user groups need to be approached about contributing content and especially ‘core workflows’.

7. Groups in Communities

• Fellow researchers in domain – find other people in your field for possible collaboration – explore “neighbourhoods” (“people who are interested in similar things as I am”) – find people that fill gaps in my, or my group’s, expertise – bring together people with overlapping expertise & interests – trust networks – project working – show other people your workflows – find workflows related to your subject, analysis, task – users network – forthcoming meetings for domains, e.g., links to calendars or sites.
• Priority: ?
• Related heading: 1, 9, 11
• Comment: various search criteria based on the notion of community member profiles. This assumes that community members are searchable entities as well as workflows.

8. Me-Section/Incentives

• Overcoming reluctance to share – generating citations for data / services used -> “reuse index” – citation index: users, labs, … - must include some “prestige” to encourage participation to help achieve critical mass, featured projects? – generate expert or network group […?] e.g., google sidebar – “me” section, incentives, prestige – preserve my intellectual property.
• Priority: ?
• Related heading: 7, 9, 11
• Comment:

9. Feedback/Connecting People

• Linking potential […?] allow connecting on your project: “you might be interested in the work of X”.
• Priority: ?
• Related heading: 7, 9, 11
• Comment: the capacity to define ad-hoc neighbourhoods of community members. This could be done automatically through profiles or explicitly through some sort of joining mechanism. The latter implies someone acting as coordinator, etc.

10. Technology Repository/Best Practices

• Use groups of workflows to find “gold standards” and accepted methods – open lab network (labs willing to share techn.[?]) – technology repository: where that technique (wet lab, compute) is available.
• Priority: ?
• Related heading: 5, 20
• Comment:

11. Peer Reviewing/Ratings

• Coolness rating, encouraging blue sky research – peer reviewing – rating of things (workflows) put in.
• Priority: ?
• Related heading: 2, 7, 9
• Comment: key to mobilising the community.

12. Tagging

• Google over everything – search by tags – tagging as a game – tag workflows – include some free […?] about this service/workflow – easy discovery mechanism, self tagged rather than fixed vocabulary: contentious? – tag autocompletion – simple tagging for discovery – allowing users to annotate services / workflows.
• Priority: ?
• Related heading: 2, 7, 9, 11
• Comment:

13. Grouping/Clustering of Services and Workflows

• Clustering the services supporting same function – find examples of workflow patterns – service harbour, telling me which are: easy to used, with examples of journal papers based on them, which are alive and kicking.
• Priority: ?
• Related heading: 16, 17
• Comment:

14. Ontologies

• Ontologies of service description – “virtual professor” (finding things) – meta stuff: library of analyses performed linked to data used & reports/results (searchable) – ontologies: of domain, who uses, who develops, what mappings.
• Priority: ?
• Related heading: 2, 7, 11, 12, 13
• Comment:

15. Finding Publications etc. Related to Workflows

• Picture + citations pop up when you download/use my WFs/data – reference site for workflow used/cited in literature; publish paper, link to WF that will be there in 5 years.
• Priority: ?
• Related heading: 16, 18
• Comment:

16. Registry and Searching

• “Go shopping” for services, workflows, data, publications, funding – a big registry across workflows worldwide.
• Priority: ?
• Related heading: 13, 15, 17
• Comment:

17. Basic Workflow Functionality

• Workflow repository: download, execution, searching – create workflows, upload workflows, store workflows related to an ontology, public repository, documentation of workflow.
• Priority: ?
• Related heading: 19, 24, 26
• Comment:

18. Writing a Nature Paper

• “write a nature paper portal” – act as a supplementary work / data store for publications.
• Priority: ?
• Related heading: 15
• Comment:

19. Semi-Automatic Workflow editing

• Templates of “general” solutions + examples – translation service to help link various data sources (shims?) – building & tweaking workflows – pimp my workflow/service.
• Priority: ?
• Related heading: 17
• Comment: Distinct from full editing capabilities available in the workflow ‘author’ environment, part of the workflow ‘enactors’ environment.

20. Demos/Guidelines

• Demo movies “ruby on rails” style – demos, guidelines, documentation etc. FAQ – support for end users: FAQ, known bugs, online support – examples of usage of web services / workflows.
• Priority: ?
• Related heading: 5, 7, 26
• Comment: Important but do we have the resources to do this as well as development?

21. Complete Integration/Interoperability

• Mash-up-able – Taverna integrated in BIRN portal – run in every browser – Taverna rewritten in AJAX (browser only interface) – Seamless desktop, web, email – cross linking different myExperiments.
• Priority: ?
• Related heading:
• Comment:

22. Data Management and Visualisation

• Data visualisation in general + management – data mangling functions, advanced data types – data vs. processed data vs. interpreted or annotated data – make a result pool, a database of databases with saved transformations – access to quant[…?] data, summary description of data/database (searchable), formal description of data in database that allows linkage, link to work done using the data in the database.
• Priority: ?
• Related heading: 24.
• Comment: Data management potentially very generic and possibly better afforded outwith the portal where users can access their preferred tools. Data visualisation has potential relevance to interactivity but, again, raises issues about integrating the portal with users’ preferred tools.

23. Confidentiality

• Data confidentiality – security – authentication + authorisation that is painless – non-seccure space allowing anonymity, a la wikipedia.
• Priority: ?
• Related heading: 1, 3
• Comment: How much will users be prepared to have disclosed about, e.g., which workflows they are using and to whom?

24. Interactive Workflows

• User interaction as processor in workflows in portal – workflow steering, interaction at runtime.
• Priority: ?
• Related heading: 17
• Comment: This came up a number of times. From one perspective, this begins to look like a requirement for supporting the research equivalent of the ‘business process’ rather than simply workflows.

25. No Errors/Only Solutions

• No errors, but solutions – Problem: Taverna consumes memory: destroy objects by telling garbage collection to start – assume it will work + have process for exceptions + the community – graceful failure.
• Priority: ?
• Related heading:
• Comment: More of a core Taverna workflow engine requirement.

26. Pre-Run Workflow Diagnosis

• Can I run this workflow? – Pre-run workflow diagnosis – Problem: many web services failed -> ping web services, if they are not available, don’t show workflows – diagnose whether the services are usable or workable – knowing about consequences of actions, e.g., compute intensive jobs.
• Priority: ?
• Related heading: 17, 20
• Comment: Suggests a range of test functions need to be available including basic pre-run tests for the enactor environment. Pre-run tests would be more straightforward if workflows were published with test data.