Skip to end of metadata
Go to start of metadata

Agenda

  1. Intro and Agenda Bashing
  2. Update on current status
    1. Prototype 
    2. Sandbox
    3. Tests
  3. Holistic view of Architecture
  4. Roadmap for future work
    1. Architecture definition
    2. Prototype 2
    3. myExperiment integration
  5. Discuss next Arch face to face meeting
    1. Agenda
  6. Dissemination
    1. Papers/demos
  7. AOB

Participants

  • Graham
  • Jits
  • Pique
  • Jose
  • Guillermo
  • Piotr
  • Raul
  • Jun
  • Dave (parts)
  • Kevin
  • ...?

Coordinator: Stian

Date: Wednesday 8th June 2011

Time: 11am (BST) / 12pm (CEST)

Duration: max 60 minutes.

For WebEx details see: Telecons

Chat Log

Stian is scribing.

Graham:
It seems strange to disconnect from one webex conf, then connect to another, to
talk to the same people again (smile)

Agenda bashing

Stian: Rough agenda:

  1. Intro and Agenda Bashing
  2. Update on current status
    1. Prototype 
    2. Sandbox
    3. Tests
  3. Metadata discussion 
  4. Holistic view of Architecture
  5. Roadmap for future work
    1. Architecture definition
    2. Prototype 2
    3. myExperiment integration
  6. Discuss next Arch face to face meeting
    1. Agenda
  7. Dissemination
    1. Papers/demos
  8. AOB

Suggests to drop Metadata discussion (this agenda was just copied from last meeting).

Graham: HAving discussions on technical requirements, thinking of poissible frameworks, structures.
to make sure we get a holistic view of the requirements to make sure it doesn't fall into the gaps of WP
as architecture TF which has project wide view, it might be worth raising this

Prototype update

Stian: Moving to Update first - Prototype

Raul: Not much done in the last few weeks
trying to develop a proof of concept prototype/iteration
for using editions - a mechanism for saving/backing up ROs
might be an non-explicit mechanism like versioning - more frequent without user
'tagging'
started working wit this to see if it can be implemented with dLibra
workin progress

Stian: Is it looking at existing solutions like Git/Mercurial?

Piotr: yes - looking at using dLibra initially
editions - small versions - only differences between different editions
kind of mercurial-style
but designed for publications, not code files
see how we can go on with this before the Oxford meeting
then see if this is what we want - or need more

Moving on to ROBox

Jits: There's an action on tests

Graham: Made progress on tests - ran into the versioning issue
Overcame most of the other blocking issues
Can respond quickly to functionality changes

Jits: RoBox - not much work except bug fixing
make it work on the sandbox and with Graham's test suite
An important question - we were in prototype 1 - we are waiting for F2F before
moving on to prototype 2
should we do something else in the meantime

Jose: need to converge in the scope of the architecure TF

from Graham Klyne to Everyone: I think it would be unfortunate if we stop all
progress until the architecture F2F... but not sure what we can usefully do on
the implementation front.

Jose: Should now think of what we want to present for this years review
mainly means - how can we reflect advantages of this work
Reflect this to convey what Wf4Ever is all about
Discussed tis in the poject management board - PMB meeting 20110621

See how to move this forward - a new prototype - trying to start converge
towards what we'll have in the end - including myExperiment

Jits: That's in line with what we are thinking
What I mean by prototype2 is - bring align myExperiment, RO Model - everything
we found form RObox. Prot2 is what Jose said - what is the question is if we
should do something now to help with the decission making

from Graham Klyne to Everyone:
So where does the driving force to convergence and impact come from? Users for
sure. But also a research vision we should reflect?

Jun: Not been following proto1 much as Graham has focussed on that - coming
from our side - how to move to next step - interesting user feedback and
outcomes from protoype 1 - but has this been circulated to the project like the
RO TF? thinking of the next 3 months

Jits: Will circulate the wiki pages on the feedback
this has already fed back to the deliverable
say te distributed nature of research objects
Marco fleshed out many of his reqs based on ROBox work
As Graham said we need to consolidate and move forward

Jun: Lots inthe deliverables - but where is each piece
scattered around
related to the Holistic agenda item
To keep all the feedback in order to possibly write it up

Jits: Tried to avoid duplication - as we focused on deliverables - it might be
scattered around now - need a report to see what we found
fair bit of work - but we can do that

Jun: Yes -b ut don't make it too heavyweight. Jose?

from Jits and Pique to Everyone: 2011-04-11 Demo to Users
is some of the findings from the first impression from users

Jun: Do we want a write-up of what we found from prototype1 - would be useful -
but don't want to make it into a full-fletched deliverable-style - just a
lightweight doc

from Jits and Pique to Everyone: Dropbox RO Connector (ROBox) is the overall ROBox work
ROBox use case are the use cases we were meeting

Jose: Don't bring too much overhead onto the archtecture team - but collect the information

Jose: Most of it on the wiki - just structure it a bit

from Graham Klyne to Everyone:
My refelctions here: Reflections on Month 6 deliverables creation process

  • not what asked for, but suggests a lightweight style

from Jits and Pique to Everyone:
Research Objects Digital Library on ROSRS

Jits: Pasted in some links - people to look at this.. or we can put this in an email

Graham: Just have a simple page with links with indication of what might be there

Jun: Different indexing

from Graham Klyne to Everyone:
I think the wiki is not easy to navigate, too.
Doesn't help that it's rather slow sometimes.
@stian, +1, tree structure not enough

Stian: Think about updating other pages when a new page is added to the wiki

(Jose - muted)
Jose - 've muted you - let me know when you want to speak

Raul: Do plan something to have before the f2f
For prototype2 work to have some initial discussion - .. (question) with myExperient - how to do some initial integration
Will have some experiences before the f2f

Dave is here from the car
Dave: Hoped to do the early myXP work in the run-up to the deliverables - but this didn't happen as
Don was unavailable
Plan B - in the runup to archtecture f2f we'll do some work together - David Newman and Danius in Southampton

from Graham Klyne to Everyone:
Ah, yes, doing some integration with other systems would provide useful input, I think. Would be of
value reagrdless of what we conclude at F2F.

Dave: Lots of potensial - discussions on what to do.
Something new to do on myXP we'll probably do in Oxford
Interested is to consider the .. (question)
intermingled world

from Graham Klyne to Everyone:
... the world where we have succeeded.
... what would it look like?

Dave: myXP can talk about existing ROs - places we can import research objects from

Jun: Agree - really try to show added value of RO by taking up low-hanging fruit as soon as possible
find shortcuts using existing works

Dave: Just looking at existing things like papers
Had a presentation at MS IEEE eScience workshop in September

from Graham Klyne to Everyone:
Jose was asking for a focus on showing advantages: I think Dave's suggestion
of contemplating the world where we have succeeded would provide a useful
counterpoint to the immediate userrequirements.

Dave: next stage is to do what we said - see what's succeeded.. (question)
Computational different from executable

from jun (privately):
@graham, yes. that's the kind of user-driven and research-driven balance we need to take care

from Graham Klyne to Everyone:
Propose for requirements gathering:

  1. User req acquisition in WP5 / WP6 - collect to living document.
  2. Technical requirements extraction for all URs under architecture WP umbrella - provide holistic view
  3. tech WPs select relevant tech requirements and develop designs for these

Holistic view of Architecture

Graham: Make sure things get carried through from the user
3 phases. Some of these we do already
User requirement acquisition - should belong in user taskforce - we're doing to much of that in the
technical work packages without appropriate user focus. should fit in WP5/6
Bring together user groups to see how technical requirements can come from this - in
terms of the architectural choices
Do this across all user reqs - without regards for technical work packages - something
we can do under the Architecture WP
to respond to the users and give a technical vision
This would generate the technical reqs - from there the individual WP can identify the reqs relevant
for their goals and work on the designs
by getting all the reqs together in one group at least we can be sure that we are
looking at all and not missing big gaps
We might actually save some work as we did some duplicate work

Piotr: Should this be done before/after the Oxford meeting? What stage?

Graham: We already have somethin to work on - I'm attending to continue on this, already
working on this with Marco/KRistina on Bio side

Raul - I've muted you - let me know if you want to talk

Graham: to guide some discussions on the f2f

Raul: While we have workking documents on the wiki with summary of functional, technical..
Already started using these documents as inputs
driven by reqs by users, like versions of ROs, compare, undo

from Graham Klyne to Everyone:
yes, and I thimnk it's important that we don't lose these requirements, but we need living
documents that can go forwards

Raul: Already having intiail work on this, but would like to have a working document with
a more stable state

from jun to Everyone:
+1

Graham: ave the reqs in the delievables - important not to loose this. Have gone trough the
Deliverables to make sure nothing was lost in the live documents that I aim to create
in a deliverable they are fossilised - need to make it lving

Kevin: Sounds very good - agree on this. Step 2 sounds useful. Hoping for the arch meeting
By putting step 2 in, we wonder how this continues as the process goes iterative/refines -
putting a step between people who implement and users in step 1
Do we have good enough representation .. (question)

Graham: No clear answer now - but we should be aware of this - do we have the right thing
at the right time

Kevin: Not always sure that we've got everything in architecture
repeating steps: Step 1: User req acuqisiution 2: req extraction 3) select reqs

from Jits and Pique to Everyone:
(as an aside, the wiki is down (sad))

Stian: Graham - are you making this live document?

Graham: Started working with Marco/KRistina - but not in position to do this for astrophysics

Graham will try to make sure nothing fall in between the gaps

Stian: Manchester will work with Pique in filling in more Astrophysics reqs

Pique: move to get a common document for both communities
Can go behind Graham's document and map it with our reqs - will do this here in Manchester

from Graham Klyne to Everyone:
BTW, I'm currently thinking of keeping the live documents as spreadsheets in the code repository

Stian: Could be problems with conflicts
Wiki has died!

Stian: difficult to talk about uture work right now

Stian: enough to work with until F2F?

Graham: enough to keep busy!

Dissemination

Raul: Should we try to write a short paper about what we've done already

Stian: Might be difficult to show something as we don't hav emuch results yet

Raul: Perhaps not a paper - but a demo or short proposal paper

Jits: Any particular target in mind?

from Graham Klyne to Everyone:
On dissemination, I'll be attending a provenenace for linked data workshop in edinburgh next
week. Also, ongoing provenance WG work. Not sure that this is relvant here.

Raul: like some demo of the prototype

Jits: Might be more useful to... he's summarising a page of the prototype 1 - might be useful
to make a video to show how it works
disseminate it amongst our selves before going more public

Jun: Might make a demo at ISMB - but we've missed the deadline
It might be difficult at the moment - digital preservation conferences?
a demo proposal for tose kind of conferences

from Graham Klyne to Everyone:
Also, documentation for our users to use the sandbox/demo might be useful?

Jits: Prototype is a bit small - would it be a good ting to demonstrate?

from Graham Klyne to Everyone:
(I tend to agree with Jits)

Jits: if we want to show what we are working on then we can package it differently

from Graham Klyne to Everyone:
I think the myExperiment integration might be useful to demo, when it happens.

Stian: Might be difficult to show prototype out of context - perhaps better a presentation of where
we want to go

Jun: Lourdes have suggestions for the website - too technicals

Jun: PAckage up demo to show to users coming to the project web site on what we've been working
on - ideas
some content or whitepaper to publish on our site

Jits: Using it to help website is good idea
Must make it inline with project goal - just show prototype1 does not convey much of the project
goal alone
But use it in a context of the whole

Jun: Yes - show how we've started - not at download-stage
A general one

Jose: to show precisely what it is - testing the water and see how things really are
as Lourdes propose should be fine
we have to be careful not to convey that "This is what we will deliver" - then we could have problems in the future
try to reach a balance

Stian: Who takes tis forward - Jits might you do a video?

Jits: Agree with Jose - perhaps not a video - people might think "this is what Wf4Ever is" so back off with the video

Jun: Sounds good

Any other business

Hotel/travel for architecture meeting in Oxford

jun: Dave encouraged people to make sure they are there for lunch first day

Piotr: Might go day before - london then train

Jun: can find hotel information

Kevin: Also looked into this - think some of the college rooms are available

from Kevin Page to Everyone: http://www.oxfordrooms.co.uk/

Actions

  • To extract from D2.1,3.1 and 4.1 the requirements, prioritized, and use them to inform the architecture definition
  • To extract user requirements from D5.1 and 6.1 and use them to inform the architecture definition
  • Graham to form live document of (initially Bio) user requirements
  • Stian/Jits/Pique to fill in live document with Astro user reqs
  • No labels