2010-04-27 Tuesday Meeting
Attending: Brian, Donald, Ola, Will, Jean-Marie, Scott, Chris, Andrew, Colin, Josh, Carlos, Melissa, Curtis
Agenda
Remember: Agenda must be complete (with estimated times) on the day before the meeting. Any additions after that must go at the bottom (AOCB)
Accepting minutes from last meeting
Matters Arising (<10 mins)
- Jean-Marie: Put up the PDF version of Jerome's presentation
- Donald: Validate proposal for ROI changes with Jerome
- Everyone: Go through the 'unknown' tickets (report 28) and own your tickets BY NEXT TUESDAY
- Josh & Chris: Find out why the hybernate upgrade didn't give us a performance boost
- Everyone: QA build for Friday night so we can test over the weekend
- Chris: Wipe Mage clean for weekend testing
- Jason: get agenda up for Paris meeting
- Jason: We need to define requirements for official OME-XML recognition
- Andrew: Need a compliance spec pages similiar to the minimum spec page link:ComplianceSpecification (page needs content added)
- Andrew: Send e-mail to lists asking about ROI usage in preparation for 3-dimensional additions
- Andrew: Document the deprecation of Path and PolyLine in 2010-04 (removal in 2010-09)
- Andrew: Proposal diagram to be put up on OME-XML Trac and linked in e-mail
- Donald, Jean-Marie: To solicit feedback from Jerome on VertexArray and 3-dimensional geometry.
Today's presentation: Scott talking about the background to his PhD (45 mins)
AOCB (<5 mins)
Notes
Meeting minutes
- Josh: report 28? and when do we decide what to drop?
Matters arising
- Jean-Marie uploaded PDF
- @Donald needs to contact Jerome re: ROIs
-
Report 28: some of them will need to be moved
- Josh: should be careful to not just push all tickets which have no owners
- some are important
- QA build was done along with some testing (mage cleaned up)
-
compliance page: up but is missing content
- @Andrew takes the lead on the content
- won't happen until after release
- @Brian will ask if Michael P. would like to present next week
- Should consider getting a solid testing frameworks
Scott's presentation (17:17-17:46)
- "Began by trying to understand what usability was"
- insights from scientists, but what was critical to the SD research?
- is there a flexible enough framework for such a dynamic dev team?
- scientists don't see the infrastructure underneath OMERO
-
"special requirements on scientific development..."
- "academic funding is difficult to come by"
-
Q&A
-
Jason: in the group meeting, Michael asked is there a new version?
- "...not on nightshade, it'll break everything..."
- technical challenge & the rhythm of development
- Scott: in principle, agree with agile principles * but developing in this context doesn't provide a suitable rhythm * acknowledge and allow for that instability
-
Josh: don't have a good feeling for the 'constraints of the scientific development process'. Not sure what you mean by this
- Agile 'rhythim' seems to do exactly what the scientist needs
- Chris: the users are different * Agile not ideal for our 'underlying structure/api changes'
- Donald: we actually have large changes
- Donald: what is usability?
- Jean-Marie: you only see the usability from the client POV * There's also the model-users, there's the API-users * Balance between which entry. Many point of interactions * Jason: Michael is tense because we're going to break things * Jean-Marie: community has to communicate back at some point * Broke down at some point in time. Pipeline is blocked.
- Scott: because of the nature of the communications, they can get broken * inconsistencies can arise
- Jean-Marie: is one of the constraints trying to build a community? * One goes from one workflow to the next, adding slowly * cF. FLIM, EM, ... * How can we address larger workflow in one shot? * Donald: when you ask a bigger question, they don't seem to want to answer * Jason: they answer, but just two words: "data management" * "Got to work with ImageJ" when pressing. How to get beyond that? * There was a process by which we got useful information. * Brian: keep equating users to scientists * But there are different breeds. * Seems like we should be identifying groups of 'users', and define their needs * Scott: first step is to identify that and re-evaluate over time
- Jason: according to agile, how do you handle multiple product owners? * Josh: it's more technical. if there's value, then you have to refactor to support the various views. * Scott: need to be able to tailor something * Josh: How do other projects with 'multiple user groups' handle this problem? * Chris: the difference is that there's a core group of developers who make all the decisions * Jason: ROI is a good example. Multiple consumers. * FLIM, EM, etc. * What can we do to deal with that? * Lots of good will
- Jason: what's the point of having developers with the scientists, * if we've failed to communicate what's coming? * Brian: we've misjudged before, small change (Will: cf. private) * Jean-Marie: How can we make the messages about process change clearer? * Chris: Understanding what the users are consuming helps us to know what to notify them about * "What's the public API? Who are we going to affect if we change something?" * Comes down to staffing; without going out and asking the community, there's no way to know
-
Jason: in the group meeting, Michael asked is there a new version?
Action Items
- Jean-Marie: please add the presentation PDFs to the website
- Donald: please contact Jerome re: ROIs
- Everyone: Go through 'unknown' tickets during THURSDAY's meeting
- Andrew: take lead on putting up the 'compliancy page'
- Brian: Ask Michael if he will present his tools next Tuesday
- Brian: Proposal for testing checklist (hopefully by next Tuesday)
- Jason: REMINDER need to get Sam Swift involved in OMERO
- Andrew: Add Till Phontonics to the website