2010-02-23 Tuesday Meeting
In Attendance: Brian, Josh, Ola, Will, Donald, JM, Jason, Chris, Scott, Andrew, Colin, Melissa
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)
- Everyone: next Tuesday be ready to suggest new 'demo' things at next Tuesday's meeting.
The Lab's Current Work: Jasons's presentation (45 mins)
AOCB (<5 mins)
- Everyone ok to start working on Agilo? Review on Thursday.
- QA tracking
Notes
Accepting Minutes from Last Meeting:
-
Collecting example test code
- Chris: We already have some examples
- @Will & Andrew setup an overview after reviewing existing pages
-
What happened with everyone going and having a look at the bio-foramts stuff
- Difficult
- Chris can provide information.
-
@Scott will try to monitor everyone's usage of Agilo
- likes / dislikes
- Chris: a few of the reports no longer worked
- Various bugs. Email Scott and/or Nitpick
- @Chris will put up some slides for Thursday for best use of agilo
-
Collecting example test code
Matters arising
-
Demos
-
Colin gone end of this week for the FS demo.
- We can still shoot for Tuesday next week for a brief talk on FS instead of a full demo
- Discuss possible technologies: FuseFS, etc.
- Carlos week after?
- Chris & Melissa: perhaps wait until they're ready to merge in (beg. March)
-
Colin gone end of this week for the FS demo.
-
Demos
Jason's Talk: OME Informatics & Quantitative Analysis for Biological Microscopy & HCAs
- The bio side of things? or long-term OME work / vision...
- cF. Jcb paper revision, 740 images online / one in paper
-
Perhaps we need another retreat for expressing what everyone is doing (cF. Michael & Matlab)
- Users want more sophisticated functionality
- Approaching scripting
-
We handle Images: time-lapse, HCS, phsyiology, EM, ...
- But not really true
- The systems are getting more powerful
- Tiling! (Large images) we don't do it yet but need to!
- People keep asking for it. What's the market requesting?
-
Interoperability (requirements for image informatics)
- Something we're trying to build. (technically demonstrably)
- Includes: Metadata, Interfaces, and what we will be talking about today, Analysis
- We tend to focus on "backend-y" stuff, incl. data mgmt.
- Need to put more sophisticated algorithms in.
-
We do the 'Boring' stuff?
- data management made both front cover of Nature and NY Times
- but in Nature article, work of many groups (incl. ours) was ignored
-
Being 'open' : Open Source and Open Community
- Appreciation from the community for our process
- But cF. above, we don't even know internally where the examples are
- Some of the messages just aren't getting out there.
- Quibbles with some of the stuff around the edges
- "Product owner" (horrified; defined externally & unclear what it is)
-
Concerned about labels, but not clear how to work in our environment
- Chris: if we aren't going to make our own process, we'll have to adapt to some process
- cF. trying to test permissions, didn't know what to look at
-
Specifications
- Need to push XML examples (Jason is working with Andrew on this, but we all need to work at it)
- Standard image
-
Bio-formats
- Melissa is getting hit pretty hard from us and the community
- What's missing? Scoping. Still lots of formats out there.
- ImageJ (largest consumer of bio-formats)
- Responsibility to provide what people want.
- There are non-biological folks using BF (geological stuff for example). What are we missing?
-
Architecture
- Powerful, flexible.
-
Analysis
- cF. measurement
- Doing more. using backend.
-
Annotation
- Time for us to go back to what this means
- Number of people using tagging is very small (but yes PDI)
- Still aren't getting much value.
- Something else to go to Paris with
- Using ontologies / structured descriptions for defining an image (machine readable)
-
What can it do
- Warning: post-4.2
- Examples. Needed! * There's a difference between having something doable and showing how to do it. * Going to have to put some time into these examples!
- 3 different examples: FRAP, HCS, EM * Showing the commonality * How we support all. * Concept that's non-trivial
- Simplified API * Calvinistic approach: it is complicated * Brings us back to the examples
- Storing analytic results * Currently have two ways for storing this: Structured annotation & OMERO.tables * Experience suggests that one size does not fit all. Perhaps this approach is good. * We need to go out to the community and find out what their use cases are.
-
Image Library for 'The Cell' (an Application)
- Janet Iwasa, great specs, mockups, etc.
- coded by Carlos
- work on the tools for real annotations
- automatism for large scale import
-
OME & Glencoe
- admitting the difficulties of split team
- cool tech for customers
- people are asking for more
- amounts to testing
- keeping the communication channels open
-
cost of breaking down isolation around the commercial projects, around the devs
- but there are also benefits
- what's the best way to deal with this?
-
Jean-Marie: hybrid-graffle for maintaining the commercial work?
- Chris: already an issue: server, client, bioformats, ...
- Andrew: commercially is it possible to have customer names, task items on the trac? * Jason: often ok, but there are people who would (needlessly) get upset
- Chris: noting every customer has own trac, would rather have a login for Glencoe tracs to make it visible
- Ola: bigger issue is what we're doing over several months
- Jason: at a high-level, this will also happen with the application developers
- Ola: something like "end of next month have to upgrade viewer, and it's 20% OMERO"
- Jean-Marie: building bridges, having larger group discussing issues so we can reuse * e.g. the schema for ASCB and its relation to protocol editor * worried that each time it is a one-off * what have we / can we learn from each of the applications? * Chris: certainly true for the metadata
AOCB
- Agilo: holding off for advanced functionality for Thursday
- QA tracking: @Adding items to the agenda (and move them to tickets!)
Action Items:
- Wlll/Andrew: set up a page where everyone can gather their examples (after reviewing existing pages).
- Scott: try to monitor everyone's usage in Agilo.
- Brian: confirm that the Agilo trac links in google still work.
- Jason: need someone to go see Tom about his Gel's!