2009-08-28 OME Group Conference Call
Attending:
Agenda
Beatson summary
ASCB quick update
update on XSLT fun (Group A)
-
See SchemaTranslation
- Next phase: schema to transform. Strategy
-
See SchemaTranslation
short term multi-image support - CA only solution ok?
QA
- status
- todos
4.1 schedule
- Code-gen estimate
- Fixed-time or fixed-scope?
-
Discussion list (need to make comprehensive!)
- code-gen --> export
- permissions
- QA
- dropbox
- hcs in insight/importer/matlab/cellprofiler
- opengl
- measurement tool updates, ...
- webstart/web deployment
- metadata completion for confocals (Leica SP2, SP5, Olympus, LIF)
- synchronization between bio-formats and omero
forums and emails
Notes
beatson visit
- had already installed
- trying to figure out how to bring OMERO to the users
- imagej analysis of interest
- interest in micro manager ("cast off scopes")
- good source of (diversified) feedback
ascb
- blurb submitted.
- booking hotels by the end of the month (sept)
- who's going?
xslt
- all transformations done (2008--> working copy)
- page of questions (SchemaTranslation)
- all stylesheets assume single pixels
- ...few minutes missed...
-
only one range in whole schema (constraints)
- was testing if it works
-
filters
- filterset is currently used as a container, but may not be
- avoiding modelling light path in entirety?
- already modelling more than most tell us
- who's using them (other than loci)?
- requirement: isn't reconstructing light but having list of things.
- don't let the order be a stopper (re: xslt)
-
analysis module
- filtering them from the conversion?
- supporting STD for another version (e.g. 2010-09)
- next phase: 2003-->2008-09 done by Andrew
-
organizing www.ome-xml.org (DeveloperInfo, etc.)
- reducing amount of info on front page
short-term multi-image support
- sent 2 proposals: STD & SA
- using 2003FC anyway.
- requiring an upgrade for support from us?
-
get a ticket on trac?
- everything should be a ticket
- See new "proposal" component
- anyone participating should upgrade
QA
- 80-90% done
- need validator
- josh generates test data
-
pointing all clients at QA?
- yes, need versions to work against QA for testing
-
missing testengine results
- Brian: should be straight-forward
- on file-by-file basis.
- need to agree on message
-
way for demo server accounts?
- added a register user page
- any barrier to deploying the above? no.
- need testing. when ready for testing?
- integrating validator
-
collecting metrics (from hudson runs)
- imagine for ascb visual
4.1
date from last week: Oct. 15
code-gen: on track to have stylesheets done
- code-gen can start
- but not clear how long it will take.
- should be able to do the simple image export
codegen:
- not generating the mapping files or the walkers
- but having the files in place to export to XML
-
Donald: but where's the end to this?
- right now everyone is steeped in the XML
- can do things more efficiently now.
- Chris: can't afford to break the OMERO API now.
- won't help us with the import stability
- Donald: wasn't the point of group B codegen?
- Chris: xslt solves > 1/2 the problem. only supporting single schema
- Jason: current release isn't usable by 75% of the market
-
Ever more things are code against our API
- Need notifications by when?
-
Josh: breaking everything as 5.0?
- hard to manage branches
-
Jean-Marie: data-in/data-out was the task
- also companion file with all the metadata
- middle ground to make lots happy.
- 6 weeks, freeze in 4...
-
Donald: code-gen to prevent breaking API
- point: getting distracted, don't want to forget all this
- Jean-Marie: slow it down, not like XSLT where we do nothing else
- doing whole rework over 6 months? etc.
-
Jason: get through user-driven list, and then decide on how to do next steps
- moving to NDIM structure?
fixed-time / fixed-scope
- time.
- missing some targets from Paris (data duplication)
- metadata completion is more non-incorrectness
- cF. companion file
Features
- A: export to OME-TIFF
- A: metadata in/out
- A: QA
-
A: opengl : only focuses on viewer
- realistic for time frame
-
A: hcs : in general
- insight / importer * visualization * small set of hcs files * largely done for visualization * lacking rois * adjustment of measurement tool. only thing for this stage * rois are largely preview in 4.1 * loading from server, getting table, etc.
- matlab * available by default (java)
- C: cellprofiler * a bit too much? * DEMO for ASCB? 4.1.1? No, more 4.2. Something new introduced. * Extending ImageJ at the same time * 4.1 doesn't have good roi support; want to do it well.
-
B: dropbox
- lots of small things, hopefully out of the way by end of next last week
- mounts ghislain's email: what's possible
- main problem: how to cope with mount disappearing/reconnecting
- Chris: isn't a local filesystem, won't be able to watch
- would need to setup a remote fs instance
- Josh: that's a second node. no problem.
- need work on documentation. clear, illustrated, ...
- Colin: should be okay for end of sept. tied to mount systems. * Chris: either local or remote. * Jason: needed before then. 2 weeks. * stayed focus.
-
C: permissions ~ lots of testing! needed by mid-september
- Broke more stuff than we thought
- C: webstart/web deployment
-
FLIM
- DEMO'ing at best. keep it going, but not for releases.
personnel
- Donald: in codegen, opengl, measurement. Etc.
- Everyone come back with estimates please.
- Need calendar of what by when for next 4-5 weeks.
groups
- roi:
- permissions: brian on GUI
next phase
-
cF. FLIM email-->proposal-->haven't followed up (4.2 is coming)
- Proposal(s) by end of next week. ("punt" on table.)
- "Provisional" for February
- Going to be a challenge.
-
cF. FLIM email-->proposal-->haven't followed up (4.2 is coming)
all jars: release should be synchronized
- regularly synchronizing already.
- on every "stable" release?
forums/emails
- need "Hello World" for coding on Insight
- how much configuration do we permit?