2015-01-27 Tuesday Team Meeting
Dundee: Simon, June, Jean-Marie, Ola, Will, Petr, Kenny, Dominik, Jason, Helen, Mark, Remote: Gus, Seb B, Josh, Andrew, Melissa, Kellie, Colin, Ian, Andreas, Roger, Chris, Emil,
Agenda - 2:30pm Start
Accepting minutes from last meeting Accepted
Project Timelines (2-3 minutes each)
Core (incl. CI & QA)
-
5.0.7 outline from nitpick
- see: recent nitpick e-mails. Tickets - 201 left.
- RC1 outline. Maybe push back by a day?
- Trello 5.0 rebase card looking good.
- Scenarios for Downloads and Client packaging need updating.
- Testing files need to be stored somewhere, E.g. unicode, float/double.
- ROI testing plus other small testing todos.
- We’ll try to merge open PRs ASAP and get an RC1 build.
- Few PRs opened today - lots to review tomorrow.
-
Units: next step is likely release of schema
- Any blockers?
-
Rendering: investigation into ChannelFiller and friends. API changes to be decided.
- Josh - found a way to use lookup tables without too many other major changes.
- Client devs will be able to have a look soon - see trello card.
- https://trello.com/b/Uoadzs9k/formats, will contain an explanation.
-
libbioformats: OME-XML now roundtrips to and from the model objects/metadata store
- model temporarily downgraded to 2013-06 for 5.1.0, and will be upgraded to the latest version after the release
- Currently working on OME-TIFF reading
- zip files not linking correctly - maybe use tar files
-
5.0.7 outline from nitpick
Glencoe Update (Chris) - Last few 5.0.7 PRs - floating point rendering etc
Virtual Microscope/teaching (Ola)
- upgrade successful
FLIMfit/OPT (Ian)
- WIN and Mac builds both now green with Matlab 2014b
- Progress report for Friday so other work on-hold to next week
Bio-Formats large file sets (Sebastien)
- Windows testing: on hold, waiting for resources on maxquant (libraries need building) Documentation on building etc could be improved.
- Woolz testing: on-going, goals as follows 1) build and use Woolz libraries in our CI server deployment 2) get a status of the Woolz file format for 5.0.6/5.1.0-m3 (follow-up of the file format testing)
Other consortium timelines (5 minutes)
- OMERO.figure: - started updating testing scenarios.
- Jason: need to fill out trello card with more details on our apps.
- Jason: Virtual Microscopes in more places!
- EBI repo new hires starting soon. Need to setup hardware etc.
AOB (5 mins max - technical discussions should be highlighted to relevant people and rescheduled)
Andrew - Working with the data model (20-25 minutes plus 15 mins questions max)
- http://downloads.openmicroscopy.org/presentations/2015/SchemaChange-APatterson/SchemaProcessTalk-2015-01-27.pdf
-
Overview - most changes happen in BF
- Designing changes. Ideas -> Diagrams -> Discuss doc -> Review & repeat untill all happy.
-
Need to consider OMERO DB implementation to check performance, validity etc.
- Prepare data model for change.
- All work is in a breaking build
- See our docs on Versioning the data model. Mostly just do major releases
- Update Schema namespaces and build values for ant & maven
- Move existing schema to new n+1 version (retains full history)
- Then can create a new n version for reference
-
Publishing interim data model changes
- Schemas must be live online for tests to pass
-
Full data model release
- Schema namespace changes etc.
-
Making changes to model
- Make a sample file of new change to get early feedback from others
- Building schema changes (see presentation pdf above)
-
Changes to transforms
- Upgrade & downgrade transforms - Update Insight catalog file.
-
OMERO requires some work before changes move out of breaking queue.
- Always start with clean
- Checkout breaking BF branch into OMERO
- Update mapping files
- Units - update ome-utils repo
-
Documentation
- Diagrams, png/svg. Made with Oxygen
- Sample files
- Jason: do we know how many users look at diagrams? No.
- Mark: failures will result in OMEROMetadata store build failures? (Yes - API changes)
- Mark: would be nice to have specific locations of items mentioned in presentation.