2014-06-17 Tuesday Team Meeting
Attending: Dundee: Kenny, Simon, Chris, Gus, Andrew, Balaji, Dominik, Emil, Mark, Petr, Graeme, Blazej, Ola, Roger, Colin, June, Jason, Simone Remote: Josh, Ian, Andreas, Mark H, Melissa, Seb S, Seb B, Sunil Kumar, Kelli, Liza, Douglas
Agenda - 2:30pm Start
Accepting minutes from last meeting
- accepted
Project Status (15 minutes max)
Mainline (5 minutes) - 14:31 UK
- Josh: search card is main one in flight
- Josh: looking at speeding up back end but won’t be making major changes now so people won’t need to re-index their database
- Balaji: is going to look at the nightshade ROIs with mtools
- Melissa: two tickets solved with open PR, other tickets remain
- Josh: import ticket needs to be picked up
- Josh: milestone 5.0.3 still has 115 tickets open, 74 bugs.
-
Jason: 5.0.3 goal is to make search work with documentation, server, clients much as they are now
- also good ideas from Will, Gus, Douglas, etc. for future work for 5.1
- Sebastien/DOCS: with Helen away, need to take this on.
Glencoe Update (Chris) (5 minutes) - 14:39 UK
- Andreas has been fixing bugs
- also work with Douglas on tagging, jsTree
- Carlos will be away for a while
Consortium Update (5 minutes) - 14:41 UK
- Jason: Simone Leo from CRS4 is joining us in Dundee for the summer
- Sebastien B: will be trying to teach git to Texans
-
Douglas: working with Chris on JavaScript third-party library version updates
- jsTree will be wholly JSON and perform faster
- Jason: targeting 5.1? Douglas: yes, too much for 5.0.3
-
Jason: impact on others working on web plugins, etc.?
- Douglas: can use library to assist backward compatibility for jQuery
- Chris: generally need to review how we are using jQuery and jQuery UI
- Jason: changes needed by Douglas may also benefit others
OPT and OMERO - Ian Munro and Sunil Kumar from Imperial (20-25 minutes plus 15 mins questions max) - 14:47 UK
-
problem of acquiring images of sample at different angles, then needing to extract sinograms from those planes where each needs data from every plane
- original images have camera aimed perpendicular to axis of rotation, but users will request planes that are perpendicular to the axis of rotation
- Chris/Jason: may be most easily solved with more storage in hardware rather than effort in optimizing software
-
Sunil: Yuriy wonders if sinograms can be annotated with a corresponding more-viewable image easily accessed from client
- Chris: strips along other dimensions available but puts greater burden on server
-
Sunil: possible to show a thumbnail that isn’t a XY plane from the image?
- Chris: not at present, but contributions welcome
-
Ian: could custom Bio-Formats reader help?
- Chris: need to store as scan-line or tile, then reading the other way will be slow
- Sunil: scan-line may make most sense given sinograms
- Chris: reader examples from Bio-Formats open bytes on web, showing extraction of parts of planes
-
Roger: currently limited by dimension ordering always starting X, Y
-
Jason: could perhaps be okay to be less general depending on conventions accepted in tomography
- Sunil: X-ray CT would be analogous, they may well also use sinograms
-
Sunil: will cause others to ask around about conventions
- Jason: also note that people are doing micro CT and HREM
-
Jason: could perhaps be okay to be less general depending on conventions accepted in tomography
-
Jason: what metadata do you need to store?
-
Sunil: angular spacing is important for reconstruction algorithm, may not be equally stepped
-
Roger: use modulo Z as rotation? is okay even if Z is the physical Z dimension
- Sunil: and rotation will be in Y Roger: but no modulo Y yet
-
Chris: clients probably won’t be great at handling using modulo like this
- Melissa: be cautious though Bio-Formats may not work if modulo annotations are misused
- Chris: so use Z as angle of rotation
-
Roger: use modulo Z as rotation? is okay even if Z is the physical Z dimension
-
Sunil: angular spacing is important for reconstruction algorithm, may not be equally stepped
-
problem of acquiring images of sample at different angles, then needing to extract sinograms from those planes where each needs data from every plane
AOB (5 mins max - technical discussions should be highlighted to relevant people and rescheduled) - 15:58 UK
- Jason: Seating: OME team will be in “goldfish bowl” and in nearby seating in larger area, Java people from other groups nearby
-
Paris Video Content (Dundee only)
- Chris: video is poor, Emil has experimented with using audio with presentation slides
- Jason: post the video chopped into talks, make the presentation slides available alongside?
-
Jason: experiment with Jason’s talk, synchronizing slide video with talk audio, see if anybody looks at it
- Emil: already half-done anyway
- Jason: cautious about engineering media of other people’s work
- Andrew: will check if anybody downloads the speakers' slides PDFs
Done 16:18 UK