Personal tools
  • We're Hiring!

You are here: Home Community Minutes Conference Calls 2007 2007-09-06 Project-wide 3:45 BST

2007-09-06 Project-wide 3:45 BST

Attending: Alexa, Andrew, Brian, Catriona, Curtis, Donald, Iggie, Jason, Jean-Marie, Josh, Melissa, Will, Lars (Homburg)

Agenda

  • Running the agenda
  • Data Model: changes and updates
    • http://www.OME-XML.org
    • Follow up on recent OME-nitpick email thread-- what is mandatory
    • Back to minimum specs and levels: see email thread forwarded to list.
  • http://openmicroscopy.org Updates
  • Bio-Formats: status update
  • OME-TIFF: status update
  • OME 2.6.1: Status update
  • OMERO: status update
    • server
    • importer
    • insight
    • webadmin
  • Review of who is doing what
  • Other


Notes

  • Long conversation on the XML spec and the modular component concept. Everyone seems to be on board with the proposal to go with Andrew's original email proposal for the spec.
    • All parts of the instrument are optional expect for the microscope.
    • All bits of the experiment are optional except ID - with the caviet that there must be one of first, last, email, omeuser present.
  • the Email on OME-NITPICK is a good example of what can be done with the OME Spec. Not a "minimal set" by many standards (and therefore it won't be adopted as such), but a good example of how it can be customized to meet a specific need.
  • Melissa adding color index information to bio-formats so we can expand what can be displayed and imported.
    • Jason: would be nice to concentrate on non-index based full data formats as these are quantitative and therefore most valuable.
  • The online OME-TIFF documents are being updated by Curtis for clarity.
  • Importer should have regression test and display missing metadata model information when importing in 2.2
  • Lots of nice insight changes. Reduced number of windows, smart folders, copying images across renderings, history. Tagging system "proof of concept".
  • Ola has added some of the features suggested by Hamburg as well as password expirations. Focus next week on having rendering definition copy included.
  • Andrew has added more feedback to the XML Validator which should be tested and up for testing soon.

Action Items

  • Jason: Record a new movie!
  • Ola & Brian: Get movies up on the web for people to view

Notes II


  • Trying to get Lars connected
  • Everyone can update Agendas
  • Data Model
    • ome-xml.org becoming pretty useful
    • cat: Direction to idiot's starting point from all websites ("Cute and cheerful")
    • curtis' "developer person"/"user person" is good
    • Improvision file
      • Andrew tried to make a file from their point of view
      • In order to put in magnification, had to provide an awful lot
      • Does instrument/experimenter make sense when certain things are optional?
      • Ilya:
        • Magnification without storing anything else about objective makes sense
        • Id of Experimenter only doesn't make sense; id is usually only unique to file
        • Problem with all three examples is that the Image wasn't attached to anything
        • Ilya: There's a mechanism built in to link an Image to Experimenters via LSID
        • Andrew: Interesting, but it seems that just putting in the first and last name is useful
        • Ilya: "optional" ~ how does the system operate when that things just not there. What does it look like with _everything_ optional?
        • Cat: "To be used and to be useful" ~ doesn't always give you the most technical solutions. Does that require "dumbing down'?
        • Andrew: Schema+combined with minimum spec for valid file/valid file with full instrument/... is what we need to look for?
        • Ilya: concrete - (1) magnification yes/no? (2) detector everything was in their string. Do they parse it?
        • Jason: they're trying to represent their info; lots about detector but nothing on light source, because they don't need to differentiate
        • Ilya: We already duplicate some things (contrast ~ filter, fluor<-excitation, wavelengths directly in channel)
          • Let's do it with pixel size/magnification, too.
          • Andrew: But then users don't know what to use
          • Ilya: It's a pattern already used. Do we need to get rid of that too? For now, up to the reader to know what to do.
          • Andrew: Difference is, they seem obviously different values, but magnification is the same value in two spots
          • Josh: Push the check for two identical magnification values into the validator
          • Ilya: Can also be done in documentation. But it has to be read.
          • Jason: We've run into this so many times. People don't want to spend lots of time in this. Examples. And if you have to have 50 examples
          • Curtis: 50 isn't necessarily too many. But, best thing to do is have code.
          • Ilya: usability v. full-specification. We have to draw the line somewhere
          • Curtis
            • slippery slope to say that a change for this file format is caving in
            • have to keep looking at specifc issues: magnification, ...
          • Ilya: two proposals - (a) unstructured or (b) parse your strings
          • Andrew: developer has info on objective, but zero on light source. Prohibitive.
          • Josh: But another developer might break on null light source. Breakage.
          • Andrew: Just ask levels, then say "Can't open that file"
          • Ilya: Not really levels, but specifications
          • Ilya: Let's say you have instruments with all components optional. Then next level, what's optional?
          • Andrew: The line I had made was at the instrument level. Objective, light source, etc. have to be completel. Instrument had to have model number, but could then assume default.
          • Ilya: Not really against that if it had name or description. Just ID is pointless. The user should be able to look at it and connect it to the thing in lab.
          • Andrew: Instrument has microscope which has manufacture and model number
          • Josh: Exactly. The line in the sand is what the user can understand.
          • Ilya: That's what I'm advocating. If optionals are left null, then there still has to be some value to the user.
          • Jason: That's where we were coming from. We all agree.
          • Are we doing what Andrew's suggesting? Instrument exists->give microscope element. Others are optional. Let's do that.
          • For Experimenter, we could say "You have to use a LSID" or "You should, but here's the rope (don't have to)"
          • Andrew: Weird to enforce LSID for experimenter and no where else.
          • Curtis: Better to not make the change to requiring LSIDs now.
          • Andrew: one of firstname|lastname|omename <-- validator check
  • Trademark issues. Tabled.
  • openmicroscopy.org
    • Server documentation in central place. Clearly referenced.
    • Critical to get people in without getting lost.
    • UI team is looking at it. Mocks will be put up.
  • Bioformats
    • Melissa working on indexed color model. Complaints from ImageJ plugin users. Need converter methods on bioformats. Omero can use either.
    • Jason: don't currently use it, and based on quantitative data p.o.v. wouldn't make it a priority. But understand that need.
    • Curtis: Only for performance.
    • Discussion offline on priorities. Brian had sent email to Melissa.
    • Note: At Janelia Farm gave props to bio-formats. Kudos. Nice endorsement.
    • OME-Tiff?
      • Commandline tools being produced for OME-Tiff. Schema Validation of XML. Etc.
      • Working on cleaning up website. Old documentation. (missing creation date)
      • Someone made something *very* minimum, just for defining image. Need to update docs.
      • Jason: Good to hear, thanks. Pages/docs we provide are *so* important, perhaps also unique.
  • 2.6.1
    • Private email. Fixing issues with linux (Working around Gnu Java)
    • Done any second now.
  • OMERO
    • server
      • Things looking good. Josh pushed lots of stuff. WebAdmin had lots of build changes. Ola, s'ok?
    • importer
      • Melissa more file formats
      • Brian working on regression testing
      • Trying to get XML validator before 2.2
      • Curtis: XML validator? Brian: OME-Metadata Model validator...which files are missing from OMEROMetadataStore.
    • client (a little b3 mixed in)
      • looking at lots of cool things
      • 2.2 is copying rendering settings across images
      • mostly in place. Few more methods, but if not all, can still release
      • working on smart folder
      • getImages is slow. @Will contact Josh to work on it. http://trac.openmicroscopy.org.uk/omero/ticket/657 ?
    • webadmin
      • suggestions from homburg
      • working on expiring of passwords
  • Movies
    • lots of movies. some with compression. some without.
    • jason's movie has no soundtrack.
    • can't go Ola's account. Using Brian's.

  • Push Trac meeting to next week.

Document Actions