[ome-devel] Removing protocol information from experiment XML files in OMERO.editor

[Will Moore]

Hi Adrian,

 Thanks for your feedback. 

You're asking about a tricky issue that's taken a bit of discussion and various different solutions.

Originally, in the very first implementation of the OMERO.editor, I did exactly as you've suggested and kept 'protocol' files and 'experiment' files separate.

The trouble is keeping them in sync:
 - It's difficult to ensure that you always have the Protocol file for a given Experiment. (not deleted, renamed lost etc)
 - You have to be sure that the Protocol hasn't been modified at all. Has to be "locked" as soon as you create an Experiment. Very tricky to "unlock" in the future because you'd have to keep track of (and update) all the Experiments that came from it.

Also, at least when I was testing it with users here, I found that they didn't really re-use a Protocol very much. There was always a small change they wanted to make that was outside of the experimental parameters they'd defined. I tried to minimise the need for this by adding "notes" to each step, but it was still a problem.

This will obviously vary depending on the type of experiments you're doing. E.g. high throughput techniques, or a lab where you're doing very similar experiments all the time might not need to change the Protocol very often.
It would also give you a lot more querying / searching power if you knew you had a lot of Experiments that used exactly the same Protocol (E.g. search by parameters etc).
If you were working on the server (not standalone Editor), you could store the values themselves in HDF tables, with columns defined by a Protocol's parameters.

This is analogous to the ISA-TAB format: http://isatab.sourceforge.net/format.html
Unfortunately we don't have the resources / funding to take the Editor in this direction at the moment.

NB: It is possible to build a custom search indexer for OMERO.editor files http://www.openmicroscopy.org/community/viewtopic.php?f=6&t=553 to help with querying them.


 Regards,

   Will. 



On 25 Mar 2013, at 14:00, Adrian Jacobson wrote:

> Dear OMERO developers,
> 
> The OMERO.editor is a great feature, but it might make sense to reduce the amount of information in the XML files of experiments.
> Currently OMERO stores the content of the corresponding protocol plus the values for the variables within it in the experiment XML files. This is redundant, because one already possesses the protocol.
> It might be more elegant to store just the value of the variables in an experiment in the corresponding XML file.
> 
> Regards,
> Adrian
> _______________________________________________
> ome-devel mailing list
> ome-devel at lists.openmicroscopy.org.uk
> http://lists.openmicroscopy.org.uk/mailman/listinfo/ome-devel

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openmicroscopy.org.uk/pipermail/ome-devel/attachments/20130326/b4f7b62c/attachment-0001.html>