[ome-devel] Questions about initial development.

[Jason Swedlow]

On 6 Oct 2006, at 17:44, Curtis Rueden wrote:


>
> P.S. to Ilya: Thanks for the excellent and thorough summary of the
> state of the two systems. Your writeup is so informative and useful
> that I think it would beneficial to post it somewhere on the main web
> site, in an "OME and OMERO" section or some such. I would be willing
> to create such a page using your words plus a bit more background
> information (i.e., OME is mature, while OMERO is experimental). What
> do you think?

Agreed-- would be very useful.  Hopefully it is clear, based on my  
email and Ilya's.  Some comments:

-- OME is released & supported.  OMERO is experimental, not yet  
released, and not yet supported.  Quite frankly, we don't even yet  
know whether OMERO can work in a real intensive imaging environment  
(we are testing it now, in a facility that collects ~1 TB/week; yes,  
we are starting slowly).

-- While OMERO development has been ongoing for some time, it has not  
been included on the OME website, simply to avoid any possible  
confusion.  All presentations I am aware of also state this  
explicitly.  Personally, I'd be reluctant to change the OME site  
substantially-- OMERO just isn't ready for release.  We are certainly  
testing at our site and a couple of others.  Bottom line, people's  
time and data are too precious to subject them to unstable,  
unfinished code.  We have guinea pigs here at Dundee who are bearing  
the brunt of that experience ;-)!  

-- The idea of a replacement, or even a competition, is not at all  
correct.  The unfortunate thing is that the OME project is research--  
we are still trying to understand all the ins and outs of delivering  
high performance image informatics tools for research.  We've been at  
this for 6 years, in various forms, and the problem is not yet fully  
solved.  If it weren't this hard, it would have been done already.

-- Ilya's thoughts in terms of the interactivity between an instances  
of OMERO and OME servers are spot on.  In fact, we'd like to achieve  
a tighter integration.  The technical choices we made on OMERO  
reflect what we needed to do at the time to get the project developed  
as quickly as possible.  Ilya and I differ on this choice.  We don't  
disagree on the need to get the systems back to a unified data  
structure.  Most likely that will mean reworking OMERO.  However,  
OMERO has been built for this type of flexibility.  We will finish  
our third milestone set in the next 3 weeks (see http:// 
cvs.openmicroscopy.org.uk/tiki/tiki-index.php?page=Omero) and will  
then start designing our strategy for aligning the servers more  
closely.  

-- A reiteration of some of Ilya's points: the OME server provides  
functionality that is critical for large scale data analysis,  
especially in image classification (e.g., see http:// 
cvs.openmicroscopy.org.uk/tiki/tiki-index.php?page=Computing+Image 
+Signatures+in+OME) and an integration with MATLAB .  There are many  
others.  We use the OME server for large scale data analysis, mostly  
object segmentation (see http://www.openmicroscopy.org/howto/ 
FindSpots-v2.pdf).

In principle, placing such statements on the OME web site is fine.  I  
would prefer to wait a few more weeks til we have a finished M3 and  
can point interested parties towards that, and also have an  
articulated strategy for aligning the servers.  However, if there is  
reason to do this sooner, that is fine.

Cheers,

Jason


**************************
MSI/WTB Complex
The University of Dundee
Dow Street
Dundee  DD1 5EH
United Kingdom

phone (01382) 385819
Intl phone:  44 1382 385819
FAX   (01382) 388072
email: jason at lifesci.dundee.ac.uk

Lab Page: http://www.dundee.ac.uk/lifesciences/swedlow/
Open Microscopy Environment: http://openmicroscopy.org
**************************