TrakEM2 0.3g released

[Albert Cardona]

Dear all,

TrakEM2 0.3g released with new features:

- database-independence: create either database-based or file system-based
projects. XML files for saving the project details, and images whereever they
exist in the file system.

- export database projects as an XML file and a folder of images, which can
then be opened as a database-independent project.

- semiautomated snapping of images: drag and drop images onto another and have
them be stiched together (only their position).

- freehand drawing of profiles (bezier curves) with the pencil tool: best of
both worlds. Draw fast, adjust fast.

- import grids and sequences of images with given overlaps, linking (or not)
images to each other to manage them as groups.

- no need of an XML template anymore: just cancel the dialog, edit later if
needed.

- editable XML template: the template tree can be customized on the fly. Also,
import works from .dtd files or the DOCTYPE declarations of other projects.

and many little bug fixes!

http://www.ini.unizh.ch/~acardona/trakem2.html 


This version needs heavy testing which I cannot do personally (sheer lack of
time). I would appreciate bug reports.

I am currently managing 8 Gb of images at the same time without any issues
(database project), montaging, dragging, adjusting images with ImageJ plugins
and modeling.

Remember: right-click is your friend.

Please enjoy!

Albert


--
Albert Cardona
Molecular Cell Developmental Biology
University of California Los Angeles
Tel [phone]
Programming: http://www.ini.unizh.ch/~acardona/trakem2.html
Research: http://www.mcdb.ucla.edu/Research/Hartenstein/

[Dimiter Prodanov]

>Hi Albert,

good to hear this from you. I was thinking also of adding
multiple database support in TrakEM2. If you are interested
I can try to port it to a MySQL backend.

best


_______________________________________________________________________
Dr Dimiter Prodanov, MD, Ph.D.

Neural Engineering Rehabilitation Laboratory
(Laboratoire de Génie de la Réhabilitation Neurale)
Département de Physiologie et Pharmacologie
Université catholique de Louvain
Avenue Hippocrate, 54
POBox UCL-5446 / B-1200 Bruxelles -Belgique-
Phone: [phone]
Fax: [phone]

http://www.md.ucl.ac.be/gren

[Albert Cardona]

Hi Dimiter,

> good to hear this from you. I was thinking also of adding
> multiple database support in TrakEM2. If you are interested
> I can try to port it to a MySQL backend.

Why not, almost everything database-dependent is clustered at the
ini.trakem2.persistence.DBLoader class, which you can copy verbatim and
replace a few lines of code (perhaps one could make an abstract DBLoader and
then create a PGLoader and a MyLoader and so on). At the moment there is
already an abstract class, the Loader, which both the DBLoader and the
FSLoader (file system loader) extend, and which handles the rolling memory
and the insertion of images, stacks and montages.

In the SQL language I have avoided any postgresql-specific idioms, except for
the PGpolygon class to store points. This class is also used as a convenience
in the Profile and Pipe classes of the ini.trakem2.display package, but I can
change it to a more portable array of points.

On MySQL: I tried a year ago version 4.0.* and it had performance issues
(read: very slow) when inserting large amounts of data (say, a layer with
about 0.5 GB of montaged images), which PostgreSQL 8.1.4 does not suffer
from. The situation may have changed now with MySQL 5, I haven't tested it.

It would help me Dimiter to have such a MySQL loader in that you may spot
performance bottlenecks (some I suspect, some I can't test). So just ask what
you need!

Albert