Monday, December 6, 2010
Enterprise Architect (EA) has played an important role in the EDNA project from its start till now. EA is a very powerful tool for UML modelling and has served it purpose well. However, the fact that EA is a commercial tool, only runs (well) on Windows and does not support the XMI standard very well has made us to look elsewhere for an EDNA UML modeller.
Since the majority of the EDNA developers use the Eclipse IDE for their developments it was felt that the ideal situation would be to have an Eclipse-based UML modeller for the EDNA developments. Peter Keller (Global Phasing), who has investigated many different solutions, recommended us to try the Topcased UML Editor. This Eclipse-based modeller is open-source and can easily be incorporated into an Eclipse installation.
Unlike EA, the Topcased UML modeller does not create XSD (XML Schema Definition) files from an UML model. These XSD files are necessary for automatically produce the data-binding objects used by the EDNA plugins. On the request of the ESRF Karsten Thoms has produced an Eclipse plugin "UML2XSD" with which it is straight forward to transform an UML model into an XSD file.
Both the kernel and the MXv1 data models have been fully transformed into Topcased UML diagrams and the EA data models are deprecated. The idea is to then slowly migrate all existing EA UML models to Topcased UML models. This migration will soon be simplified thanks to further developments by Karsten Thoms.
For more information on how practically the Topcased UML modeller is used in EDNA see: Data Modeling with the Topcased UML Editor
Minutes from the EDNA full meeting, EDNA training slides and the Kernel Code Camp summary are now available online
On November 15th-16th an EDNA training was organised at the ESRF. The slides from this training are available: EDNA Introduction (PDF) and Edna Tutorial (PDF).
The summary from the last code camp which was dedicated to kernel developers is also available online: EDNA Kernel Code Camp November 2010 (PDF).
Tuesday, November 30, 2010
Using the EDNA framework the DIMPLE pipeline takes a PDB file and an MTZ file and uses software included in the CCP4 suite to automatically calculate a difference density map. The results of the pipeline are a map file, a set or refined coordinates and a small script to automatically start COOT.
The pipeline has been based on extensive discussions with a number of users and an initial proof-of-concept application developed by Graeme Winter, solve-o-matic. In collaboration with CCP4 (namely Ronan Keegan and formerly George Pelios) and with grateful guidance from Mark Basham, the proof of principle application has been developed into an EDNA framework application. We have deployed this pipeline on Diamond beamlines where we will hopefully iron out any further problems but we would be delighted to hear if anybody else tries to use the pipeline and how you get on! Work is already underway to improve some of the reporting and usability aspects.
The software is made available under the LGPL v3 license and is available from the EDNA SVN or drop us an email. We will distribute with future releases of the CCP4 package. More details can be found from an early presentation at: http://www.edna-site.org/wiki/images/3/30/DIMPLE_-_Difference_Map_Pipeline.pdf
Graeme, Ronan and Alun
Monday, August 30, 2010
Thursday, April 22, 2010
The areas which will be covered in detail are as follows, they are based on CCP4's first use of EDNA and also comments from DLS from the use of MXv1. This is not an exhaustive list, and any other kernel issues which could do with addressing will be welcomed and added before the meeting.
- Automatic unpacking of input objects
- Definition of methods inside plugins
- Allow use of state in plugins
- Define plugin types in data model
- Should input XML be validated
- EDNA run folder location (Possibly an MXv1 issue, but may run deeper)
The Agenda to try to address these issues is as follows (Times are UK) :
- 09:00 - Coffee + Setup Maratech etc.
- 09:30 - Discuss Kernel changes as stated above, and divide the documentation writing tasks as Olofs plan http://www.edna-site.org/joomla/index.php?option=com_wrapper&view=wrapper&Itemid=6 Should include a code review of some DIMPLE plugins as examples of where issues came from.
- 12:00 - Lunch
- 13:00 - Open session for kernel development
- 4:00 - Code Reviews
- 14:30 - Coffee
- 15:00 - Documentathon - try to get as much doc in place in 30 mins.
- 15:30 - Open Session for kernel development
- 16:30 - Code Reviews
- 17:00 - Close
- 09:00 - Plugin Development (i.e. try to use the new kernel changes in your own code, also the new docs)
- 10:00 - Documentathon - try to fill in any missed bits of doc
- 10:30 - Coffee
- 11:00 - Discussion over kernel changes made, and there implementations.
- 12:00 - Lunch
- 13:00 - Final Session for Kernel Developments
- 14:30 - Code reviews
- 15:00 - Coffee
- 15:30 - Closing discussions etc plus workshop outcomes.
The whole thing will be done using Maratech, and is really aimed for seasoned Python programmers who have written EDNA plugins and are happy to help with making kernel changes.
Friday, January 29, 2010
Friday, December 18, 2009
The schedule of the workshop will be more of an introductory level for new developers on the Monday then more of a code camp style where tasks for kernel developments and specific plugin developments (e.g. MXv1 and V2) will be undertaken in breakout groups for the remainder of the time.
We anticipate that the workshop will be very popular so expression of interest/commitment to attend early is recommended! We will try and organise accommodation at http://www.stfc.ac.uk/About/Conts/Find/Coseners/Introduction.aspx though all travel and accommodation expenses will be the responsibility of the attendee and their home institution.
Early expressions of interest would be very much appreciated! I hope you haven't all gone away for Christmas already, if you have, I hope your having a good time!