[Edserplo_users] Update, projected progress and tidal analysis
Steve Loch
sgl at bodc.ac.uk
Mon Jul 23 14:18:50 BST 2007
Bugs/enhancements resolved
711 - Input of analyses stored in Oracle tables now supported
712 - In certain circumstances the settings file name not carried through to Series or Port page from invocation
716 - Unable to save settings file
720 - Edit mode flagging problem with non-blank flags
719 - Problems with losing currency on series when moving between pages;
also interference between pop-up menu and datacycle window
[721 - Direction of panning - we're investigating best way to deal with swapped mouse buttons]
722 - Oracle access files on Windows: can now be found in current directory
723 - Font on the AnalysisPage has been improved - however the production presentation is "damaged". Hopefully sorted tomorrow.
730 - Pointer and datacycle window not in step - some lesser used pages still need attention.
Most of this work is down to Muhammad. He is now, or shortly will be, working on the update of the AnalysisPage buffer. He then has to code the recalculation of residuals for user-selectable tidal analyses. Once that's done and I've resolved a problem with the output of ModelResidual format we should be at Edserplo 2.0. The User Guide/Tutorial needs to be worked on too, over the autumn, as part of our (hoped-for) KT customisation process.
We also need to build a comparator to check that sets of statistics held in identical tables in two different Oracle IDs match reasonably - one produced by Edteva and the other by Edserplo.
---------------
Tidal Analysis now works with an inclusive flag mask rather than an exclusive flag mask. This change was associated with setting up some tidal analysis demonstration code (hardwired for TASK format and QXF) in Matlab. This will be worked on in the course of time to provide an automated QC facility for shelf CMD (comparing tidal model output - allows a check on timing). The Matlab is a wrapper for Java TIRA which is inside Edserplo.
See
>>help demoTidalAnalysis.
Steve
More information about the Edserplo_users
mailing list