[Edserplo_users] Major update - Analysis page now working

[Steve Loch]

1) Bug 572. The CTDPage now identifies the casts correctly using the #tt option

2) Bug 923. The XY-Plot can now manage timeless series (e.g.CTDs).  See below. Courtesy of JOSS.

3) Bugs 893,928.  The Analysis page can now display the set of constituents in a properly synchronised manner. Use of help no longer locks the program and indicators of analysis selection are now present. Numeric KP  7,4 and 9,6 will pan the analysis and constituent selection. The rapid panning through the use of the middle mouse button is still supported. Also you can use the latest analysis to revise the tide and residuals on the Timeseries page. The analyses can be stored in, and retrieved from,  Oracle. It is now possible to use ETA tables as an alternative to the POLAP database (see below). The associated controls are now in line with the Edteva manual. Almost all of this is courtesy of MOIF.

4) Bug 925. The incorrect write toggle on the Series page has been addressed (related to BUDS series).

5) Bug 899. Tutorial not working. The reason was that the associated files were not included in the JAR file. I'm putting in time fleshing out and correcting the entries. This will go on for the next month or so as I add, edit, extend and revise  elements of the old user guides  within the new tutorial system.

6) Jonathan has corrected the messaging problem on the Output page which, given its longevity, is much appreciated by Linux users like me.

7) Bug 896(2). TASK Format. The cycle flag was incorrect on output.

---------------------------

There is an option on the XY-Plot page to provide a colour ramp mapped to time. This is toggled through the 't' option.  This has been extended to cover timeless series where (implicit) cycle number is used instead. Should there be a mixture of timed and timeless series you will be asked to decide on what  display options you want to use.  

PS I have just tried this out on AMT17 and I wasn't impressed by the axis scaling for T v S - so I think there is further work to do.

-----------------------------

Edteva stuff: AnalysisPage

Processing data for which there was no set of constituents within DLB's database was a problem because unless there was such an analysis it was impossible to generate (via the Porch mechanism) the tide and more particularly the residuals. This has a particular resonance for external users of Edserplo who will not be able to access DLB's database. Luckily we can store analyses in the ETA Oracle tables. Thus the procedure for working up an unknown port is to do a tidal analysis and store it in Oracle  - it doesn't have to be particularly good. Restart Edserplo but this time include a reference to the Porch files in the driver and you will then have the residuals and tide. Refinements can then be made to the analysis, if desired, and the results replotted as in the Edteva manner. 

Steve