Comments on Fieldbook V1 b29/04/11
Red entries have not been addressed up to 18/5/2011.
- You have not changed the version number. Mine is still called Prototype V1.0 – the same version as my previous comments dated 14th March. We MUST get on top of the version numbering problem!
- I managed to get Fieldbook to connect to my IVIS-CENTRAL-GMS and IVIS-TL1 by typing those DSNs into the fields on the Options page. (It would not read my ini file). It would be better if the application looked for the file called ICIS.INI on the Temp Directory at startup and connected to the databases indicated there. In this way, users with multiple crops can use the existing launcher to connect all tools to the correct databases whenever they change project. I know we will be making a ne Launcher and this might use a different method, but for now this is all we have and it is dead easy to program so we might as well do it. Furthermore, there are a couple of other settings which might be usefully read from a [IBFieldBook] section of the ini file – the path to a default directory where templates are stored and one to a directory where germplasm lists are stored (if files will be used to specify germplasm lists).
- I can get a list of germplasm lists from the database, but I cannot open any list. I cannot get a list of traits. This was resolved be adding field LLRECID to the local LISTDTA table and adding fields TraitGroup and Ontology to the local TRAIT table (also made TRAITDESC memo instead of Text but not sure if this is necessary).
- Some changes to the introduction. Please replace all the text on the opening page with:
Please try the available functions and give comments and suggestions on the user interface. At present the only applications on the workbench are the trial fieldbook wizard and trial Editor.
The Workbench opens with an assumed active 'breeding project' and an empty list of 'studies' in the StudyExplorer. A 'breeding project' is usally formed around a technical objective such as 'Breeding drought tolerant rice for the Mekong region'. The project defines the databases to be connected and the users who have access. A 'study' is unit of work within a project which generally has a specific objective such as '2011 multi-site evaluation of promising lines' and provides the structure to contain data sets relevant to that objective.
The fieldbook application requires an active 'study' so the first thing to do is create a study by clicking the 'New Study' button (or right-clicking the study header in the StudyExplorer). Once you have a study highlighted in the study browser you can create a trial by clicking the 'New Trial' button (or right-clicking on the study name in the StudyExplorer and selecting 'Create New Trial'). These actions will open the 'Create Trial Wizard' which will lead you through eight steps required to construct a Trial Fieldbook by merging together a Trial Template, a Germplasm List and a Field Design.
- Note, the fieldbook should indicate somewhere what project is open. I suggest we print the field INSTLN.IDESC next to the title IBFIELDBOOK on the blue strip at the top of the screen. (This will also encourage users to put sensible descriptions in that field). (See point 2 of IB Fieldbook Prototype V 1.0.doc which points to Point 4 of IB Filedbook Prototype V3.doc and still needs to be addressed.)
- I still don't think the File menu should be called the File menu and I have suggested 'Study' menu (see Point 5 of IB Filedbook Prototype V3.doc and tell me what you think).
- All the tool tips have vanished. I know we had a plan to develop a help system, but something is needed urgently.
- Testing template IBFB1-Generic MultiFactorialMET.xls, the study condition is the PI ID and PI NAME with property PERSON and SCALEs DBID and DBCV. I correctly get a pick list but from the canned data of the alpha version. This should read the PERSONS table from the local and central databases.
- The spin box for Trial Instances on the 'Study Conditions' wizard page still says 'Study Instances'. We agreed it should say 'Trial Instances' and contain the name of the Trial Instance factor in parenthesis. Eg Trial Instances (TRIAL) in the case of template IBFB1-Generic MultiFactorialMET.xls.
- On the Trial Conditions page of the wizard for template IBFB1-Generic MultiFactorialMET.xls you should fill in the SITE ID and SITE NAME with lookups from LOCATION, but only the DBID is filled (with canned values at the moment). (See point 8 of IB Fieldbook Prototype V 1.0.doc – still needs fixing).
- I cannot load a germplasm list from the database. I see the set of lists but nothing gets loaded when I select one. (Resolved – see point 3 above) I must use the germplasm files supplied. When I do, I still do not see the ENTRY number in the table of germplasm displayed. (See point 9 of IB Fieldbook Prototype V 1.0.doc which points to point 9 of IB Filedbook Prototype V3.doc and still needs fixing).
- BLK is still listed among the Other Treatment Factors and this is a bug.( See point 10 of IB Fieldbook Prototype V 1.0.doc)
- On the trial Constants to be Measured page of the wizard the Description text box does not show the description of the highlighted constant trait.
- In the Trial Editor the Entries tab should be called Germplasm Entries and it should precede the Other Treatment Factors tab. (See point 14 of IB Fieldbook Prototype V 1.0.doc which points to point 13 of IB Filedbook Prototype V3.doc and still needs fixing).
- On the Other Treatment Factors tab of the Trial Editor you should not have the 'Value' column. (See point 11 and 13 of IB Fieldbook Prototype V 1.0.doc which points to points 10 and 12 of IB Filedbook Prototype V3.doc and to point 2 of IB Fieldbook V2 Issues.doc and to point 10 of Comments on IB Fieldbook V2 201011082200.doc and still needs fixing).
- On the Design tab of the trial editor you should note points 12 and 15 of IB Fieldbook Prototype V 1.0.doc.
- On the Measurements tab of the Trail Editor, the design column headings Trial, Rep Block, SubBlock, Plot and Entry should be the names of the labels from the template with properties TRIAL INSTANCE, REPLICATION, BLOCK, PLOT IN BLOCK, FIELD PLOT and GERMPLASM ENTRY respectively, If no label exists with one of these properties there should be no column for that item in the measurements table (eg SubBlock – this is just returned by your design generator but is not specified or needed in the template). Note your problem here is simply to merge some columns from the table returned by the design generator with data set columns defined by the template. The merge key is the PROPERTY which must match known columns from the design generator and the column headings in the measurement tab must come from the template and not the design generator table. (See points 15, 16 and 18 of IB Filedbook Prototype V3.doc which also contain other issues which need to be resolved).
- You still do not account for the Other Treatment Factors in the measurement Table.Pleas look at the example I sent you with IB Filedbook Prototype V3.doc, point 16. With template IBFB5-Wheat MultiFactorialMETwith Labels.xls there are two other treatment factore FERTLE and PDATE and the first has a label FERTAM, these must appear in the measurement table in combination with the main treatment factor – ENTRY.
- Now to try a new template: Cowpea_SST_Template.xls
- In the Study Conditions page of the wizard the LOCATION and LOCID conditions have PROPERTY LOCATION and scales DBCV and DBID but no look-up list is provided (canned or otherwise).
- This is a single site trial and has no factor with PROPERTY TRIAL INSTANCE. Therefore you should not show the Trial Instance spin box on the Study Conditions page of the wizard – there is only 1 trial instance! (See point 7 of IB Fieldbook Prototype V 1.0.doc which still needs to be resolved).
- Since this is a single site trial (has no TRIAL INSTANCE factor) it cannot have any Trial Conditions so you should not show the Trial Conditions page of the wizard or of the trial editor.
- Template Cowpea_SST_Template.xls has no Other Treatment Factors and so the Other Treatment Factors page on the wizard and tab on the Trial editor should not be shown.
- On the Experimental Design page, if the number of entries is prime you are not given any choice of Block Size and also no error message so you may not know what is wrong. Again we need to add some more designs urgently (RCBD and UNREPLICATED – they are easy).
- Now to try a new template: Cowpea_MST_Template.xls
- No Trial Conditions appear on the Trial Conditions page of the wizard. In fact there are four in the template - COOPERATOR, COOPERATOR ID, SITE, SITE ID. What might be happening here is that in this template the Trial Conditions are appearing in the FACTOR section of the template and not the CONDITION section of the template. (This also happens with IBFB7-Rice MET.xls) In fact the Trial conditions can occur in either section and the section they occur in indicates how the fieldbook should be printed. If they occur in the CONDITIONS section then the fieldbook should be printed (or saved) in separate files for each instance (with the Trial Conditions on the Description page). If they appear in the FACTOR section, this indicates that the fieldbook should be printed as a single book (or file) with the Trial Conditions appearing as columns in the Observation (measurements) sheet. In fact this decision can be taken later (at the time of printing (or saving) the fieldbook, but this convention gives us the opportunity to have a flexible default option.
- The Other Treatment Factors page is picking up TRIAL as another treatment factor and this is clearly wrong as it has PROPERTY=TRIAL INSTANCE.
- It is good that you have the save fieldbook function working, but you might as well save the fieldbook to a Workbook formt since this is just an excel file, you know the format of it already (since it is the same as the template), and we could then use the Workbook applications to perform a lot of functions which are not yet available in the IB Fieldbook – such as saving to the database, but there are many others.