Put in which file?

[Charity]

OK. I have two files one with GUI and one with the data.  Now I want to make value lists but both files seem to be able to create the same value list in their own file using the table occurrence there but also can create a value list from the other file.

 

Is it better to create the value list in the data file since it is closest to the real data?  Or is it better to create the value list in file my people will work in, the UI since that is closes to them?

 

When would I use the part that says "use value list from another file" and why would I use it?  This is being set up where everything happens in the GUI file.  I can find nothing in searching so the question is probably dumb but it stumps me which to use and when and why.

[IdealData]

"Use value list from another file" is necessary when the value list is a custom value, i.e. a static list. Maybe a "master" list that would be in your data file. It can of course be used on a dynamic list - one that is based on content.

 

VLs based on content can be built either as the content of a specific field in the 'local' file, or as you have realised via relationships. A VL based on a relationship can effectively be 'shared' between files by using the VL from another file so you don't need to make another relationship in the file.

 

Watch out for relationships based on global fields as VLs may not work when the global has no value.

 

Not too sure about performance, but it's logical to consider that VLs from another file mean there's only one VL 'doing the work'

Experiment and find out what's best for you.

[Charity]

Thank you, Mr. Ideal Data

 

Then that means I can use one less table occurrence group in the GUI since I have the primary relationship already in data.  Now this is nice idea.

 

How would I go about testing which way would be faster?  Do you have any ideas?  I will then need to decide whether a new table occurrence group in UI or using value list from data which exists is the most speed effective.

[Charity]

Does anyone have any ideas about which would be most efficient?  Or how I could conduct a test to find out?  It seems to me that it would make most sense to keep all value lists in the same table just for ease of maintenance which would be the GUI but if it saves resources to put some in Data file I would do it.   Also I wonder why that option isn't talked about more than it is.  Mass searching only produced 3 links about it.  

 

Do other people use it?  Do other people wonder about where to put their value lists?  Do other people wish I would shut up?  

 

Honestly, people must make decisions on where to put them.  Speed, accessibility, reduction of table occurrences and other reasons?  I would give up accessibility if it were faster in data file.  Everything should be considered in these decisions shouldn't it?  Thank you anyone willing to give their thoughts.

[jbante]

Where I put the value list depends on what the value list is for; I try to put value lists in the file consistent with the purpose of the value list. If the value list is driven by data values or defines what values are valid, I'll put it in the data file. If the value list drives some interface convenience, I'll put it in the UI file.

[Charity]

Hello Mr. jbante, thank you for answering on this thread.

 

I have been studying.  I think I now get it about value lists except can you explain what you mean here please?  I put the specific part that I do not understand in red.   "If the value list is driven by data values or defines what values are valid, I'll put it in the data file."

 

Conditional value list is example of uncertainness and this may be what you mean but I would have thought that it needs to go in GUI since it uses globals and the table occurrences group is in GUI.  How else might I "define what values are valid" so I will know they go in data file?

 

Any additional thoughts you can provide me would be great.

[jbante]

Charity, some value lists defined on a field just show what values have already been used in that field, which is what I mean by a value list "driven by data values."

 

In contrast, some value lists with custom values are used as the set of allowable values. For example, a value list for currencies might contain "dollars", "pounds", "euros", "yen", etc. Then user interface controls might use that value list with a pop-up menu to minimize entry of inappropriate values like "meters" or "kilograms", and the validation settings on the field can prevent entry of values not in the value list entirely. This is what I mean by a value list that "defines what values are valid," and since it's more a function of the data model than the user interface, I put this kind of value list in a data file in a separation architecture.

 

You are correct in thinking that conditional value lists have to live in a user interface file.