Methods tried:
Interchanging of Load priorities were not changing the load sequence in both Data access Tab and Preferences table.
Change in JSON sequence changed the order but behavior was not consistent
Ex: If RNA-Seq was first element in JSON then it would load before IGB Quickload but if Blueberry is first element, it would not load first so the behavior was inconsistent.
After analyzing, I understood that loadPriority of value "2" was missing. Load priority was 0, 1 ,3, 4, 5...
Issue was fixed when I added load priorities in proper sequence.
Now data providers load according to their Load priority irrespective of their placement in JSON file.

Also fixed test case associated with the same.
Branch link: https://bitbucket.org/swathark/integrated-genome-browser/branch/IGBF-1221-DPOrder

Please explain:

How does data provider order in the preferences JSON file affect top-to-bottom order in the IGB Available Data Sets section of Data Access tab?

We just reviewed our tests to be sure.

The order of the data provider nodes in the file igbDefaultPrefs.json has no affect on the order of the data providers in IGB.
The order of the data provider nodes in the preferences file (com.affymetrix.igb.plist) has no affect on the order of the data providers in IGB.

In IGB, data providers are displayed in Available Data panel, and in the Data Sources table in the preferences window.
In both cases, the order of the data providers is dictated by their load priority.
Note: The order may be inconsistent when multiple data providers have the same load priority. This issue is produced when a user adds data providers. Any time the user adds a data provider, it is given a load priority of -1. See issue IGBF-1225.

To extend from Sneha's email...

When the user adds a new data provider, it gets a load priority of -1 and it appears at the top of the list because -1 < 0.
When it is moved up and down the table, it swaps its priority with the data provider of current position.

As mentioned in IGBF-1220 and IGBF-1225, problems arise when additional data providers are added by the user because they are all given load priority of -1. When multiple data providers have the same load priority, the swap has no affect.

I am grabbing this for first level review.

In this file:
core/igb-preferences/src/test/resources/igbDefaultPrefs.json

I think ordering the nodes within the file to match the load priority is a good choice. --good job!
I think making all data providers have consecutive load priorities (no skipped numbers) is a good choice. --good job!

----------

In this file:
core/igb-preferences/src/test/java/com/lorainelab/igb/preferences/IgbPreferencesParseTest.java

Why did the package name need to change?
When I look at the master branch, Netbeans highlights it because the package name in the file (org.lorain...) doesn't match the package its in (com.lorainelab...).
But the package it is testing is called "org.lorainelab...".
Would it make more sense to change the package name to match the file and the other package?
Do the two packages need to match? The project was able to build, so it may not matter. Might just be a best-practice thing.

We could stand to improve this test. With this new test, we don't really know if the prefs are being read in correctly, just that its reading something, and it got at least one something. We should consider adding something like:

String aName = prefs.getDataProviders().get(0).getName()
assertTrue(!Strings.isNullOrEmpty( aName ) )

int loadPriority = prefs.getDataProviders().get(0).getLoadPriority()
assertTrue(loadPriority instanceof Integer)

Then we would know at least the first one had at least some symptom of being the right data type.
But at least for now, this modified test looks fine.

Digging into the tests here is outside the scope of this issue.

Sneha, I emailed you a note about doing a git rebase.
Please review that and the notes above and make what changes you see fit before passing the issue to the next 1st level reviewer.
(I think we are still doing 2-layers of 1st level review).

Yes, I liked the idea of modifying the test case in the way Ivory mentioned. Its more clearer and specific. I will change the test case.
Package name was throwing error. Packages are meant for separation. Test file is in com.lorainelab... but it had package name as org.lorain.. So i changed it to the right package it belonged to.

Fixed Test case as per the suggestions given by Ivory in First level review.
Moving the issue to "First level review" column.

Tested using the IGB-master.dmg downloaded from bitbucket main master...
After reseting preferences, the default data providers appear in this order:
IGB Quickload
RNA-Seq
Bio-analytic Resource
(I think that's wrong)

I also downloaded the latest version of master (with this branch merged in) and when I run it the data providers appear in the correct order.
This baffles me.

The test seem to work. The build fails when I make bad changes to the igbDefaultPrefs.json file.
I'm happy with the code, I'm just baffled by the installer producing a different result.