Help understanding Jira's Lexorank and how to use it with drag-and-drop.

[_ HigoChumbo _]

Hi there, I am trying to persist changes to a react-beatiful-dnd list using a third party implementation of Jira's Lexorank, but despite having watched this explanatory video several times I still can't wrap my head around how it works. My main questions are:

1.  In a rank that looks like 0|hzztzz:i what does the colon mean and do the characters added after it have any special significance?
2. What exactly is the marker row and why do I need it? What are the different types of marker rows (0, 1 and 2) and do they have anything to do with the 0, 1 and 2 buckets? None of the implementations I've found seem to use them.
3. What is the point of balancing if we can continue appending characters almost indefinitely? Is it just to keep ranks tidy and as short as possible? If so, should I do balancing after a fixed period of time? After a specific event happens?
   
   

Cheers =)

[_ HigoChumbo _]

I am using it outside of Jira, trying to persist a react-beautiful-dnd implementation in my own code, that's why I'm struggling, since I believe I'll have to do the balancing and all these things myself =)

Thanks for the reply.

[Nic Brough -Adaptavist-]

Ok, well, the rank field in Jira is built for Jira, not for other systems.  If you want to use a Lexorank system in your own code, you would be better off looking at a standard Lexorank implementation, not try to co-opt the modified one Jira uses.

[_ HigoChumbo _]

Thanks for the tip, I'll see what I can find.

The drag and drop implementation I'm using (react-beautiful-dnd) is also made by Atlassian and since I could not find any specific information on how to persist the changes made with it, I assumed that they were meant to work together in some way,

[Nic Brough -Adaptavist-]

1. It is part of the value, it has no special significance you need to worry about it
2. No, you don't need to think about the marker row either
3. It makes space for further re-ordering in the future and optimises the data so it doesn't take ages to read.

Essentially, the lexorank system is a black box and you should never have to care how it works in Jira.  You send ranking commands to Jira via the UI or REST and don't have to look at the rank at all.