expose a structure's "order" so JQL queries and filters can be created to list issues in that order.
The order's created would be very handy to be able to query in JQL - something like: project = MARSCOLONY AND assignee = "cosmonaut" ORDER BY "taskorder1-structure" ASC
This would allow exposing structures in Confluence and other places in a useful way.
Not sure what to do about different levels, but initially just ignore issues that aren't at the top level?
I support the need of Patrick Metz question regarding "order by structure_order".
I have to prioritize epics therefore the RANK function is not useful because the rank is changed by everyone who is using the epic on several boards.
I am using "Allow manual adjustments of generated content" to order the epics. A query is calling the structure to use the same content on a Kanban Board - but the order is not the same I have in my structure :(
It looks like you export in fractions of days so I can use a formula to get it to a compatible format.
It would be better just to export the numerical number of hrs/secs though.
On import to project I still get some errors and crashes but that is perhaps MS Projects problem.
When you export to excel the indentations on the task name are incompatible with MS Project import. I have got round this by removing the indents so that is not a big deal.
What i'm finding problematic is the Original Estimate / Time Spent. These fields are converted using a custom format. MS Project see's this as invalid and just drops 0 hrs in each time I synchronize. If you could export to excel in simple text format the calculated hours value that would be better. Or, I think the native format of the data in JIRA is seconds, just export the seconds and a simple excel formula can be used to convert before MS Project Import.
A final, trivial point, is that sheetname seems to have invalid characters in it preventing excel from accessing the sheet (this could of course just be due to my structure name). I just rename the sheet just now. If your standard export function just left the sheet name as "Sheet 1" then the problem would not occur.
Thanks for your feedback. Making it possible for JIRA query / export result to be in the structure order is not very simple, due to JIRA's limitations.
However, it seems to me that you only need this once in a while. So what if, say, there's a button "export order to custom field". You press it, select a numeric custom field, and then run the export. Structure updates values in this custom field for all issues in the structure to 1, 2, 3, etc (or even in WBS notation - 1, 1.1, 1.2, 1.2.1 etc). Then you can use JQL with "order by <customfield>".
Note that this is one-time operation, and there's a possibility of a conflict if two users export orders from different structures to the same custom field.
Would that work for you?
Also, regarding Excel export - could you please tell me which columns are in Project-incompatible format?
Followup to this, I spotted the 'x' in the latest structure version views and exported my structure to excel. This is really useful, but the native data types exported to excel are not compatible with an MS Project import which is a shame.
Even something as simple as an export of the struture 'tracking' view to excel would facilitate my need.
I would like to see something like this such that the order of the structure would be used in an export of the JIRA fields. We are working on projects where we still use MS Project for longer range dependancy forecasting and end of project estimation. Through a simple JIRA filter export we can export all tracking to the Gantt chart for easy update and synchronization. If we were able to also export the order within the structure it would make it easier to maintain the Gantt on a biweekly / monthly basis such that the task depedancies could be built quickly in the Gantt chart to reflect that which has been "structured". All I need is a filter ordered as per the structure to make this super easy.
Patrick Metz commented
@Igor Sereda: Thanks for your reply.
Perhaps, when the time comes and this idea has enough supporters, you could make it optional:
so that one could choose between loosing performance and exposing the "structure order", or keeping performance and not exposing the "structure order".
Hi Patrick, thanks for your comment! Unfortunately, there's no good solution when it comes to ORDER BY. Having an ORDER BY would mean having a custom field (one per structure!) where this order is represented (as with Rank in GreenHopper) and issues reindexed every time the order changes. It is doable though.
Patrick Metz commented
I support this vote.
This morning I plugged JIRA and PhpStorm together with Atlassians IDE Connector and tried to get a part of our current projects Structure to be displayed in the "Atlassian" tab of the IDE.
So I created a filter in JIRA and then selected it in the IDE. It worked, but was only partly satisfying, since issues could not be sorted by their "structure order".
This is the SQL query that I used:
issue in structure("Project Name", "[status != closed] and child of ID-123")
But this is what I actually needed:
issue in structure("Project Name", "[status != closed] and child of ID-123") ORDER BY "structure order"
By the way: are there any plans to integrate Structure support into Atlassians IDE Connector, or to create a dedicated ide plugin for Structure?
Excellent. Very excited about that.
The gadget is coming in the next version (1.2) within a couple of weeks.
Where can I find this confluence gadget to display a Structure in confluence?
Hi Brendan, thanks for your idea!
Is the purpose only to show Structure in Confluence? We will have a gadget that will display the Structure widget on JIRA dashboard or in Confluence - so that you can also edit the Structure from there.
I'm not sure if "structure order" custom field makes much sense. It won't be indexable (otherwise any operation on structure would cause reindexing, which is unacceptably slow) - and therefore sorting or searching by that field would be very slow. It won't be consistent - because the "order" is not actually a property of a single issue, so there's possibility of timing issues (what happens if you rearrange structure while JQL result is being sorted).