planner files are difficult to diff because of constantly changing ids
Submitted by Martin-Éric Racine
Link to original bug (#318480)
Description
Please describe the problem: The output format for planner, both to database and to file, has a bad problem. It supports an id="" attribute for all items in the file, and uses this to cross-reference - fine. However, planner generates these ID's as integers numbered-from-1 every time the plan is output.
This means that it is impossible to do concurrent work on a plan!
If the plan is kept under a source-code-management system, the ID numbers make it impossible to merge plans. Likewise in databases.
The in-program datastructures don't make use of the ID.
If the ID were a string, generated like "<username>
-<host>
-<datestamp>
",
which was GQuarked on load/save of the file, then it would be possible
to store the plans sensibly under CVS, which would make this program
useful.
Steps to reproduce:
Actual results:
Expected results:
Does this happen every time?
Other information: This is an old bug report remaining in the Debian BTS.
It probably was fixed already but I'm hereby polling the upstream authors to verify. If it's fixed, I will close the corresponding bugs in the Debian BTS.
Version: 0.12