Closed GoogleCodeExporter closed 8 years ago
Notes:
* Relevancy is only tracked on total active time spend on each file (currently, no Degree of Interest calculated)
* Folding should be done by the user (remembered, in the future could be used in a more predictive way)
* No event activity logging yet (read/edit patterns could be recorded based on line numbers where interaction took place?), no artifact detection (the explorer could be helpful)
* A simplified Database API was needed (inspired by python's "shelve" dict-like persistence), with abstraction and performance issues fixed
There are some improvement still to be implemented, like a better integration
with the PSP module (mixin), and a cleaner separation of concerns between UI
and logic (data uses a more consistent relational approach)
Original comment by reingart@gmail.com
on 2 Nov 2014 at 6:13
TODO:
* Task merge tracking additions / deletions (line numbers should be recalculated with hg!)
* Store hg revision in the task (so line numbers are meaningful in the future, as said in the previous point)
* Find a way to detect (and persist) the most read/written lines of code ("r/w relevance"?)
* Store more context beside files, breakpoints and folding (i.e. arguments, webpages?)
* Allow user Favorites?
Original comment by reingart@gmail.com
on 2 Nov 2014 at 6:30
Original issue reported on code.google.com by
reingart@gmail.com
on 20 Oct 2014 at 3:29