View Issue Details
ID | Project | Category | View Status | Date Submitted | Last Update |
---|---|---|---|---|---|
0012156 | MMA | UPnP / Casting | public | 2014-08-29 20:22 | 2015-07-22 18:03 |
Reporter | rusty | Assigned To | |||
Priority | urgent | Severity | minor | Reproducibility | always |
Status | closed | Resolution | fixed | ||
Product Version | 1.0.7 | ||||
Target Version | 1.1.3 | Fixed in Version | 1.1.3 | ||
Summary | 0012156: UPnP: download manager status doesn't indicate status within a file | ||||
Description | If the user downloads a file with the UPnP > Download function, the status indicator is binary--i.e. it doesn't indicate download status within a file--it just goes directly from 0 to 100%. For a large file (or a slow network) this makes it look as if the download isn't working, even though it is. | ||||
Tags | No tags attached. | ||||
Fixed in build | 461 | ||||
|
Notification was refactored according Android design standards - system GUI - added remaining time Fixed in build 460 |
|
Tested 560 and it's generally working, however, the x minutes to finish indicator seems to fluctuate much more than it should (it continually changes every half second resulting in estimates that don't really make sense). I would suggest that calculations should be based on measurements of running averages of the last x seconds in order to limit such fluctuations, and perhaps only update the metric every 1-2 seconds. |
|
The main reason is that the wifi speed really fluctuates a lot. In my case 400kb - 2 MB/s. It was already implemented as exponential moving average. The smoothing factor was 3:1 (previous speed to speed of current sample). Sample was measured each 0.5 seconds. Time is not shown till enough samples is obtained. Now the factor is 5:1 and sample length is 1 second. It also counted the average speed per track - not per whole download process. This is fixed now. The speed is shared between tracks. I have also removed some testing code that caused failures. I have added 2 seconds to remaining time per each track too. This approximately simulates the connection and DB insertion time per each track and the remaining time is more accurate now. Fixed in build 461 |
|
Verified 461. |