twoolf / circator

Metabolic Compass: A Circadian Monitoring Mobile Health Stack
https://www.metaboliccompass.com
Apache License 2.0
6 stars 1 forks source link

Daily Progress Screen: Sometimes unresponsive #55

Open twoolf opened 8 years ago

twoolf commented 8 years ago

probably data related, but a reentry to daily progress screen after an initial failure to populate the screen seems to prevent a reentry to the screen; updates still on first and second dashboard; but locked from third --

adding onto this -- on cleaning up data on iPhone, I had multiple entries for testing purposes on some events (e.g. sleep was sometimes more than one entry). This might be a cause for the error, though the delays and crashes are still present --

further adding, did get a version to come up, but it took six minutes of patience, way too long for the average user -- nothing obvious in the logs to indicate why so long -- this was after several cycles of data cleaning and complete clean/re-compile -- noting that some sleep events are not reported correctly on the plot -- maybe due to different entry sources (MC Watch vs HK entry?), though if so that is also not consistent, since it picks up some, but not all of each entry type; might be related to events that are deleted from HK, but that had been present earlier -- not sure whether it is somehow reading deleted events and then using those as the entry points.

should a horizontal screen be supported for this?

adding on: ten minutes into first launch of App on phone (3rd complete clean/reinstall after data cleaning), only now with Daily Progress up, takes way too long to load -- still oddities to data display, with some events in HK not seen on display and vice-versa --

adding on, screen is consistent, as being the same from repeated runs, but not consistent with HK data; note that first three days are fine, its the last four that are off; perhaps its something within the memory allocations for the later days?

more adding on: the first four lines are correctly mapped to HK, but on the 5th line we've somehow mapped an eating event to a sleeping event (probably not related to mode of entry, since on the first four days there are both HK entered and MC entered events and they are fine); from that point on the plot doesn't make sense, for example there is nothing else plotted after that on the 5th day, and 7th day has an eating event that was not in HK

conny-the-unicorn commented 8 years ago

@twoolf 1) landscape mode is not supported through the app 2) other points added to JIRA to double check and fix