archive-ht.com » HT » H » HIGHLIG.HT

Total: 6

Choose link from "Titles, links and description words view":

Or switch to "Titles and links view".
  • Highlight Developer Blog
    us troubleshoot problems If we detected that a particular user s battery was draining more than it should the app would send back detailed reports so that we could quickly discover issues and respond with bug fixes Geofencing The biggest improvement in power savings came through the addition of geofencing in late 2012 Any time a user stopped moving we would set up a geofence around them This meant we didn t have to request another location point until the user moved and punctured the geofence In the meantime Highlight would draw no power Figure 1 How geofencing works To make sure we didn t miss a puncture we began setting up concentric geofences To increase efficiency while the user was moving we adjusted the size of the geofence based on the amount of time it would take to traverse it given the user s speed so geofences became bigger when users were driving and as small as possible when they were still Eventually we had optimized it enough that we could get GPS level accuracy with dramatically less battery drain Testing While new services promise to make this easier testing for battery life when the app is in the background is still a surprisingly manual process The most accurate approach we found was to set up two identical test runs one performed with Highlight running in the background on the phone and the other without Each time we would carry the device around without unlocking it until it ran out of battery and compare the resulting times By ensuring we were using a new device with a freshly installed operating system we were able to measure Highlight s impact on battery and improve it over time The results of the tests were as follows Figure 2 Relative standby time Though these tests have been very helpful it s important to note that they don t fully capture all the different ways people move with their mobile devices Other factors influence battery life such as the strength of the your network signal the presence or lack of nearby WiFi networks and the other applications that are running on your phone Heavier motion patterns and the use of additional radios like Bluetooth LE can also increase the battery drain Future improvements Our goal for 2014 is to reduce worst case background power draw by an additional 1 2 per month and we are confident that we can get there Some of the areas we are researching now include Network adaptation One of the biggest drains on battery for background location apps is having to boost the power of network radios when a user enters an area with poor cell reception We can detect these situations by monitoring data from the operating system and timing the throughput and speed of network requests With this information we can make smarter decisions about when to send location updates from Highlight Sensor fusion and indoor sensing Highlight allows users to opt in to using Bluetooth

    Original URL path: http://devblog.highlig.ht/ (2016-02-11)
    Open archived version from archive

  • Highlight Developer Blog: Archive
    helped us troubleshoot problems If we detected that a particular user s battery was draining more than it should the app would send back detailed reports so that we could quickly discover issues and respond with bug fixes Geofencing The biggest improvement in power savings came through the addition of geofencing in late 2012 Any time a user stopped moving we would set up a geofence around them This meant we didn t have to request another location point until the user moved and punctured the geofence In the meantime Highlight would draw no power Figure 1 How geofencing works To make sure we didn t miss a puncture we began setting up concentric geofences To increase efficiency while the user was moving we adjusted the size of the geofence based on the amount of time it would take to traverse it given the user s speed so geofences became bigger when users were driving and as small as possible when they were still Eventually we had optimized it enough that we could get GPS level accuracy with dramatically less battery drain Testing While new services promise to make this easier testing for battery life when the app is in the background is still a surprisingly manual process The most accurate approach we found was to set up two identical test runs one performed with Highlight running in the background on the phone and the other without Each time we would carry the device around without unlocking it until it ran out of battery and compare the resulting times By ensuring we were using a new device with a freshly installed operating system we were able to measure Highlight s impact on battery and improve it over time The results of the tests were as follows Figure 2 Relative standby time Though these tests have been very helpful it s important to note that they don t fully capture all the different ways people move with their mobile devices Other factors influence battery life such as the strength of the your network signal the presence or lack of nearby WiFi networks and the other applications that are running on your phone Heavier motion patterns and the use of additional radios like Bluetooth LE can also increase the battery drain Future improvements Our goal for 2014 is to reduce worst case background power draw by an additional 1 2 per month and we are confident that we can get there Some of the areas we are researching now include Network adaptation One of the biggest drains on battery for background location apps is having to boost the power of network radios when a user enters an area with poor cell reception We can detect these situations by monitoring data from the operating system and timing the throughput and speed of network requests With this information we can make smarter decisions about when to send location updates from Highlight Sensor fusion and indoor sensing Highlight allows users to opt in to using

    Original URL path: http://devblog.highlig.ht/archive (2016-02-11)
    Open archived version from archive

  • Highlight Developer Blog — How we improved Highlight's battery life 5x
    problems If we detected that a particular user s battery was draining more than it should the app would send back detailed reports so that we could quickly discover issues and respond with bug fixes Geofencing The biggest improvement in power savings came through the addition of geofencing in late 2012 Any time a user stopped moving we would set up a geofence around them This meant we didn t have to request another location point until the user moved and punctured the geofence In the meantime Highlight would draw no power Figure 1 How geofencing works To make sure we didn t miss a puncture we began setting up concentric geofences To increase efficiency while the user was moving we adjusted the size of the geofence based on the amount of time it would take to traverse it given the user s speed so geofences became bigger when users were driving and as small as possible when they were still Eventually we had optimized it enough that we could get GPS level accuracy with dramatically less battery drain Testing While new services promise to make this easier testing for battery life when the app is in the background is still a surprisingly manual process The most accurate approach we found was to set up two identical test runs one performed with Highlight running in the background on the phone and the other without Each time we would carry the device around without unlocking it until it ran out of battery and compare the resulting times By ensuring we were using a new device with a freshly installed operating system we were able to measure Highlight s impact on battery and improve it over time The results of the tests were as follows Figure 2 Relative standby time Though these tests have been very helpful it s important to note that they don t fully capture all the different ways people move with their mobile devices Other factors influence battery life such as the strength of the your network signal the presence or lack of nearby WiFi networks and the other applications that are running on your phone Heavier motion patterns and the use of additional radios like Bluetooth LE can also increase the battery drain Future improvements Our goal for 2014 is to reduce worst case background power draw by an additional 1 2 per month and we are confident that we can get there Some of the areas we are researching now include Network adaptation One of the biggest drains on battery for background location apps is having to boost the power of network radios when a user enters an area with poor cell reception We can detect these situations by monitoring data from the operating system and timing the throughput and speed of network requests With this information we can make smarter decisions about when to send location updates from Highlight Sensor fusion and indoor sensing Highlight allows users to opt in to using Bluetooth LE if

    Original URL path: http://devblog.highlig.ht/post/78783186365/how-we-improved-highlights-battery-life-5x (2016-02-11)
    Open archived version from archive

  • Highlight Developer Blog: Archive
    Developer Blog Month Filter by post type All posts Text Photo Quote Link Chat Audio Video No posts yet 2014 January February March April May June July August September October

    Original URL path: http://devblog.highlig.ht/archive?before_time=1394143380 (2016-02-11)
    Open archived version from archive

  • Highlight Developer Blog: Archive
    helped us troubleshoot problems If we detected that a particular user s battery was draining more than it should the app would send back detailed reports so that we could quickly discover issues and respond with bug fixes Geofencing The biggest improvement in power savings came through the addition of geofencing in late 2012 Any time a user stopped moving we would set up a geofence around them This meant we didn t have to request another location point until the user moved and punctured the geofence In the meantime Highlight would draw no power Figure 1 How geofencing works To make sure we didn t miss a puncture we began setting up concentric geofences To increase efficiency while the user was moving we adjusted the size of the geofence based on the amount of time it would take to traverse it given the user s speed so geofences became bigger when users were driving and as small as possible when they were still Eventually we had optimized it enough that we could get GPS level accuracy with dramatically less battery drain Testing While new services promise to make this easier testing for battery life when the app is in the background is still a surprisingly manual process The most accurate approach we found was to set up two identical test runs one performed with Highlight running in the background on the phone and the other without Each time we would carry the device around without unlocking it until it ran out of battery and compare the resulting times By ensuring we were using a new device with a freshly installed operating system we were able to measure Highlight s impact on battery and improve it over time The results of the tests were as follows Figure 2 Relative standby time Though these tests have been very helpful it s important to note that they don t fully capture all the different ways people move with their mobile devices Other factors influence battery life such as the strength of the your network signal the presence or lack of nearby WiFi networks and the other applications that are running on your phone Heavier motion patterns and the use of additional radios like Bluetooth LE can also increase the battery drain Future improvements Our goal for 2014 is to reduce worst case background power draw by an additional 1 2 per month and we are confident that we can get there Some of the areas we are researching now include Network adaptation One of the biggest drains on battery for background location apps is having to boost the power of network radios when a user enters an area with poor cell reception We can detect these situations by monitoring data from the operating system and timing the throughput and speed of network requests With this information we can make smarter decisions about when to send location updates from Highlight Sensor fusion and indoor sensing Highlight allows users to opt in to using

    Original URL path: http://devblog.highlig.ht/archive/2014/3 (2016-02-11)
    Open archived version from archive

  • Highlight Developer Blog: Archive
    Developer Blog Month Filter by post type All posts Text Photo Quote Link Chat Audio Video No posts yet 2014 January February March April May June July August September October

    Original URL path: http://devblog.highlig.ht/archive/2014/3?before_time=1394143380 (2016-02-11)
    Open archived version from archive