Acknowledged
5 months ago

BUG: EDGE 1040 - CIQ 4

Something changed today. SDK Manager downloaded some new devices. And all of a sudden, EDGE 1040s return monkeyVersion: 4.2.4! That is broken.

Yesterday they properly returned 5.0.0.

EDGE 840 devices still do the right thing, returning 5.0.0.

So my fetch process is relegated to the background when emulating a 1040 now, rather than the foreground.

JUST FYI - use an 840 when testing in the simulator so you aren't limited by the 5 min delay.

DaveBrillhart
Parents
  • 5 months ago

    As per the other thread, this problem was resolved by moving away the existing device folder and re-downloading it via the SDK manager.

    Which raises the question: was a bad device update temporarily made available? If so, how many people did this affect? Was it not possible for the SDK manager to replace the bad update with a good update, if and when the problem was found on Garmin’s end?

    I’m just wondering how widespread and frequent this kind of problem is, and whether it’s possible in principle for the SDK manager to automatically recover from it, rather than the dev taking manual steps.

    After all, not everyone will necessarily know that they received a bad update.

    This also reminds me of another bug where a developer apparently received a bogus device that nobody else has (“instinct3solar”) in addition to the valid device everyone else has (“instinct3solar45mm”), which caused export issues when both were selected. If in fact the SDK manager received a bad device at one point, it seems that again, there was no way for the SDK manager to automatically recover from this issue.

    https://forums.garmin.com/developer/connect-iq/i/bug-reports/connect-iq-sdk-7-4-3-export-duplicate-entry-error-for-instinct-3

Comment
  • 5 months ago

    As per the other thread, this problem was resolved by moving away the existing device folder and re-downloading it via the SDK manager.

    Which raises the question: was a bad device update temporarily made available? If so, how many people did this affect? Was it not possible for the SDK manager to replace the bad update with a good update, if and when the problem was found on Garmin’s end?

    I’m just wondering how widespread and frequent this kind of problem is, and whether it’s possible in principle for the SDK manager to automatically recover from it, rather than the dev taking manual steps.

    After all, not everyone will necessarily know that they received a bad update.

    This also reminds me of another bug where a developer apparently received a bogus device that nobody else has (“instinct3solar”) in addition to the valid device everyone else has (“instinct3solar45mm”), which caused export issues when both were selected. If in fact the SDK manager received a bad device at one point, it seems that again, there was no way for the SDK manager to automatically recover from this issue.

    https://forums.garmin.com/developer/connect-iq/i/bug-reports/connect-iq-sdk-7-4-3-export-duplicate-entry-error-for-instinct-3

Children
  • 5 months ago in reply to FlowState

    The instinct3solar45mm device config folder was briefly named instinct3solar and then later updated. Having both would have duplicate part numbers and cause build errors. At the moment the SDK manager doesn't have a process for automatically deleting devices that are no longer relevant. This was intentional in the case that someone wanted to continue developing with an old preview device. A way to improve this is being explored.