ROG Failure Email Patch

[ExtremeFiretop]

Currently with ROG routers using ROG builds of the firmware, post reboot the email notifications function believes that the firmware update failed when in fact it did not.

User will receive a "failed update" email even though the router updated successfully on rog routers using rog UIs.
Seems I had an issue with the post-reboot hook since the update happened successfully and it waited 3 minutes to send me the notification. Something was evaluated wrong in the post reboot.

Post reboot file variables:
savedInstalledVersion = 3004.388.6.0_rog
savedNewUpdateVersion = 3004.388.6.2

$fwInstalledVersion:
Pre-Update = 3004.388.6.0_rog
Post-Update = 3004.388.6.2_rog

There's a potential issue in comparing version strings with and without suffixes (like _rog)?
If the script strictly compares these strings, 3004.388.6.2 would not be considered equal to 3004.388.6.2_rog, and thus, the post-update scenario might not be recognized as successful.

[Martinski4GitHub]

Post reboot file variables:
savedInstalledVersion = 3004.388.6.0_rog
savedNewUpdateVersion = 3004.388.6.2

$fwInstalledVersion:
Pre-Update = 3004.388.6.0_rog
Post-Update = 3004.388.6.2_rog

There's a potential issue in comparing version strings with and without suffixes (like _rog)? If the script strictly compares these strings, 3004.388.6.2 would not be considered equal to 3004.388.6.2_rog, and thus, the post-update scenario might not be recognized as successful.

Great catch!!

[ExtremeFiretop]

Post reboot file variables:
savedInstalledVersion = 3004.388.6.0_rog
savedNewUpdateVersion = 3004.388.6.2
$fwInstalledVersion:
Pre-Update = 3004.388.6.0_rog
Post-Update = 3004.388.6.2_rog
There's a potential issue in comparing version strings with and without suffixes (like _rog)? If the script strictly compares these strings, 3004.388.6.2 would not be considered equal to 3004.388.6.2_rog, and thus, the post-update scenario might not be recognized as successful.

Great catch!!

Thank you! I noticed this behavior while testing this in dev, but at the time, I figured it was because I was just going 388.6 to 388.6 and that the comparison would always think it failed because the version numbers didn't change.

But today, the version numbers did change, And I still got the failure email, so it sent me down the hunt to find where the little sneaky bug might be hiding.