View Issue Details
|ID||Project||Category||View Status||Date Submitted||Last Update|
|0003482||3 - Current Dev List||Bug||public||2019-10-07 18:24||2019-10-19 10:18|
|Target Version||Fixed in Version|
|Summary||0003482: Improve timeout detection in call sign lookup operation completion|
|Description||The call sign lookup operation splits work between the main thread (for the Logbook, UCSDB, and country list operations) and a worker thread (for any other source). |
It's conceivable that something goes wrong in the worker thread, and results are not mmade available when the main thread expects to present them. This has always been a problem in the call sign lookup code, and isn't new to the rewrite. Code exists to do various message pumps and sleeping, but the bottom line is that an individual call sign lookup operation who's source has timed out or failed to respond will end up causing problems for the application.
This code could be tightened-up to be more reliable, and also transparent to the user about its error disposition.
|Tags||No tags attached.|
some work with this check in, for the related bug (which was actually about a deadlock)
This checkin adds code to assure the completion callback on the lookup client is only called once.
|2019-10-07 18:24||K7ZCZ||New Issue|
|2019-10-10 11:42||K7ZCZ||Relationship added||related to 0003485|
|2019-10-10 11:42||K7ZCZ||Relationship added||related to 0003486|
|2019-10-12 11:55||K7ZCZ||Relationship added||related to 0003484|
|2019-10-16 16:09||K7ZCZ||Note Added: 0008836|
|2019-10-19 10:18||K7ZCZ||Note Added: 0008865|