HamApps Support (VK3AMA)
On 25/01/2019 1:38 am, Ed wrote:
There is no real disk-access bottleneck, except for the delay at startup as config data is read into memory.
There is also disk-access delays where querying your log file for each decoded callsign. This can add a second or two when the band is very busy, BUT, JTAlert mitigates the effect of these log lookup times by caching in memory the results of a lookup. This way there is only one disk-based lookup for each unique callsign encountered during a JTAlert session. This affect can be seen when first starting JTAlert when all callsigns are new and don't have a memory-cached record.
What response time are you trying to minimize? Is it the time between when WSJT-X has produced decodes and when JTAlert displays the Callsigns? What sort of delay are you seeing? If your trying to get JTAlert to display Callsigns within the short 2 second window at the end of a FT8 period, that is unlikely especially with a band full of signals.
de Laurie VK3AMA