to quote from Slamlander's post.
Indeed, if this is a "real time response machine" that is some seconds "off", it should only be off if compared to an EXTERNAL clock. Besides, en external clock COULD re-sync the machine.
But in the case of a missile-detection-and-response system, this is NOT the main problem.
How does it calculate the response? On boot-time or ... after detecting the approaching missile?
If it already detected the scud on boot-up, and made the calculations, then scheduled the response time ... it would have missed by said 0.4 seconds
OTOH, if the whole detection was done AFTER the scud had been launched, the calculated response time would have had the very same offset as the detection time. Therefore the reaction should have been adequate.
So go on, blame the chip-maker for trying his best, blame anybody but NOT the people mis-coding the whole thing. Because this should be something that already had appeared in test-situations. They CAN build a "hollywood" village, put test-systems in it, fire rockets to it and measure the (in)accuracy of their systems BEFORE the army buys it ... why don't they? And why does the army buy it if it hasn't been tested? |
