Some note worthy stuff on firmware 76 and 6.5.2: Two days ago, I replaced a malfunctioning SQB with a fresh boxed SQB. The malfunctioning SQB was not exhibiting transition problems, but was not accepting a static IP address (DHCP ok). I guess you could call the behavior spontaneous reboot, but since the SQB was never fully connected, I consider it a failure to initialize. The SQB responded with "Invalid IP address". Something I had never seen before. The only difference between this SQB and the boxed SQB was that it was briefly connected to squeezenetwork and used to play ogg radio streams (that failed to play since the bitrate was not supported). Factory reset, firmware update.. all had no effect on restoring the prior ability to assign a static IP address.
Another SQB, also part of radio testing (but from slimserver in referral mode, not squeezenetwork), was having transition problems, while another SQB (not part of squeezenetwork or radio referral from slimserver) connected to the same server (unsynced) was not having transition problems. The faulting SQB is in the process of being replaced and tested. Have you connected your SQB at any time to a radio station or to squeezenetwork? Tried to play ogg files? I agree that 6.3.1 did not exhibit this transition trouble. Neither did version 6.2.2. Prior to that, transition trouble was sporadic across versions. I have also noticed an apparent correlation between custom MySQL and more frequent transition problems. No theories on why that might be the case. You might want to try using the 6.5.2 default mysqld to generate the fault. Using the default mysqld will get more attention from developers. As for the direct xover test, the basis of the test is that Ethernet is a broadcast implementation. That means that every frame is seen by virtually every NIC. The NIC that eventually decides that the frame is being sent to it (by comparing the frame MAC address to its own MAC address - among other things), takes ownership of the frame. A faulty NIC/driver might decide to garble the frame, or engage in some other non standard behavior. Besides, reducing the number of components required to create the fault brings one closer to a reproducible case, or identification of a faulty component. Richard -- RichardG ------------------------------------------------------------------------ RichardG's Profile: http://forums.slimdevices.com/member.php?userid=801 View this thread: http://forums.slimdevices.com/showthread.php?t=33044 _______________________________________________ beta mailing list [email protected] http://lists.slimdevices.com/lists/listinfo/beta
