I was running Fedora Core 7 on my machine until this weekend when I upgraded to Fedora 11. This meant upgrading from nut 2.0.5 to nut 2.4.1. I have a Powercom Black Knight 600AP UPS, and this was always driven by the powercom driver, type KIN1500AP. Now I had to use type "BNT" to connect to the UPS. My rather long-winded question: the powercom driver now does some kind of "battery test" when it connects to the UPS. Is this correct? I did some checking in the svn repo, and this change was added in revision: http://svn.debian.org/wsvn/nut/trunk/drivers/powercom.c?rev=1236&sc=0 In method "upsdrv_initups" there is now an extra section that does the "ser_send_char(upsfd, BATTERY_TEST)" for some reason: ---------- code ---------- if (!strncasecmp(types[type].name, "BNT",3) || !strncasecmp(types[type].name, "KIN",3) || !strncasecmp(types[type].name, "IMP",3)){ if (!ups_getinfo()) return; if (raw_data[UPSVERSION]==0xFF){ types[type].name="IMP"; model=raw_data[MODELNUMBER]/16; } if (raw_data[MODELNAME]==0x42){ types[type].name="BNT"; model=BNTmodels[raw_data[MODELNUMBER]/16]; } if (raw_data[MODELNAME]==0x4B){ types[type].name="KIN"; model=KINmodels[raw_data[MODELNUMBER]/16]; } linevoltage=voltages[raw_data[MODELNUMBER]%16]; sprintf(buf,"%s-%dAP",types[type].name,model); modelname=buf; upsdebugx(1,"Detected: %s , %dV",modelname,linevoltage); if (ser_send_char (upsfd, BATTERY_TEST) != 1) { upslogx(LOG_NOTICE, "writing error"); dstate_datastale(); return; } } ---------- code ---------- The outer "if" (with the string comparisons to BNT, KIN and IMP) does not make a lot of sense to me? So my questions are: . why the battery test? . the if looks a bit strange - is it supposed to look like that? Thanks! Albert -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.alioth.debian.org/pipermail/nut-upsuser/attachments/20091011/95bf8d32/attachment.htm>