Bug#529882: Perhaps found the soruce of the problem

2009-09-21 Thread Vincent Lefevre
On 2009-09-20 19:32:46 +0200, Vincent Lefevre wrote: > Due to this bug, guessnet is unusable, and several users could > reproduce this bug. So the severity should be grave. The attached > patch (based on Dietz Pröpper's comments) fixes the problem here. > There's still the 5-second timeout problem,

Bug#529882: Perhaps found the soruce of the problem

2009-09-20 Thread Vincent Lefevre
severity 529882 grave tags 529882 patch thanks Due to this bug, guessnet is unusable, and several users could reproduce this bug. So the severity should be grave. The attached patch (based on Dietz Pröpper's comments) fixes the problem here. There's still the 5-second timeout problem, but that wou

Bug#529882: Perhaps found the soruce of the problem

2009-09-20 Thread Vincent Lefevre
On 2009-09-20 12:05:59 +0200, Vincent Lefevre wrote: > Here 2000 also works, even 1000. But this always introduces a long > delay: I get the (correct) output only after 5 seconds. It seems that even though a packet may arrive much earlier, 1000 is the minimum timeout value for libpcap. I suppose t

Bug#529882: Perhaps found the soruce of the problem

2009-09-20 Thread Vincent Lefevre
On 2009-09-02 12:28:55 +0200, Dietz Pröpper wrote: > after reading a little bit of guessnet source and pcap documentation, > one of the possible reasons seemed to be a too short timeout for packet > capture (4th parameter of pcap_open_live). This is set to 500ms in the > original sources. Changing

Bug#529882: Perhaps found the soruce of the problem

2009-09-02 Thread Dietz Pröpper
Hi, after reading a little bit of guessnet source and pcap documentation, one of the possible reasons seemed to be a too short timeout for packet capture (4th parameter of pcap_open_live). This is set to 500ms in the original sources. Changing this to 5 (5000 should also suffice I think) sol