Too many rejected connections and weird things

[jordi120]

Hi,

I've done a little app with the WCL, and one of the basic features is too show the statistics. The people for whom I work, are telling me that there are a lot of rejected connections in comparison with their old software.

Then I tried the proximitySenderDemo and I'm founding a lot of weird thins. First of all, i put a timeout time of 120 seconds, and each proximitysender round, there are a lot of devices who send a 'timeout' (WCL_E_TIMEOUT) in the first 15 seconds.

Therefore another lot of devices send rejected connection in the first 15 seconds too (WCL_E_CONNECTION_REJECTED). On top of them, a few of them throws a internal driver error and others obex_forbidden and stuff like that.

What could be the reason of this? the usb drivers? the usb dongles? there are actually a lot of people looking at the movile in each second and they reject the request inmediatly?

Am I doing something wrong? do I have to put in the black list some manufacturers like IPhone and blackberry(I already put this two manufacturers MAC on the black list, so they are beeing ignored).

I will apreciatte any kind of help

[Mike Petrichenko]

Hi,

There is no one phone except SonyEricson which will wait more than 15-20 seconds. So that is absolutly ok that you have timeout in 15 seconds or so. Some devices may send Reject code even timeout appeared. OBEX_FORBIDDEN - usualy sent by SE/LG as user reject code (LG as phone reject even timeout). Actualy you got good result. Just may be wrong result code interpretation.

I can recomend to try our solution: http://www.btframework.com/textblue.htm it is WCL based application.

[jordi120]

Thanks you, actually de app I'm developing has a few especial features. It connects to a remote server, and download the campaigns, upload usage stats and stuff like this, this way the clients can check the stats from the web, and make comparisons between diferents usb dongles on diferents locations.

By the way, thanks for the response, i'm going to change the way I'm 'reading' the codes. I'm testing now the textblue and found that the stats are pretty similar. I'm testing bluesender too, and the same thing.

So i think the last software my company used was stat-manipulated.

Have nice day!

[Mike Petrichenko]

That is great idea to use remote server to collact stats. I just recomended TextBlue as a other software you can test functional and compare to your.