Function url.begin() returns succes when supplied with a dead link

[Frank-Bemelman]

Trying to make an internet radio, and of course I have more urls to play with than just the example, I used the VS1053 example.
Many urls work just fine, certainly the one used in the example.

But many other give problems.

Here is one example, trying to connect to http://theoldiestation.dyndns.org:8020/stream.mp3 which is in fact a dead link.
If I try that link in a webbrowser, I immediately get a response from the browser that this url is unreachable, which is ok.

Below is dump from the serial monitor in Arduino IDE 2.3.6:

18:04:43.986 -> url.begin ->http://theoldiestation.dyndns.org:8020/stream.mp3
18:04:46.955 -> [E] HttpRequest.h : 243 - Connect failed
18:04:50.459 -> [E] MetaDataICY.h : 246 - icy-metaint not defined
18:04:50.505 -> [W] ICYStreamT.h : 77 - url does not provide metadata
18:04:50.505 -> url.begin finished -> time elapsed = 6521
18:04:50.505 -> url Started with bool result: 1

Code snippet:

Serial.printf("url.begin ->%s\n", RequestedStationUrl);
lapMillis = millis();
return_result = url.begin(RequestedStationUrl,"audio/mpeg"); // what about "audio/mp3"
currentMillis = millis();
Serial.printf("url.begin finished -> time elapsed = %ld\n", (currentMillis = millis()) - lapMillis);

if(return_result==1){
Serial.println("url Started with bool result: 1");

The serial output reports connect failed but takes almost 3 seconds to find out. (yes I changed #define URL_CLIENT_TIMEOUT 3000;)
My webbrowser decides in a tenth of second about this. Is there any reason to keep this URL_CLIENT_TIMEOUT so high?

Secondly, there is the question why does url.begin(..... return a solid 1 while obviously it is not succesful?

Thanks in advance for any clarifications!

Frank.

[pschatzmann]

#2071