Android Question Log RAW HTTP request

That's not how you usually do it. Instead you have something sitting between the device and the backend.

Historically I've used Charles Proxy for some things:

Currently I'm using ngrok:

The latter might seem complicated, but once you install it, it's almost dead simple to use.

There are many alternatives to the two above:

Sandman said:

That's not how you usually do it. Instead you have something sitting between the device and the backend.

Historically I've used Charles Proxy for some things:

Charles Web Debugging Proxy • HTTP Monitor / HTTP Proxy / HTTPS & SSL Proxy / Reverse Proxy

Charles Web Debugging Proxy - Official Site

www.charlesproxy.com

Currently I'm using ngrok:

ngrok | API Gateway, IoT Device Gateway, Secure Tunnels for Containers, Apps & APIs

ngrok is a secure ingress platform that enables developers to add global server load balancing, reverse proxy, firewall, API gateway and Kubernetes Ingress to applications and APIs.

ngrok.com

The latter might seem complicated, but once you install it, it's almost dead simple to use.

There are many alternatives to the two above:

https://alternativeto.net/software/charles/

Click to expand...

So you're telling me that from B4A I can't print the request that it itself is making?

Nope, that's not what I'm telling you. I'm just telling you how most people do it.

If you want the info directly from the phone, I imagine you would have to root it and install some sort of network sniffer. Seems like a lot of work, but go ahead and please report back to the forum - I'd love to hear how it went.

Also, in fairness, it might be possible to get OkHttpUtils2 to produce some sort of verbose report, but I've never heard of it.

there are online test servers available or you could run one locally. just send your
request to it and display what it sees: headers, body, everything.

if a server crashes, there is no network connection. nothing to report back, no way to
report it. that someone could crash a modern-day server with an okhttputils2 http
request is pretty amazing on its face. i reckon erel is more of a genius than we thought.

in the event of some exception, an http server will report what it has been
programmed to report; in most - if not all - cases, that will be a 50x
http error. occasionally, as in the case of ssl failures, the caller may receive
a little more information. what the server may report to itself will be to itself. okhttputils2
does not have any enhanced interrogation methods to find out.

okhttputils2 is a conduit using square's ok technology. square has developed all
the request- and response-related code. okhttputils2 wraps it. blaming okhttputils2
is a stretch.

drgottjr said:

blaming okhttputils2 is a stretch.

Click to expand...

I don't think anybody did that.

drgottjr said:

that someone could crash a modern-day server with an okhttputils2 http
request is pretty amazing on its face

Click to expand...

Not really. If by "server" you mean the backend that is called behind the actual webserver (Apache, for instance). Any bug in there might very well result in a crash in the backend code.

But yes, if you did mean the actual webserver (Apache, for instance again), then it is exceptionally unlikely.

Mike1970 said:

the server crashes and I have no way to see why.

Click to expand...

The server is not developed by you?
Otherwise why not handle the error using Try-Catch?

You can use Requestbin. Click public bin to Generate URL. Use that Url as endpoint in your request.

mcqueccu said:

You can use Requestbin. Click public bin to Generate URL. Use that Url as endpoint in your request.

View attachment 146738

Click to expand...

awesome tool!

aeric said:

The server is not developed by you?
Otherwise why not handle the error using Try-Catch?

Click to expand...

Yes, the backend is developed by me, but the client receives an error BEFORE the request arrives on the server , so it is a big mistery.

Edit: the problem is solved, somehow the internal configuration of the webserver has changed autonomously (maybe some update of the webhosting center, i don't know) and the maximumRequestLenght parameter has changed so the requests were trashed before reaching the code because too long for the "new" limit...

BTW would be useful to have such a manner to log the http request i'm sending directly from B4A.. you know.. debug purpose, understand what the phone is actually sending out