In fact this was supposed to be completed two years ago by my SoC student. Unfortunatelly it wasn't and still remains in my todo list..
My idea was not to use any external software, not because the NIH syndrome, but because we already have the software in place. So my idea is to use the current gcov website's code to display other's machine test results. The results would be stored in the DB in the same way as they are currently stored, but with some machine ID (I was thinking in opening this service for trusted machines only).
It seems reasonable. But how do machines get to be trusted? Would you modify run-tests.php so that it only attempts to post if the machine is known? If not how would you deal with input from non-trusted systems?
The initial implementation relies on a second script (similar with what's behind the gcov stuff), that has a username+password to upload the data directly to the gcov server. The test machines delivered the data already processed to the server, instead of submiting just the run-tests.php log (there's a valid reason to do like that).
The problem with input from non-trusted systems is that they could potentially inject html/javascript code in a php.net domain. We have to be safe :)
So, if you want to go this way you have two choices: do it yourself and get moderate help from me until July,
This suits me. I have a bunch of other things to do so may not get to it either but I have a little free time at the moment. I read your technotes on gcov but am not that familiar with the code so I have a quite a bit more looking around to do.
OK, sure. I'll be more than happy to help you.
Thanks for your interest,
Nuno