Forum Moderators: coopster
Ive checked all the _SERVER vars for anything that might be useful besides IP/HOSTNAME and so on, i want to be able to allow and ban people from the same IP address, ive considerd numerous combinations IP/HOSTNAME/USERAGENT and so on however the end user (if an attacker) could use this to their advantage if aware of how the system works (Unless they have more than one or two machines to work off, this wont cause any issues - if they know what there doing in my mind they would hijack someones pc to do the attacking from anyway?!?).
Is there anything i can use to identify the end user as unique to the rest of the users on the same network/gateway
TIA
IP + USERAGENT is the best that I can think of immedietely, bu chanigng user agent is trivial.
You could of course set a cookie on their machine, but trivial to remove again.
I'll keep thinking. Let us know what you come up with!
This page gives an overview of that for Intel pentiums and later. You'll need some adaptive code to cover other chip types and platforms:
[support.intel.com...]
It is also possible to make some close statistical guesses about a machine based on the setting of its internal clock. That really only requires some Javascript to return the machine's clock time.
Similarly, with Javascript, you can gather version numbers of components that are running. Enough of those, and you have a profile that will identify if not a particular machine, at least a smallish group.
But the more you sniff a machine's state, the more likely you are to be identified as a hacker or virus precursor and be treated accordingly
On another note, making things difficult, is it possible to only allow connections on a predefined port in my script (i.e. in a config.inc.php file stored outside of the root docs folder)? I'll take a look into this on le search engine however thought i would probe here first. I know the likes of cpanel and other system i.e. Raven Core do this however im unsure if this is done via .htaccess, php.ini (or setting manual at start of script?), httpd.conf or the scripts them selves.
Also (Sorry for banging on here) for peace of mind, is there any utility out there which can phish php scripts (unparsed)? I recall someone babbling on about it a year or so back however if it was the case im sure the whole php industry and all its offspring would have collapsed if so (i dare say it was (i dont like this word) a n00b who spouted such babble)).
TIA
identify a remote user uniquely
Or, consider making them log in with user/password, if you're actually concerned about unique people, rather than unique machines.
only allow connections on a predefined port in my script