G.I. Jerk":5f528]
"verification of the integrity of files is a pretty basic area and one which has been successfully implemented using standard techniques for some time without and need for variation..... ie if you knew anything you would have learned this shit years ago."
Uh gee.... I mentioned exactly said premise already. Perhaps you missed it? As you say a CRC check is pretty old hat man. BTDT eons ago 
[/quote:5f528]
then why suggest a far less usefull verification.... ie that a file of that name exists?
[quote:5f528]"to actually check for the existance of particular files within PK3s such as skins or TIK files, as you have suggested, is completely useless in both verification of stock and foreign files."
Not so I do believe. Unix DIFF or DOS FC commands can verify contents changed VERY easily EVEN IF (assuming text as in the TIK example) the byte count remains the same. Sorry Sir, you stand corrected. Ask your linux/perl/sql "guru" [/quote:5f528]
i dont stand corrected at all. you suggested checking if a file of a particular name exists as a method to identify cheats.... this is still completely useless and trivial to circumvent regarless of the operation of any programming language you wish to mention.
[quote:5f528]"so lets run through both these scenarios in order of probably occurance.... firstly to verify the integrity of the stock PK3s from the retail installation. the problems with employing such a check for this purpose are that verification of path does not address any known problem and that such verification is useless in verification of integrity of the file you are checking"
How so? You've already admitted you haven't the foggiest inkling how to do it via code. Care to take a stab at it? Send me the source code?[/quote:5f528]
you really dont get it do you? there is no code that can be created to explain this by me... you .... or anyone else.
as i pointed out there are no problems with people removing files such as TGA, TIK or SHADER..... it simply does not happen.... therefore checking that they are present is of no value. these files are present just the same on a default installation as they are on the most heavily modified system.
to use my example.... i can turn my manon skin into bright red without touching any part of my default PAK files. likewise i can modify my weapon TIK files without removing or replacing those files from the original PK3s therefore it is pointless to check that they still reside in the PK3s.
exactly how could you better understand this by seeing source code?
if you agree that there is a large section on this website called "maps, mods and skins" that hosts a variety of PK3 files that do work to modify your game without altering the original PK3 files then you must concur.
[quote:5f528]"there is no problem (and i presume never will be) in that i have deleted my skin out of a PK3 file for the "manon". so the implementation of such a check is completely useless in that it does not actually address any problem."
It is extreemly easy to (via code) overlook the missing. To use simple syntax for your benefit. --- IF NOT EXIST *.whatever (texture) GOTO NEXT. That is to say, what isn't THERE can harldy be construed to be possesing a cheat can it?[/quote:5f528]
granted it may be quite easy to find what is and isnt missing.... however if missing files are not a problem at all now of in the future of what benefit is it?
[quote:5f528]"there is however a problem that i can alter the "manon" skin to be a brilliant pink to allow me to see the bitch in the dark and against any background. in any implementation this will require that i alter the TGA graphics file that is used by the system to add this colouration. so if your sharper than a brick youll probably realise that actually checking the file integrity is the only option for preventing this type of cheating."
Well obviously this brick is sharper than your comprehension of my earlier posts on THIS thread as it is EXACTLY what I had proposed. No offence intended. [/quote:5f528]
so after saying im too dumb to comprehend the useless method you wrote you are telling me what i just outlined is what you initially meant when you said ? i got that right did i ?
[quote:5f528]"now onto the other possible scenario from your example which takes us to a much harder area to deal with - and please do let me know whether i have addressed this to your satisfaction
if you want to identify known cheats by way of applying any examination of the files contained in PK3 files.... whether stock or foreign.... you can, but with for no effective period worth implementing.
lets say you have a single cheat which consists of a spiked skin by way of an altered TGA file. if this circulates the net and is actually caught by your system, the kick you employed will be effective up until the name of the TGA file or its path are altered.... in which case your ststem will not find the cheat it is looking for. "
Not using checksums as BOTH of us have been thinking. No this is NOT to my satisfaction. PLEASE make Manon "brilliant pink" as you suggest. Then place it for me on the net...ftp...http....gopher...CVS or whatnot. (send URL) I will then write a script (using various executablesof my choosing) that will differentiate in a heartbeat betwixt the two. Then to PROVE the script is doing the work and not my human eye, I will video capture the execution thereof for you to see. I'm putting my money(and reputation) where my mouth is. Will you?[/quote:5f528]
uh.... my suggestion was that this method was easily circumvented so why would i want to attest to its effectiveness?
the greater problem in examining foreign PK3 contents for spiked PK3 files is that there is no code that can be written that can determine "artistic merit" of which i have prreviously written.
for example the shoot-bot that was originally discussed here used spiked blue skins for its allied soldiers.... which in turn make the players highly visible even without the auto-shoot functions of the rest of that hack. you can certainly tell that this skin is loaded rather than the default whatever skin that it replaces, however you cannot write anything that distinguishes it from the god-awfull sailors outfit in the skins section of aa.com. i doubt that it is a small thing to analyse a graphics file to determine its count of a particular colour, and even if it could you simply cannot remotely differentiate between a skin designed for high visibility and a skin that just plain sucks arse.
however we seem to have got to the point where you are willing to write code just to make a point..... why not just start writing the fucking thing instead of waiting for people like Crow King to do it?
i will most certainly put my reputation on the line where i think the test has merit.... however sending you a PK3 for you to locally point a program at does not. so create and release this tool and i (without any programming knowledge) will circumvent it. sound good? this process would actually aid in the development of the tool as opposed to you writing something to detect a single skin.
as i said dont think there is any weight to be attributed to me sending you a PK3 file which you are able to examine on a local system before creating something to detect it as servers to not have this priveledge.
[quote:5f528]"on the other hand if the system used the actually reliable checksum system employed for the verification of the stock files it would fall down in a similarly short period for similar reasons. simply changing the colour of my 255 red spiked skin to 254 red will alter its checksum to one which is not blacklisted in your system."
Well, enough arguing. Put your money where your mouth is. (only those who never take on challenges, never lose... I'm prepared to.) If I turn out to be wrong then SO BE IT! We would then know that this is a failed method and lets move on to the next. I assume you as much as I wish to find the answers to this dilemma? Right?[/quote:5f528]
well a pissing contest would interest me much less than the development of an actual working admin tool. once again if the tool is to have any use it needs to account for these variations beforehand...... so making it to find a file that i send you is of no merit, for it could be nothing more than a timestamp check to achieve the results you seek. you need to make it so that it has the ability to find foreign modifations outside of your reach.
(also if you disagree with me here and persist with this challenge as though it would have any merit..... then i will be happy to send you a batch of 200 skins, all but one of which being from respected skinners from our mods section and you can write something that doesnt tell us they are developing cheats yet detects the one that is - get my point?)
however if you wanted a challenge to prove that something of this nature would actually work a more usefull test would be to download ...say... 50 custom scopes from aa.com and write a tool that could identify which was loaded in the client and produce file info of that scope. if you think about how any tool of this nature is to be deployed you will understand the value of such a test over checking one particular PK3 file. the common nature and the file structure of scopes and in particular the fact theat the overwrite each other is also of benefit, along with the rest of the HUD.
[quote:5f528]
System? I haven't even begun to start on such project. In fact I never said I had the time to take on such an undertaking. .[/quote:5f528]
then why not do so? i cant fathom how any of these people develop things on their own when the beta testers run into the thousands..... so get yourself and the rest of these people who already have MoH programming interests together and start development. get an email list between yourselves and just do a bit when you can.
r3mix