Digital Rights Management

[monele]

http://www.gamesetwatch.com/2008/02/opi ... _pirac.php
http://www.gamesetwatch.com/2008/02/opi ... ames_-.php

[Hentai Senshi]

Well then, does anyone know any software that make trial version. I will make the trial period around 20-30 minutes.

Timed trials are inherently relatively easy to crack; if you do want a timed trial anyhow, you're better off writing it yourself than using a stock trial-maker - popular wrapper programs are likely to have a generic cracking technique out there that works on _EVERYTHING_ protected by them, making it trivial to unlock them. At least to crack your custom timed trial engine the would-be cracker must have three brain cells to rub together instead of 'DURRR... download generic patch'.

Also, consider how you intend to sell the game - if you want to sell it through large corporate portals, they have their own wrapper software that they will want to use for the timed trial. This wrapper will often be ridiculously ineffective - one was foiled by the simple expedient of not quitting the game while playing, another can only be defeated by a hacker skilled enough to copy a file.

[Hentai Senshi]

Update: I checked out DLSite too and it seems CuriousFactory.com is their english/american agent. You can publish with them, the product appears on DLSite.com and you can get copy protection for free (although they obviously charge a fee for every purchase). I like it because it's being handled entirely by DLSite and in case of problems I can point to their support service.

Galaxy reports that, from the customer's perspective, DLSite's copy protection is quite painful to use (See
http://lemmasoft.renai.us/forums/viewto ... =22&t=2967)

[papillon]

Timed trials are inherently relatively easy to crack; if you do want a timed trial anyhow, you're better off writing it yourself than using a stock trial-maker

... well, if you're comfortable enough with code to do that and make it safe, secure, and reliable. Since game development and application development are generally different skills, there are an awful lot of people who really shouldn't try to write their own timed trials. Self included.

A system that works as intended but can be cracked by someone willing to try cracking it is better than a system that screws up unexpectedly and punishes legit paying users. And a lot of people trying to roll their own for the first time may not have thought of even the simplest and most obvious workarounds that can be used by non-crackers.

When GMC kids try to design their own security systems I try to talk them out of it. I don't think the majority of RenPy users would be in a good position to build their own security either.

On the other hand, yes, if you do know what you're doing, your own security may be superior to generics.

[Hentai Senshi is too lazy to log back in]

Timed trials are inherently relatively easy to crack; if you do want a timed trial anyhow, you're better off writing it yourself than using a stock trial-maker

... well, if you're comfortable enough with code to do that and make it safe, secure, and reliable. Since game development and application development are generally different skills, there are an awful lot of people who really shouldn't try to write their own timed trials. Self included.

A system that works as intended but can be cracked by someone willing to try cracking it is better than a system that screws up unexpectedly and punishes legit paying users. And a lot of people trying to roll their own for the first time may not have thought of even the simplest and most obvious workarounds that can be used by non-crackers.

Given many of the DRM stories out there, I get the impression that companies making copy protection _also_ often haven't thought of the most obvious workarounds, and that buying someone else's system isn't really a reliable guarantee that it will Work. How much junk is out there that explodes if you don't have IE, or don't have it set to '0wn me plz'? How many commercial DRMs explode under legit users? If you write it yourself at least you understand how it works.

Yes, you're going to need to be reasonably good with code to make any sort of DRM that's not an embarrassment. It's not possible to make a timed trial system truly secure, but not sucking is doable.

When GMC kids try to design their own security systems I try to talk them out of it. I don't think the majority of RenPy users would be in a good position to build their own security either.

Don't the GMC kids usually want to punish warezers by deleting their Windows directory and setting their wallpaper to Goatse? *grin*
(Also, if it's a Ren'Py project, I'm not sure if there are any crossplatform demo-wrappers.)

On the other hand, yes, if you do know what you're doing, your own security may be superior to generics.

[PyTom]

I'll note that if you're using Ren'Py, anything that simply encrypts the executable will not work. That's because there's nothing tying the game directory with a given executable, and you could just put the game directory into a freshly downloaded instance of Ren'Py.

[000]

(dont laugh, I am reeeally clueless here)

Well, if we would not speak abut dedicated-cracker-proof protection (as if someone is truely into it, anything is breakable), but about protection from casual copying, then wouldn't selling unique keys be suffice? I mean, have game generate some code on its first run on given system and soves it somewhere deep into OS, user sends a code and money to you, you generate a matching key and send it to user, user uses it to run his copy of a game. That directly ties copy of a game to the machine, so casual coying will result in game that won't go past key prompt. Downside is, you have to have a list of legitimate users to ensure they can restore the game if their system crashes. But this option should be limited in use, to prevent said user from activizing multiple copies. Also, key generation algorithm should be somewhat intricate.
Pro is -- this system is fairly simple to do and (i think) perfectly doable in Ren'Py.

[Jake]

Pro is -- this system is fairly simple to do and (i think) perfectly doable in Ren'Py.

Unfortunately, the downside is that it doesn't allow the user to move the software from one machine to another without generating a new key, and then you have to start putting arbitrary limits on how many activations a user's allowed before you assume they're generating keys for all their friends as well, and then users are reluctant to buy your software in case they go through n+1 computers before they want to stop using the software.

Not to mention that it's an extra step in between the user wanting to buy your product and getting to use it, which will put some people off buying it in the first place.

...and then - especially if the software is desirable - someone will eventually work out what process you're using to generate keys and write a tool that does the same thing, and distribute it over the web, meaning someone has to jump through fewer hoops to get an illegal copy working than a legal one. Which is bad.

[noaccountyet:o]

...and then - especially if the software is desirable - someone will eventually work out what process you're using to generate keys and write a tool that does the same thing, and distribute it over the web, meaning someone has to jump through fewer hoops to get an illegal copy working than a legal one. Which is bad.

What if you don't have any program to "generate" the keys and just type random crap like; EJDLR2873jjd89j8?

[PyTom]

What if you don't have any program to "generate" the keys and just type random crap like; EJDLR2873jjd89j8?

Well, assume you 26 upper case letters, 26 lower case letters, and 10 numbers to work with. That means with one "digit", you have 62 possible results. With two digits, that goes up to 3844. If we assume 16 digits, a reasonable length for a combination like that, how many combinations do you think there are?

47,672,401,706,823,533,450,263,330,816

That's a lot, and it makes it more likely that you'll be killed by a meteor then crack the code.

[Jake]

That's a lot, and it makes it more likely that you'll be killed by a meteor then crack the code.

On the other hand, if the question is "why not make your game that you're selling accept that random string only as a valid key so people can't use keygens to break it", then that just makes it easier for the people trying to illegally play your game; the game program will have that bit of text in it somewhere, they just have to take it apart and find the 'key'. Not to mention that you have to give it out to customers who've bought it, so once one person leaks it to some sharing site or something everyone can play with the same password.

The relative advantage with generated keys being that you can write the algorithm such that the key only works in conjunction with some other piece of information, like the user's name, so potential illegal users would have to also put up pretending to be "Mr Alec Smart" or whoever.

[josh]

Idea:
Pull a Infocom or a Startropics with DRM.
if the game is shipped, provide a booklet with a bit of info, or something, that to type in in order to proceed in the game.

[papillon]

Now that we have The Internets, booklets don't really help, anyone who feels like distributing their copy will scan and distribute the book as well. It's also annoying if you own the game but have misplaced the booklet that came with it and have to find a pirate site in order to play it...

OTOH, shipping *nice looking* physical stuff with your game can be an incentive to buy a legit copy... but then you start feeling like you're wasting your time making games if people are only paying for the toys.

(There's also the interesting possibility of having a code needed for late in the game sneakily printed ON THE CD. Since people wouldn't automatically think to scan that, and you're less likely to lose the game disc than the booklet, generally.)

[josh]

Put the thing thats on $20 bills on the booklet, you know the thing that makes scanners not work. (Edit: or on the CD)

[Jake]

Put the thing thats on $20 bills on the booklet, you know the thing that makes scanners not work. (Edit: or on the CD)

Uh... I believe the thing that makes money difficult to scan is just that there are patterns printed in the background which certain software recognises and refuses to copy. But then, where are you going to get a digital representation of these patterns from, if you can't scan a $20 bill (or whatever) to get at it? I expect the US treasury isn't going to be rushing to publish high-res images of their patterns publically, because that would just make counterfeiting easier.

Not to mention that you hardly need a perfect high-res scan to read text, where you do need a perfect high-res scan to counterfeit banknotes - so even if you did get hold of these patterns and integrate them into your manual background, people would just be able to make low-res scans or photograph the booklet from a distance and get all of the important information with none of the pattern data Photoshop balks at.



...not to mention that this still doesn't solve the "it's annoying for genuine users" problem. Which, to be honest, is the thing I'd worry about most when choosing a DRM/copy-protection process, because people who want to copy stuff illegally will find a way one way or another regardless of what you do; all you're really doing is putting off casual home copying.