"For video game developers, the CD-ROM was an odd beast. The capacity far exceeded the quantity of assets they were able to produce. A few titles, like 7th Guest (1993) or Phantasmagoria (1995) introduced Full Motion Video (in a world where only part of the screen could be animated). "
------
The title that made me go out and buy a CD-ROM drive was "Wing Commander III", released during the holidays of 1994, a year and a half before Quake 1. This was peak "Silliwood", when games started to use FMV cut-scenes starring bankable actors like Mark Hamill. Games were still being released for the SNES at this point, so imagine the contrast!
In the age before wikipedia or ubiquitous internet access, it was also pretty amazing to have access to a multimedia enhanced encyclopedia. Storing large amounts of data was still a tricky thing. It would take a few years for CD-R's to arrive, and there would be a plethora of competing technologies like Zip drives. Looking back, it seems like computer technology was moving especially fast in the 90's.
> “In the age before … ubiquitous internet access”
For some bizarre reason when upgrading from the family Mac SE to a Quadra, I had an idea in my head cd-rom drive was unnecessary. Thankfully in less than ten years I was proven correct.
The first CD-ROM gaming experience I remember was the Turbo Duo console from late 1991. It came with 2 disks of games. The first and primary disk contained Bonk I, Bonk II, Gates of Thunder, and a hidden Bomberman 3. The second disk contained a combined Y's I and Y's II.
The Y's combo really stood out. You have to understand that late 1991 was still the age of 8-bit NES gaming. This game was very much an 8-bit game that could have run on the NEW and yet it also contained almost DVD quality animated cut-scenes with top quality digital audio. It really felt next level, because the quality of audio was like a studio music album and the animation quality was better than what was on television at the time. It really felt next level compared to the NES.
And yes, I think we both lived through the most rapid period of technological change (so far). Moore's law was real, the improvements each year were both felt and needed, and the software was coming along for the ride.
Quake was huge when it was launched - there'd never been anything like it. Of course, on my crappy Cyrix 6x86 and some similarly lame graphics card, I recall having to play it at something like 15fps. That wasn't great.
When I was in university living in some crappy room at a friend of my parents, it was freezing and there were some transportation issues because of it, so I didn't go to classes and was bored around programming (I wrote Pascal those days for fun and profit), so a struggled to the closest shop to ask what's interesting; Phantasmagoria it was.
Good memories playing that one on that massively heavy monitor that was normal back then, under the blankets with ice on the inside of the windows and on top of the blanket.
Ah Phantasmagoria, that was an interesting case with my CDROM drive at the time, I think it was a Mitsumi one and there was a compatibility issue with specific discs from the 7disc bundle, so I buy the game, going home and start playing with some friends - that was common practice back then - we finish day 1 and I enter the 2nd disc and the FMV is choppy as hell! Try again and again and again and I assume the disc is problematic, so we go back to the shop and they ask what CDROM drive I have, I say Mitsumi and they tell me that it is a known issue with some drives and there is no workaround, so they offer to exchange the game for a different one, oh this "Harvester" game looks intriguing :) Fun times!
I find it so funny that there are new games that are are smaller than Wing Commander 3 despite the 32 year gap. To be fair, they obviously pack their data very differently.
Meh, in the early days of DVDs I recall a game that came with uncompressed 48kHz audio. Compress it to mp3 (which the game naively accepted if you had the correct Windows filters) and it fits in a CD.
Uncompressed audio tracks may have been necessary to get acceptable performance. e.g. GTA 1 came with some PCM tracks on the CD. If you didn't like them, you could pop in a different CD and get different music on the radio in-game. Playing mp3's was, in the early to mid-90's, actually pretty intensive. There were still a decent number of 386's around that could barely run a mp3 player in a Linux console-only setup. You needed a 486 or better to even think about playing mp3's with a full graphical OS running, and probably a Pentium before you have enough CPU left over for the rest of a game.
Even as late as in 2008 I had such a slow machine that when I had to use NetBeans and I wanted to listen to music at the same time, I'd just play music CDs with headphones plugged into the headphones port of the cdrom drive :)
I never noticed any problems, but I did have the mmx version. It wasn't fast but never did the playback stutter , even with dreamweaver and several browser windows
For reference: I had a 486 DX2 66 MHz, and I played MP3s by decoding them in batch to PCM and then playing that as it couldn't really do that in real time. Perhaps only if you didn't do anything else and the MP3 was low bitrate enough and ideally mono.
The turbo button was always on, or it should be. you would turn it off to downgrade your pc to run at the speed of the original IBM PC in the mostly vain hope that whatever old game that was programmed for that exact speed would start to run right. It was slightly be fore my time(we had a 286 with the button, but I was to young to be curious about it) Would the pc take the speed change without problems or was it best to reboot? I sort of hate whatever marketing genius figured out to label it "Turbo" instead of "Slow" But they definitely earned their bonus.
The real weird one from that era was the 7 segment led MHz displays that were popular in the 386/486 era. It was weird because the the chips could not change speed in the first place. And I don't want to say every system was like this, but every one I ever saw was set by a set of static jumpers or a dip switch. When you upgraded your cpu you put the little sticker that came with it on the chassis and changed the jumpers, they served the same purpose.
The way I remember the turbo button it just halved the clock speed when not depressed, but it didn't bring it back to the original IBM PC's speed.
I do remember using it to make older games play less bad, but, as you say, this rarely worked, I think precisely because the speed was never the original one (I don't remember the clock speed of my first PC, but the first one with a turbo button ran at 33 MHz which I'm sure is a lot more than twice the speed of the first one; seems closer to 8x from what a quick search implies).
They're doing that with Blueray games now, writing gigabytes of uncompressed audio for no particular reason other than to use up space and sucker the occasional audiophile.
When Metal Gear Solid 4 released on Ps3, they used uncompressed audio. I think part of it was just to show off how much storage Blu Ray had.
The game Titanfall on PC has the audio in it twice, compressed and decompressed. The decompressed was for machines with dual core or less so that there was enough overhead to keep the game running.
The video game Uru shipped with the sounds as ogg files. Post installation they were decompressed to wav files, I'm guessing so that they can squeeze some more performance out of the game
Given two files, a compressed file which must be uncompressed to play, and an uncompressed one which does not require extra processing - why would the game ever bother uncompressing te compressed file when the same data is already available uncompressed?
With hard disks, I could see reading the compressed version from disk and decompressing giving you the bytes faster than reading the uncompressed version. With SSDs that advantage likely went away.
Chris Roberts is still doing his thing. Elite:Dangerous is still getting updates and there's also Star Citizen/Squadron 42, both of which could either turn out to be epoch-defining or fusterclucks of truly epic proportions.
I did exactly this thing when I was a broke teenager. The files in my ID1 directory that I shuffle around from computer to computer to this day came from that disc 30 years ago. (I still have the disc, but not the case.) I did, however, purchase Quake II and III when they were released. And many years later bought Quake on Steam. I think they got their money's worth from me after all.
(There were some who believed that making the shareware disc easily-crackable was an intentional stroke of genius. It got people who couldn't afford the full retail version to buy the game and expand its popularity. The argument being that $10 was a fairly ludicrous amount of money to pay for a single shareware game. Shareware CDs tended to be $5 on the high end, or free with most computer/gaming magazines.)
> There were some who believed that making the shareware disc easily-crackable was an intentional stroke of genius.
I doubt it. I would believe that the id guys weren't too choked about the lost revenue after the fact, but the business guys probably believed the scheme would work (especially since they contracted a company that said "yes, this will work").
Aside from that, the shareware cd was incredible marketing - the digipak cd case was the kind of art and paper construction that nin was using on their discs (like the Broken EP), and the disc itself has nice art with the four logos in cardinal directions. I bought a copy of the registered disc from ebay years later and it's in a jewel case with really crappy artwork vs the shareware version, it was a surprising disappointment.
Surely anybody who could download the keygen could just as easily download the pirated game? Only difference is time and dialup costs, but neither of those would’ve pushed somebody into the buy category.
If I remember correctly, actual pirated copies would have been maybe tens of megabytes, since they didn't include the CD audio. But that would still have been a hefty modem download.
Oh, but that's why you went to the computer lab of your uni - those machines were on a 10mbit LAN and had a decent connection to the internet!
Although you still wouldn't download a CD, simply because... you wouldn't have been able to take it home with you! Hard drives were in many cases smaller than a single CD. Anything more than a few 10s of MBs would have been considered excessive.
Connecting to office LANs and going on the internet was a magical feeling back then. The speed coming from a 56k dialup was jarring. We'd have network games of Quake (rocket jumps included) and watch Real Player versions of grainy South Park early episodes at er, lunch. Good times.
>Only difference is time and dialup costs, but neither of those would’ve pushed somebody into the buy category.
well, if dialup cost more than just buying the game, it makes sense to buy it rather than pirate. also, if you care about audio then you'll get the cd, becuse downloading 400MB of audio over dialup isn't fun.
Dial-up cost will be discovered later by upset parents, and you already have all your nice pirates games, which you would need to pay upfront with your own pocket money (if that existed).
But in any case physical copy + crack was the best solution at the time.
Glad I am not the only one who carts an id1 directory around. Mine originally came from a retail disk with a funny story, I had bought it for a friend. a birthday present. and a couple years later I bought their old pc from them and that quake disk was in the box of parts. It had came back me.
This strategy makes sense in developing markets. People might be too poor to buy your game right now so you distribute it for free hoping that in the future they will get rich and then they'll actually buy your newer game.
It’s worth getting the Quake shareware disc for the NIN soundtrack. Official track names are now available following the vinyl release a few years ago. This is the only CD release of the soundtrack. Just don’t forget to skip track 1.
There is very slightly more to it than just the mixed-mode sessions: the audio tracks on the Quake CD were mastered with pre-emphasis, so your ripper or player needs to apply corresponding de-emphasis EQ in order for them to sound correct. 1996 was actually a strangely late year to find any CD mastered with pre-emph; it was mostly an '80s thing: https://www.studio-nibble.com/cd/index.php?title=Pre-emphasi...
On top of being a technique that had fallen out of common use by that time, Quake has the pre-emphasis metadata flagged in the subcodes but not flagged in the TOC, so some software players/rippers like (annoyingly) Exact Audio Copy won't heed it: https://wiki.hydrogenaudio.org/index.php?title=Comparison_of...
And if you're curious, here's the pre-emphasis subcode channel Q flag and pre-emphasis frequency response as described in the Red Book standard, and sox_ng's code for undoing it:
Although if someone has been unknowingly listening to the pre-emphasized tracks for thirty years, maybe the ‘incorrect’ one is what sounds right to them? :)
IIRC, it's actually debatable if the pre-emphasis flag is supposed to be set. I don't remember the exact details anymore, but I think it was discovered that there multiple releases with bit-identical audio tracks, but some of them have the flag set while others don't. They can't both be correct, because the audio data should've changed if the emphasis was removed.
Make sure that it isn't a case of the software you're using to inspect the disc intentionally looking the other way with regard to that flag.
> Quake has the pre-emphasis metadata flagged in the subcodes but not flagged in the TOC, so some software players/rippers like (annoyingly) Exact Audio Copy won't heed it
> For a short while, up to version 0.95b3 [in 2004], Exact Audio Copy had a manual function to actually scan the audio area of the disc also [talking about subcode channel Q versus the TOC] to find if tracks used pre-emphasis. This was removed due to some perceived legal threat. Now, basically, a standards-compliant disc that doesn't use the TOC kludge won't show "pre-emphasis"
The adjacent bit is “copy prohibited” versus “copy permitted”, which is not a piece of metadata I knew Compact Disc had. Makes a lot more sense why EAC would remove that feature in the legal climate of the P2P era, because to look for one flag also must look for the other, and this way it's out of sight and out of mind and Gee Your Honor We Didn't Know You Couldn't Do That if someone uses EAC to rip a “copy prohibited” disc. I wish I could get this feature back in 2026 though :p
I actually remembered more about this after my initial post. I ended up with two copies of Quake (one is the original shareware release and one is a later release of the full game), and one of them has the PRE flag set in the TOC and the other one doesn't. Now I'm wondering if I ever checked the subcodes for the one that's missing the flags in the TOC. I'll have to dig it out sometime and check.
I did verify that the audio is identical down to an offset, though, between the two.
> The adjacent bit is “copy prohibited” versus “copy permitted”, which is not a piece of metadata I knew Compact Disc had.
I wonder if any equipment ever actually used those other flags for anything?
Now I'm curious if it's possible to make a data track that has at least some resemblance to actual music. That could be an interesting challenge if it can't obviously be ruled out.
In each sector 2048 of the data bits are just bits, but each data sector also includes 12 bytes of sync, 4 bytes of metadata and 288 bytes of ECC that are not really controllable (if you want the disc to be readable). Then there’s the ISO 9660 on top that will impact the first few sectors. If you have a single file (say a zip/tarball/rar, image) you could limit that. Assuming you could mask the noise from the ECC at 75hz it might be possible.
That's how I tentatively "imitated" Aphex Twin and Autechre when I was a teenager using impulse tracker. Selecting random files as samples until I would find a section that sounded the way I wanted.
I remember I had the CD but rarely used it because having to pop in the CD just to play the game was annoying. So, I mostly played it without the sound track. Also, I never was that much into nine inch nails. Just not my genre. But my memory of the game is that it was fine without music.
"The CD was announced[4] on July 3, 1996 and released on August 30th[5]. The hacker group GNOMON released Quakecrk.zip only 39 days later[6]."
While not being familiar with the Windows side of things, I did frequent the a.b.m.a newsgroup being a broke student. It was always impressive to me to see how quickly someone would post a binary before the crack was posted. I remember one app I was very interested in posting on Friday, and the crack available when I checked back on Monday. Even apps that needed a physical dongle for the ADB port were cracked and not just the apps that needed a key typed in to unlock. Later, the apps that would clone a USB dongle came along, and one software company that used USB dongles actually advised using this tool when we started running their software on Blade servers that did not support adding 24 USB devices.
This was me as well. +HCU and searchlores.org were so much fun. The coolest thing I remember was the hash-maze but I couldn't find it again. I saw Fravia live in 1999 :D
Switching 0x74 to 0x75 and vise versa lead to me ordering the free Intel books and diving into assembly and osdev for a while. There was something zen-like wading through disassembled code with a complete beginner mind and looking for patterns. This and reading Phrack around the "Smashing the Stack for Fun and Profit" times are probably my favorite memories related to computing because I only had a very vague understanding of what was going on mixed with a sense of "wait you can do this" and a lack of readily available information.
And when the crack came out on the same day as the CD release it was called a zero-day. A term that has long since lost ita original meaning, in fact it has been inverted.
I remember old games shipping with huge lists of codes on paper. Paper made, on purpose, too big for an US legal/A4 photocopier. And moreover the codes would be using "light blue characters on slightly less light blue paper". Very annoying for casual piracy (as in: copying your neighbor's version of the game).
> Later, the apps that would clone a USB dongle came along, and one software company that used USB dongles actually advised using this tool when we started running their software on Blade servers that did not support adding 24 USB devices.
The sentence "This created an only-too-familiar situation where illegitimate users enjoyed a better experience than paying customers." is a direct reference to these system which annoyed people who actually paid for their game!
Siemens for some software uses a license system that uses "secret" files stored on "dongle" that is just USB stick, and software that manages copying to hard drive etc.
I have heard of several cases where Siemens technicians and support staff would use a well-known "crack" (actually more of a keygen combined with license file manager) or guide customers how to obtain and use it...
Remember, legally (usually), the really important part is the proof of contract and payment, not exact way the software was activated
> The packaging was actually pretty high quality for a shareware product. On my copy, a sticker on the front clarifies this is the "Shareware version" with instructions to call 1-800-669-9342 (or 1-800-ID-GAMES) to unlock the full game.
> The phone number is still active today. However you don't reach the unlock center since CompUSA went out of business. The Superstore chain closed between 2007 and 2008. Instead of an unlock operator we get an automated message to sell us elderly stuff.
I remember how stuff like that presented an insurmountable obstacle as a kid. Buying stuff in the store, sure, but calling phone numbers or writing letters to companies? No way.
I like how it has come full circle today, with the phone number being inaccessible once again...
> By June 1996, after three years of hard work, id Software had completed their next title, Quake. As for their previous title, they were going to release both a shareware version and a full version of their game. Since it used a mere 22 MiB of storage, people at id Software had the idea of leveraging the remaining capacity of a CD-ROM. Why not include encrypted versions of the full id catalogue of games? Not only this would cut out the middlemen, it would give instant access to gamers with a simple phone call and a credit card.
That wasn't a new idea at the time. In 1994 Apple distributed a "software store" CD-ROM called Software Dispatch that implemented the same idea: https://68kmla.org/bb/threads/apple-software-dispatch-from-1.... The thing was full of demos, and if you wanted to buy the full version, you just had to call some number to get an unlock code.
I think I had at least one other CD-ROM that implemented the same concept. I'm sure there were tons of others.
As soon as I saw "Quake Shareware" in the title of this post, I thought "Pray to the one you will pay!" which was what appeared on the screen when you ran QCRACK.
I always wondered why Final Doom didn't unlock with QCRACK - I always assumed it was a bug with the keygen, never realized it wasn't possible to purchase it from the disk at all!
From the title I was expecting this to be about them packing 700MB of data on the CD, a common thing at the time by spacing the tracks slightly closer together, as some CD-Rs did.
How would the challenge/serial have worked if implemented correctly?
Since the challenge rotate every 5 minutes, there must be some way for a user to enter against an old challenge, right? Else they call, get a code and nothing.
But even with that out of the way, the decryption key must be the same across all discs, since there is only 1 master disc.
So the challenge/serial must combine to create a decryption key, which can be shared across users.
I remember Microsoft had a time limited version of Office on cd-rom that was sold for a couple bucks at Kinko's. You could replace a .dll in it and that removed the time limit.
There was something similar (but I can't recall what) where you just set your clock far in the future (like 2010!) and then the trial would last that long + one month.
Man, those CompUSA and Computer City pictures take me back. Spent too much time there (when the 7300 was my regular Mac, the CompUSA "Apple store" was my primary source of software). Probably still have some of those free Computer City floppies in a drawer somewhere.
Yes, the game just played it as a regular CD. In fact, if you put a different disc in the drive it would play the tracks from there instead. There were also console commands to change the track, stop the music, loop, or even reload the disc information if you needed to refresh it after swapping CDs.
Yeah, and I remember if you didn't have the audio cable that went from the CDROM into your sound card, you would never hear the tracks. It wasn't your computer decoding the music, the CDROM drive was simply doing the decoding, and passing the audio out into a CDROM input (just a line-level audio signal) on your sound card.
I'd forgotten about those little cables. If I recall correctly, they were small round cables, with maybe four conductors and a little flat connector with a clip?
No. That cracking group would have known a priori the phone call was a pretense, because id was very unlikely to press distinct CDs. So they could focus on conventional cracking techniques.
Maybe have n versions of the CD (maybe 16, or 32), so there are n decryption keys. When the buyer recites the code over the phone, the program run by the salesperson can identify which key is used based on the number.
The hacker groups would be a little frustrated trying to find all of the different CDs.
The setup costs for a pressed CD are substantial - manufacturing many different designs would have been much more expensive, and would have only slowed down crackers marginally. It's unlikely that this would have recovered enough sales to be worth it.
A reminder that CD-ROMs are pressed from glass masters, not burned like CD-Rs. I wonder if anyone has tried something like this with burned discs, though.
The multisession standard actually afforded the possibility of CD-PROM; a hybrid disc with a pressed read-only session followed by a writable section. It was only used for Kodak Picture CD (not the same as Photo CD) to hold software in the pressed session. It also apparently standardized a magneto-optical hybrid that never made it to market.
Hmm, it could work then if they’ve pressed most of the game data first and then recorded the encrypted part for each disk. But I guess the tooling for such setup would still be too expensive.
With CD-Rs it was already possible in the mid '90s, if we're talking mass production. CD-Rs are a little more expensive to produce, but not much more so. A slight concern is that readers of the time might not have been able to read the discs, because CD-Rs are not quite as reflective as CD-ROMs.
You would need some sort of hardware serial number system for PCs and a whole security infrastructure to verify that the serial number being used for the unlock wasn't faked. Then you could do per-PC unlocks. I'm not aware of anything like that being widespread back in the Quake days, though these days every computing device you own probably has a unique ID burned into ROM somewhere.
"Never attribute to malice what you can attribute to stupidity. And never attribute to stupidity what you can attribute to time pressure. -- Fab's Razor".
Surfacing this gem of a razor here, as it's buried deep in the (fantastic) blog post. Indeed, all too often, what we attribute simply to stupidity/ignorance, as per the classic Hanlon's razor, is actually downstream of market/client/management/deadline etc. pressure.
Am I reading this right, and was only the file header encrypted?
Also, wasn't ID the company that did not use serials or other DRM? Maybe that explains their naivite, maybe they underestimated how easily games get cracked.
Not sure about the early days - I got a Quake II CD in the early 2000s that didn't need a CD key. Doom 3 and Quake 4 required CDs, at least at release, but after a little while no CD patches were released.
But it sounds like it wasn't Testdrive's fault; iD implemented it incorrectly. Their phone support weren't issuing the decrypt code, just a simple checksum:
> Described as is, there is no flaw in this process. The secret seed comes from the unlock server, tied to a CHALLENGE/SERIAL that could not be reused. But the hacker team GNOMON found a way.
> The QUAKE unlock program FLOW.EXE that ships on the CD is capable of generating the SERIAL from the CHALLENGE on its own. All it does is check that its own locally-generated SERIAL and the SERIAL entered by the user match! The entire protection mechanism relies on security by obscurity.
If the phone support issued the decrypt code, you could just replay that same code for every CD stamped from the same master.
If phone support issued an encrypted decrypt code that could only be used with your challenge code to decrypt the decrypt code, replay wouldn't be as trivial.
Every CD is identical, so no matter how iD went about this there's only ever one decryption key (per title, I assume) and those keys must either have been encoded in iD support's response, or already stored on the CD. TestDrive sold iD on the notion that the process was too hard for hacker groups to reverse engineer, and it wasn't.
Those photos of the inside of CompUSA are just stunning. Bright sodium lighting, not a dim dank cave (one of the benefits of CRTs over modern OLED screens and their color range from black to slightly-less-black). Colorful packaging everywhere (this was before fun was made illegal in 2012). Big signage, so you know where things are (the idea of a customer knowing where to find something without being diverted to an antivirus upsell by a "helpful" "associate" didn't bother MBA CEOs back then. Nowadays the idea would cause them them to spontaneously combust).
> Upon contacting the operator, users were to also communicate a "SOURCE CODE". It played no part in generating the Unlock code. It may have been a way for the distributors to claim a transaction fee. Browsing eBay, I found many with names indicative of past/present retailers. 12-BSTBY BestBuy, 24-CCITY Computer City, 22-CUSA CompUSA, 88, 11-1111, 38-EB Electronic Boutique, 44-FTRSP Future Shop (Canada), 34-EGGH Egghead Software, and 56-MCTR Media Play/ Musicland.
Why do we think that CCITY is a code for Computer City and not Circuit City?
There was a lot less competition in the digital entertainment industry back then. Nowadays you still have AAA games going for $80 in a much more crowded market.
This screw up ended up costing id Software millions in sales. With the Doom franchise they only included the first episode in the shareware version. But to save a buck I guess they decided this was better than mailing a second CD with the full game on it. Then doubling down on that greed by putting their entire game library on the disc. This is right about the time John Romero and John Carmack parted ways, can't help but wonder if this debacle played a role. And by all accounts Quake's development was painful, so imagine this happens after going through hell getting a game out the door.
COD:BO7's single-player uses dedicated servers. Terrible, yes, but that's as foolproof as it gets when it comes to piracy. They also probably did this to combat all the 'services' offering to unlock camos/cosmetics that you'd normally have to grind for; players would give someone their credentials (probably stolen later and sold to cheaters) and the person on the other side would then use a hacked client to trick the backend into thinking they completed all the challenges. With a server-side campaign, that probably got much more time consuming, I reckon.
Every game separately encrypted then compressed into solid archives. Pressed in a way every game lands on different precise height of CD, padded before compression if necessary so it takes enough space on CD. Separate machine burning tiny holes in CD from the back, but not so tiny that CIRC could fixes them. At least one hole per file.
Install locates the damage and gives you code encoding locations. Over the phone You receive ECC code. No more magic keygens, pirates have to distribute whole multi megabyte games.
I think most (all?) Steam games link steam_api.dll (or something like that) which somehow figures out the running Steam process and thus the Steam ID, and then does a HTTP API request to check that you own the title.
Like 15 years ago Steam launched CEG (Custom Executable Generation) - I remember reading a write-up a while ago, but each user had a customer launcher generated on install I think, and I think there was an extra segment put into the .exe file and that was put to use somehow, though I don't know if CEG is still a thing.
The ownership check you are referring to is a local check. There is another check (involving EncryptedAppTickets) which is not local, but far fewer games use that.
CEG is long dead/was broken immediately in a pragmatic, automatable way.
Is it really fair to call these such algorithms "security through obscurity"?
By that logic, wouldn't real symmetric crypto also be considered the same, since it's still just math that similarly swaps the data with other bytes, based on an "obscured" password or key?
I guess they could have done a public/private key pair in 1996, but doing that over a phone call was going to be painful. A simple password read over the phone and verified by the software was the best solution at the time.
Mind, this was also intended for a world where internet connectivity was not to be expected. The majority of PC owners were not online.
How would your scheme work? You use RSA to encrypt the entire thing? Or you use it to get a key for a block cipher and then use that?
I feel like either way, you'd need the key to be different for every CD, otherwise you could just share the shared key. But if the encryption of every CD is different, why not just share the block cipher key directly? They can have a list of CDs and the associated key(s).
What additional security or functionality does an RSA like step add here?
Sneakernet was common also. Say you have a class of 30 kids, and one of them has access to Usenet to grab qcrack.zip. He'll put it on a floppy disk and share ;)
If I ask you to pass on a message from me to someone else, and the message is encrypted using a symmetric key known to me, and to the recipient, but not to you, there's no security-by-obscurity in place.
If I ask you to pass on a message to someone, and the message is encrypted using a symmetric key known only to me, and I expect the recipient to be able to decrypt it because he knows that the key is written on the back of the paper, that's security by obscurity. You're just as able to do the decryption as the intended recipient is; I'm just hoping that you won't realize that.
That was an image I found in Google Image search. I should have done a better job to research the provenance and I appreciate being called on it. I have updated the photo.
Kind of sounds like the low‑background steel [1] problem all over again.
I have a rather badly written tool [2] I made about 5–6 years ago. If you connect it to your Google Custom Search Engine (along with the appropriate CSE keys), you can quickly search, copy, and download images from before 2022. It also has a ban list so you can filter out Shutterstock, Alamy, and other heavily watermarked sources.
The app is a hot circle of garbage code, but it might still be useful as a starting point.
I fed the image into Gemini and it could not detect that it’s legit or AI, but in addition to the wrong store name, the other banner that’s supposed to day SOFTWARE says TWARE or something like that. It sure seems like AI to me.
Okay, I haunted CompUSA in my teen years and I purchased my first 386 system there. So the photograph included in this blog seems bona fide.
Except their logo or name was never "Compu USA". Where in the world did they get a store hung with signs that all say "Compu USA"? Is this some Mandela Effect that has me in its clutches? The business was definitely never named or branded that way. I did my Wikipedia and Bing Images research. There are no serious search hits. Gemini is giving some truly b.s. explanations that I do not believe.
Is this a concrete example of a USA store tolerating mis-branded signage... inside their own store? Or is it some third-world knockoff? The merch on the shelves, the bros shopping for themselves, the tie-wearing salesmen, they all seem totally legit. What kind of big-box computer store lets that happen to their signage?
I made a mistake when I researched photos for it and found "one" without sources. You are right this photo looks shady. I have replaced it (with source).
"For video game developers, the CD-ROM was an odd beast. The capacity far exceeded the quantity of assets they were able to produce. A few titles, like 7th Guest (1993) or Phantasmagoria (1995) introduced Full Motion Video (in a world where only part of the screen could be animated). "
------
The title that made me go out and buy a CD-ROM drive was "Wing Commander III", released during the holidays of 1994, a year and a half before Quake 1. This was peak "Silliwood", when games started to use FMV cut-scenes starring bankable actors like Mark Hamill. Games were still being released for the SNES at this point, so imagine the contrast!
In the age before wikipedia or ubiquitous internet access, it was also pretty amazing to have access to a multimedia enhanced encyclopedia. Storing large amounts of data was still a tricky thing. It would take a few years for CD-R's to arrive, and there would be a plethora of competing technologies like Zip drives. Looking back, it seems like computer technology was moving especially fast in the 90's.
> “In the age before … ubiquitous internet access”
For some bizarre reason when upgrading from the family Mac SE to a Quadra, I had an idea in my head cd-rom drive was unnecessary. Thankfully in less than ten years I was proven correct.
Obviously I didn’t play enough games.
The first CD-ROM gaming experience I remember was the Turbo Duo console from late 1991. It came with 2 disks of games. The first and primary disk contained Bonk I, Bonk II, Gates of Thunder, and a hidden Bomberman 3. The second disk contained a combined Y's I and Y's II.
The Y's combo really stood out. You have to understand that late 1991 was still the age of 8-bit NES gaming. This game was very much an 8-bit game that could have run on the NEW and yet it also contained almost DVD quality animated cut-scenes with top quality digital audio. It really felt next level, because the quality of audio was like a studio music album and the animation quality was better than what was on television at the time. It really felt next level compared to the NES.
Myst was the CD-ROM killer game for me.
And yes, I think we both lived through the most rapid period of technological change (so far). Moore's law was real, the improvements each year were both felt and needed, and the software was coming along for the ride.
Quake was huge when it was launched - there'd never been anything like it. Of course, on my crappy Cyrix 6x86 and some similarly lame graphics card, I recall having to play it at something like 15fps. That wasn't great.
..so.. ..you know about RealMyst, right? A 3d first-person adaptation of the original Myst?
What's the point you're trying to make?
When I was in university living in some crappy room at a friend of my parents, it was freezing and there were some transportation issues because of it, so I didn't go to classes and was bored around programming (I wrote Pascal those days for fun and profit), so a struggled to the closest shop to ask what's interesting; Phantasmagoria it was.
Good memories playing that one on that massively heavy monitor that was normal back then, under the blankets with ice on the inside of the windows and on top of the blanket.
Ah Phantasmagoria, that was an interesting case with my CDROM drive at the time, I think it was a Mitsumi one and there was a compatibility issue with specific discs from the 7disc bundle, so I buy the game, going home and start playing with some friends - that was common practice back then - we finish day 1 and I enter the 2nd disc and the FMV is choppy as hell! Try again and again and again and I assume the disc is problematic, so we go back to the shop and they ask what CDROM drive I have, I say Mitsumi and they tell me that it is a known issue with some drives and there is no workaround, so they offer to exchange the game for a different one, oh this "Harvester" game looks intriguing :) Fun times!
I find it so funny that there are new games that are are smaller than Wing Commander 3 despite the 32 year gap. To be fair, they obviously pack their data very differently.
Meh, in the early days of DVDs I recall a game that came with uncompressed 48kHz audio. Compress it to mp3 (which the game naively accepted if you had the correct Windows filters) and it fits in a CD.
Uncompressed audio tracks may have been necessary to get acceptable performance. e.g. GTA 1 came with some PCM tracks on the CD. If you didn't like them, you could pop in a different CD and get different music on the radio in-game. Playing mp3's was, in the early to mid-90's, actually pretty intensive. There were still a decent number of 386's around that could barely run a mp3 player in a Linux console-only setup. You needed a 486 or better to even think about playing mp3's with a full graphical OS running, and probably a Pentium before you have enough CPU left over for the rest of a game.
Even as late as in 2008 I had such a slow machine that when I had to use NetBeans and I wanted to listen to music at the same time, I'd just play music CDs with headphones plugged into the headphones port of the cdrom drive :)
CDROM and 386 would have been an odd combination though?
My 486 had a single speed CDROM but it only had a 500mb and a 100mb hard drive, so still wasn't playing mp3s
Back in the day, my 200mhz Pentium remained just-barely usable while listening to WinAmp, but there was noticeable latency.
Yeah, I remember an mp3 rating >50% CPU time on one of those early pentiums we had
Yeah. Seeing that made me very happy, being the proud owner of a dual CPU (no, not dual core!) AMD K6. :-)
I never noticed any problems, but I did have the mmx version. It wasn't fast but never did the playback stutter , even with dreamweaver and several browser windows
For reference: I had a 486 DX2 66 MHz, and I played MP3s by decoding them in batch to PCM and then playing that as it couldn't really do that in real time. Perhaps only if you didn't do anything else and the MP3 was low bitrate enough and ideally mono.
MP3 decoders have, thankfully, gotten faster since the 90s. But yes, a 66MHz is pushing it indeed.
And don't forget to engage your "TURBO" button, to make the playback slightly better!
The turbo button was always on, or it should be. you would turn it off to downgrade your pc to run at the speed of the original IBM PC in the mostly vain hope that whatever old game that was programmed for that exact speed would start to run right. It was slightly be fore my time(we had a 286 with the button, but I was to young to be curious about it) Would the pc take the speed change without problems or was it best to reboot? I sort of hate whatever marketing genius figured out to label it "Turbo" instead of "Slow" But they definitely earned their bonus.
The real weird one from that era was the 7 segment led MHz displays that were popular in the 386/486 era. It was weird because the the chips could not change speed in the first place. And I don't want to say every system was like this, but every one I ever saw was set by a set of static jumpers or a dip switch. When you upgraded your cpu you put the little sticker that came with it on the chassis and changed the jumpers, they served the same purpose.
The way I remember the turbo button it just halved the clock speed when not depressed, but it didn't bring it back to the original IBM PC's speed.
I do remember using it to make older games play less bad, but, as you say, this rarely worked, I think precisely because the speed was never the original one (I don't remember the clock speed of my first PC, but the first one with a turbo button ran at 33 MHz which I'm sure is a lot more than twice the speed of the first one; seems closer to 8x from what a quick search implies).
mp2's files were almost usable on 386 but they used tons of space.
They're doing that with Blueray games now, writing gigabytes of uncompressed audio for no particular reason other than to use up space and sucker the occasional audiophile.
When Metal Gear Solid 4 released on Ps3, they used uncompressed audio. I think part of it was just to show off how much storage Blu Ray had.
The game Titanfall on PC has the audio in it twice, compressed and decompressed. The decompressed was for machines with dual core or less so that there was enough overhead to keep the game running.
The video game Uru shipped with the sounds as ogg files. Post installation they were decompressed to wav files, I'm guessing so that they can squeeze some more performance out of the game
Why would it play the compressed version if it had the decompressed audio right there though?
Because with lossless compression you can save a lot of storage space without losing any quality. You are thinking about lossy compression.
I think you misunderstood his question.
Given two files, a compressed file which must be uncompressed to play, and an uncompressed one which does not require extra processing - why would the game ever bother uncompressing te compressed file when the same data is already available uncompressed?
There's no true use case for that that I can see.
With hard disks, I could see reading the compressed version from disk and decompressing giving you the bytes faster than reading the uncompressed version. With SSDs that advantage likely went away.
I would guess to save on installed size?
Ripper (1996) with Christopher Walken and Paul Giamatti came on 6 CDs
Also https://en.wikipedia.org/wiki/A_Fork_in_the_Tale - not a good game, but an impressive number of CDs at 5
I miss the entire WC series... alot. I wish we could get something like that again today.
Chris Roberts is still doing his thing. Elite:Dangerous is still getting updates and there's also Star Citizen/Squadron 42, both of which could either turn out to be epoch-defining or fusterclucks of truly epic proportions.
The latter.
Thank you for playing Wing Commander!
The most polite error message ever
I did exactly this thing when I was a broke teenager. The files in my ID1 directory that I shuffle around from computer to computer to this day came from that disc 30 years ago. (I still have the disc, but not the case.) I did, however, purchase Quake II and III when they were released. And many years later bought Quake on Steam. I think they got their money's worth from me after all.
(There were some who believed that making the shareware disc easily-crackable was an intentional stroke of genius. It got people who couldn't afford the full retail version to buy the game and expand its popularity. The argument being that $10 was a fairly ludicrous amount of money to pay for a single shareware game. Shareware CDs tended to be $5 on the high end, or free with most computer/gaming magazines.)
> There were some who believed that making the shareware disc easily-crackable was an intentional stroke of genius.
I doubt it. I would believe that the id guys weren't too choked about the lost revenue after the fact, but the business guys probably believed the scheme would work (especially since they contracted a company that said "yes, this will work").
Aside from that, the shareware cd was incredible marketing - the digipak cd case was the kind of art and paper construction that nin was using on their discs (like the Broken EP), and the disc itself has nice art with the four logos in cardinal directions. I bought a copy of the registered disc from ebay years later and it's in a jewel case with really crappy artwork vs the shareware version, it was a surprising disappointment.
Surely anybody who could download the keygen could just as easily download the pirated game? Only difference is time and dialup costs, but neither of those would’ve pushed somebody into the buy category.
it would have been like 520 hours to download 500mgb. Assuming it didn't break at some point.
That was pretty tough in that day and age. Especially from pirate sites.
mIRC had some good servers but getting someone to keep a direct connection with you for that long was problematic too.
And then if you didn't have a dedicated phone line for the modem...
If I remember correctly, actual pirated copies would have been maybe tens of megabytes, since they didn't include the CD audio. But that would still have been a hefty modem download.
I got 3kB/sec, or 6 minutes per meg in ‘96. 10 meg an hour wasn’t terrible even I. The U.K. where you paid a penny a minute.
That's why you had a download manager and left the machine running all night.
Correct
Oh, but that's why you went to the computer lab of your uni - those machines were on a 10mbit LAN and had a decent connection to the internet!
Although you still wouldn't download a CD, simply because... you wouldn't have been able to take it home with you! Hard drives were in many cases smaller than a single CD. Anything more than a few 10s of MBs would have been considered excessive.
Connecting to office LANs and going on the internet was a magical feeling back then. The speed coming from a 56k dialup was jarring. We'd have network games of Quake (rocket jumps included) and watch Real Player versions of grainy South Park early episodes at er, lunch. Good times.
According to the article: the game was 22MB and only the first 32kB of the .exe were encrypted.
>Only difference is time and dialup costs, but neither of those would’ve pushed somebody into the buy category.
well, if dialup cost more than just buying the game, it makes sense to buy it rather than pirate. also, if you care about audio then you'll get the cd, becuse downloading 400MB of audio over dialup isn't fun.
compared to that a crack is just a single binary.
Dial-up cost will be discovered later by upset parents, and you already have all your nice pirates games, which you would need to pay upfront with your own pocket money (if that existed).
But in any case physical copy + crack was the best solution at the time.
Glad I am not the only one who carts an id1 directory around. Mine originally came from a retail disk with a funny story, I had bought it for a friend. a birthday present. and a couple years later I bought their old pc from them and that quake disk was in the box of parts. It had came back me.
This strategy makes sense in developing markets. People might be too poor to buy your game right now so you distribute it for free hoping that in the future they will get rich and then they'll actually buy your newer game.
It’s worth getting the Quake shareware disc for the NIN soundtrack. Official track names are now available following the vinyl release a few years ago. This is the only CD release of the soundtrack. Just don’t forget to skip track 1.
> Just don’t forget to skip track 1.
There is very slightly more to it than just the mixed-mode sessions: the audio tracks on the Quake CD were mastered with pre-emphasis, so your ripper or player needs to apply corresponding de-emphasis EQ in order for them to sound correct. 1996 was actually a strangely late year to find any CD mastered with pre-emph; it was mostly an '80s thing: https://www.studio-nibble.com/cd/index.php?title=Pre-emphasi...
On top of being a technique that had fallen out of common use by that time, Quake has the pre-emphasis metadata flagged in the subcodes but not flagged in the TOC, so some software players/rippers like (annoyingly) Exact Audio Copy won't heed it: https://wiki.hydrogenaudio.org/index.php?title=Comparison_of...
Here's a side-by-side comparison — my EAC rip of my MacSoft version of Quake (left) alongside the output of `sox_ng deemph` for the same file (right): https://i.ibb.co/9m18CSsk/Emi-2026-08-17-18-52-48.png
And if you're curious, here's the pre-emphasis subcode channel Q flag and pre-emphasis frequency response as described in the Red Book standard, and sox_ng's code for undoing it:
― https://audio-union.com/wp-content/uploads/2023/02/Red-Book-...
― https://audio-union.com/wp-content/uploads/2023/02/Red-Book-...
― https://codeberg.org/sox_ng/sox_ng/src/branch/main/src/deemp...
Although if someone has been unknowingly listening to the pre-emphasized tracks for thirty years, maybe the ‘incorrect’ one is what sounds right to them? :)
IIRC, it's actually debatable if the pre-emphasis flag is supposed to be set. I don't remember the exact details anymore, but I think it was discovered that there multiple releases with bit-identical audio tracks, but some of them have the flag set while others don't. They can't both be correct, because the audio data should've changed if the emphasis was removed.
Make sure that it isn't a case of the software you're using to inspect the disc intentionally looking the other way with regard to that flag.
> Quake has the pre-emphasis metadata flagged in the subcodes but not flagged in the TOC, so some software players/rippers like (annoyingly) Exact Audio Copy won't heed it
I learned that this is an intentional deficiency in EAC: https://forums.stevehoffman.tv/threads/eac-not-detecting-pre...
> For a short while, up to version 0.95b3 [in 2004], Exact Audio Copy had a manual function to actually scan the audio area of the disc also [talking about subcode channel Q versus the TOC] to find if tracks used pre-emphasis. This was removed due to some perceived legal threat. Now, basically, a standards-compliant disc that doesn't use the TOC kludge won't show "pre-emphasis"
And sure enough, the Red Book standard mentions usage of the other three flag bits alongside the pre-emphasis flag: https://audio-union.com/wp-content/uploads/2023/02/Red-Book-...
The adjacent bit is “copy prohibited” versus “copy permitted”, which is not a piece of metadata I knew Compact Disc had. Makes a lot more sense why EAC would remove that feature in the legal climate of the P2P era, because to look for one flag also must look for the other, and this way it's out of sight and out of mind and Gee Your Honor We Didn't Know You Couldn't Do That if someone uses EAC to rip a “copy prohibited” disc. I wish I could get this feature back in 2026 though :p
I actually remembered more about this after my initial post. I ended up with two copies of Quake (one is the original shareware release and one is a later release of the full game), and one of them has the PRE flag set in the TOC and the other one doesn't. Now I'm wondering if I ever checked the subcodes for the one that's missing the flags in the TOC. I'll have to dig it out sometime and check.
I did verify that the audio is identical down to an offset, though, between the two.
> The adjacent bit is “copy prohibited” versus “copy permitted”, which is not a piece of metadata I knew Compact Disc had.
I wonder if any equipment ever actually used those other flags for anything?
> I wonder if any equipment ever actually used those other flags for anything?
MiniDisc did. You could make one (digital) copy but not two.
The data track never sounds very good. A little too industrial for my tastes.
Now I'm curious if it's possible to make a data track that has at least some resemblance to actual music. That could be an interesting challenge if it can't obviously be ruled out.
In each sector 2048 of the data bits are just bits, but each data sector also includes 12 bytes of sync, 4 bytes of metadata and 288 bytes of ECC that are not really controllable (if you want the disc to be readable). Then there’s the ISO 9660 on top that will impact the first few sectors. If you have a single file (say a zip/tarball/rar, image) you could limit that. Assuming you could mask the noise from the ECC at 75hz it might be possible.
Sounds like the sort of thing Aphex Twin would do.
That's how I tentatively "imitated" Aphex Twin and Autechre when I was a teenager using impulse tracker. Selecting random files as samples until I would find a section that sounded the way I wanted.
I think it's the sort of thing Master Boot Record should do
Honestly that'd probably appeal to Reznor's aesthetic.
Mind that every Quake release includes the same soundtrack, it's not just the shareware version ;)
There's millions of such CDs in the wild, both shareware and fully purchased copies.
Oops, I can see how that reads like only the shareware version is this way. It’s often cheaper if you’re just going for the music.
I remember I had the CD but rarely used it because having to pop in the CD just to play the game was annoying. So, I mostly played it without the sound track. Also, I never was that much into nine inch nails. Just not my genre. But my memory of the game is that it was fine without music.
NIN and iD (re-?)released a vinyl record set for the 30th anniversary a couple weeks ago.
"The CD was announced[4] on July 3, 1996 and released on August 30th[5]. The hacker group GNOMON released Quakecrk.zip only 39 days later[6]."
While not being familiar with the Windows side of things, I did frequent the a.b.m.a newsgroup being a broke student. It was always impressive to me to see how quickly someone would post a binary before the crack was posted. I remember one app I was very interested in posting on Friday, and the crack available when I checked back on Monday. Even apps that needed a physical dongle for the ADB port were cracked and not just the apps that needed a key typed in to unlock. Later, the apps that would clone a USB dongle came along, and one software company that used USB dongles actually advised using this tool when we started running their software on Blade servers that did not support adding 24 USB devices.
I did reverse engineering as a hobby in my teens. The method for 90% of cracks back in the day boiled down to:
1. Launch the app.
2. Trigger the "not registered" screen.
3. Search for the string in the binary.
4. Search for code that references the location of the string and squint hard until you spot the "check if CD-key is valid" section.
5. Patch the final opcode from "jnz" to "jz". If you felt fancy, reverse engineer the checking algorithm to create a keygen instead of a patch.
Anticracking methods became mainstream in the late 90s, but they weren't very sophisticated and most amounted to obfuscation.
Long live +ORC, +Fravia and CracksLatinos.
That hits the feels. Sometimes I miss late night sessions on SoftICE hah
This was me as well. +HCU and searchlores.org were so much fun. The coolest thing I remember was the hash-maze but I couldn't find it again. I saw Fravia live in 1999 :D
Switching 0x74 to 0x75 and vise versa lead to me ordering the free Intel books and diving into assembly and osdev for a while. There was something zen-like wading through disassembled code with a complete beginner mind and looking for patterns. This and reading Phrack around the "Smashing the Stack for Fun and Profit" times are probably my favorite memories related to computing because I only had a very vague understanding of what was going on mixed with a sense of "wait you can do this" and a lack of readily available information.
I did this to an industrial piece of software that required a hardware dongle to function, that my father and his colleagues were using.
For a few hours, I was a teenage god among men.
Efnet user spotted :)
And when the crack came out on the same day as the CD release it was called a zero-day. A term that has long since lost ita original meaning, in fact it has been inverted.
I remember old games shipping with huge lists of codes on paper. Paper made, on purpose, too big for an US legal/A4 photocopier. And moreover the codes would be using "light blue characters on slightly less light blue paper". Very annoying for casual piracy (as in: copying your neighbor's version of the game).
> Later, the apps that would clone a USB dongle came along, and one software company that used USB dongles actually advised using this tool when we started running their software on Blade servers that did not support adding 24 USB devices.
That's a good one!
I have a whole collection of boxes featuring anti-piracy stuff: https://fabiensanglard.net/happy/index.html
The sentence "This created an only-too-familiar situation where illegitimate users enjoyed a better experience than paying customers." is a direct reference to these system which annoyed people who actually paid for their game!
Only to be repeated with the experience of streaming vs torrents
Siemens for some software uses a license system that uses "secret" files stored on "dongle" that is just USB stick, and software that manages copying to hard drive etc.
I have heard of several cases where Siemens technicians and support staff would use a well-known "crack" (actually more of a keygen combined with license file manager) or guide customers how to obtain and use it...
Remember, legally (usually), the really important part is the proof of contract and payment, not exact way the software was activated
> The packaging was actually pretty high quality for a shareware product. On my copy, a sticker on the front clarifies this is the "Shareware version" with instructions to call 1-800-669-9342 (or 1-800-ID-GAMES) to unlock the full game.
> The phone number is still active today. However you don't reach the unlock center since CompUSA went out of business. The Superstore chain closed between 2007 and 2008. Instead of an unlock operator we get an automated message to sell us elderly stuff.
I remember how stuff like that presented an insurmountable obstacle as a kid. Buying stuff in the store, sure, but calling phone numbers or writing letters to companies? No way.
I like how it has come full circle today, with the phone number being inaccessible once again...
> By June 1996, after three years of hard work, id Software had completed their next title, Quake. As for their previous title, they were going to release both a shareware version and a full version of their game. Since it used a mere 22 MiB of storage, people at id Software had the idea of leveraging the remaining capacity of a CD-ROM. Why not include encrypted versions of the full id catalogue of games? Not only this would cut out the middlemen, it would give instant access to gamers with a simple phone call and a credit card.
That wasn't a new idea at the time. In 1994 Apple distributed a "software store" CD-ROM called Software Dispatch that implemented the same idea: https://68kmla.org/bb/threads/apple-software-dispatch-from-1.... The thing was full of demos, and if you wanted to buy the full version, you just had to call some number to get an unlock code.
I think I had at least one other CD-ROM that implemented the same concept. I'm sure there were tons of others.
As soon as I saw "Quake Shareware" in the title of this post, I thought "Pray to the one you will pay!" which was what appeared on the screen when you ran QCRACK.
I always wondered why Final Doom didn't unlock with QCRACK - I always assumed it was a bug with the keygen, never realized it wasn't possible to purchase it from the disk at all!
You can actually unlock Final DOOM from qcrack. But you need to use `-g` to give the codename of the game[1] because sku.17 won't give it away.
1. https://rmolina-co.blogspot.com/2016/08/qcrack-y-la-belleza-...
Surprised that it's not mentioned that the CD was also an audio CD with the Quake soundtrack..! https://en.wikipedia.org/wiki/Quake_(soundtrack)
"Originally pressed on the same disc as the game, Quake was otherwise commercially unavailable until its release on vinyl in September 2020"
Soundtrack: https://www.youtube.com/watch?v=zbLttdCOgCM
From the title I was expecting this to be about them packing 700MB of data on the CD, a common thing at the time by spacing the tracks slightly closer together, as some CD-Rs did.
How would the challenge/serial have worked if implemented correctly?
Since the challenge rotate every 5 minutes, there must be some way for a user to enter against an old challenge, right? Else they call, get a code and nothing.
But even with that out of the way, the decryption key must be the same across all discs, since there is only 1 master disc.
So the challenge/serial must combine to create a decryption key, which can be shared across users.
The best part of this article was seeing the old computer store. It was all so much better before the internet...
Don't get me wrong. I like the internet, but...
I remember Microsoft had a time limited version of Office on cd-rom that was sold for a couple bucks at Kinko's. You could replace a .dll in it and that removed the time limit.
There was something similar (but I can't recall what) where you just set your clock far in the future (like 2010!) and then the trial would last that long + one month.
Man, those CompUSA and Computer City pictures take me back. Spent too much time there (when the 7300 was my regular Mac, the CompUSA "Apple store" was my primary source of software). Probably still have some of those free Computer City floppies in a drawer somewhere.
It was fun to just browse. As least as a kid it was
Ha, I remember buying the shareware disk and unlocking every game on there. Plus you got the soundtrack which was pretty cool.
Did it work as an audio CD, the way PS1 games do?
Yes, the game just played it as a regular CD. In fact, if you put a different disc in the drive it would play the tracks from there instead. There were also console commands to change the track, stop the music, loop, or even reload the disc information if you needed to refresh it after swapping CDs.
Yeah, and I remember if you didn't have the audio cable that went from the CDROM into your sound card, you would never hear the tracks. It wasn't your computer decoding the music, the CDROM drive was simply doing the decoding, and passing the audio out into a CDROM input (just a line-level audio signal) on your sound card.
Audio tracks and WAV/AU files weren't actually decoded, but straightly send to the sound card as bits. That's it, a command like
Would just work under Unix (/dev/audio under plan9/9front)Thus, making a function for Quake to read the CD bits to the sound card would create almost no CPU usage, even less with DMA.
I'd forgotten about those little cables. If I recall correctly, they were small round cables, with maybe four conductors and a little flat connector with a clip?
4-conductor audio, generally.
Later on there was a generation of two-conductor cables that used SPDIF/PCM Audio instead of analog audio for better fidelity.
I vaguely remember CD drives with a headphone jack output!
(E.g. this one: https://pavel.network/using-cd-drive-as-poor-man-audio-playe...)
Was there anyway to make this work unless every CD was burned with a different key used to do the encryption?
No. That cracking group would have known a priori the phone call was a pretense, because id was very unlikely to press distinct CDs. So they could focus on conventional cracking techniques.
Maybe have n versions of the CD (maybe 16, or 32), so there are n decryption keys. When the buyer recites the code over the phone, the program run by the salesperson can identify which key is used based on the number.
The hacker groups would be a little frustrated trying to find all of the different CDs.
The setup costs for a pressed CD are substantial - manufacturing many different designs would have been much more expensive, and would have only slowed down crackers marginally. It's unlikely that this would have recovered enough sales to be worth it.
A reminder that CD-ROMs are pressed from glass masters, not burned like CD-Rs. I wonder if anyone has tried something like this with burned discs, though.
The multisession standard actually afforded the possibility of CD-PROM; a hybrid disc with a pressed read-only session followed by a writable section. It was only used for Kodak Picture CD (not the same as Photo CD) to hold software in the pressed session. It also apparently standardized a magneto-optical hybrid that never made it to market.
It was also used by Ricoh EncryptEase - CD-Rs with encrypted backup software in an initial pressed track. https://www.youtube.com/watch?v=_5ucImqdKbY
Hmm, it could work then if they’ve pressed most of the game data first and then recorded the encrypted part for each disk. But I guess the tooling for such setup would still be too expensive.
Yes, my point, if the answer is that there is no way this works without every disc being different.
It was not possible to manufacture at the time at scale.
It's still not possible. You would need to make a unique master for every individual unit.
You'd use CD-R of course.
With CD-Rs it was already possible in the mid '90s, if we're talking mass production. CD-Rs are a little more expensive to produce, but not much more so. A slight concern is that readers of the time might not have been able to read the discs, because CD-Rs are not quite as reflective as CD-ROMs.
Some sort of public key crypto scheme + including hard-to-change local device characteristics in the challenge.
I guess the path of least resistance then would have been to patch the executable to remove the check.
Right.
You would need some sort of hardware serial number system for PCs and a whole security infrastructure to verify that the serial number being used for the unlock wasn't faked. Then you could do per-PC unlocks. I'm not aware of anything like that being widespread back in the Quake days, though these days every computing device you own probably has a unique ID burned into ROM somewhere.
Supposedly Xbox One+ discs are individually serialized which was originally going to be used for their DRM system at launch that was walked back.
"Never attribute to malice what you can attribute to stupidity. And never attribute to stupidity what you can attribute to time pressure. -- Fab's Razor".
Surfacing this gem of a razor here, as it's buried deep in the (fantastic) blog post. Indeed, all too often, what we attribute simply to stupidity/ignorance, as per the classic Hanlon's razor, is actually downstream of market/client/management/deadline etc. pressure.
Am I reading this right, and was only the file header encrypted? Also, wasn't ID the company that did not use serials or other DRM? Maybe that explains their naivite, maybe they underestimated how easily games get cracked.
Not sure about the early days - I got a Quake II CD in the early 2000s that didn't need a CD key. Doom 3 and Quake 4 required CDs, at least at release, but after a little while no CD patches were released.
Man, if I were id, I'd sue the hell out of Testdrive for selling me that broken security system...
But it sounds like it wasn't Testdrive's fault; iD implemented it incorrectly. Their phone support weren't issuing the decrypt code, just a simple checksum:
> Described as is, there is no flaw in this process. The secret seed comes from the unlock server, tied to a CHALLENGE/SERIAL that could not be reused. But the hacker team GNOMON found a way.
> The QUAKE unlock program FLOW.EXE that ships on the CD is capable of generating the SERIAL from the CHALLENGE on its own. All it does is check that its own locally-generated SERIAL and the SERIAL entered by the user match! The entire protection mechanism relies on security by obscurity.
If the phone support issued the decrypt code, you could just replay that same code for every CD stamped from the same master.
If phone support issued an encrypted decrypt code that could only be used with your challenge code to decrypt the decrypt code, replay wouldn't be as trivial.
Every CD is identical, so no matter how iD went about this there's only ever one decryption key (per title, I assume) and those keys must either have been encoded in iD support's response, or already stored on the CD. TestDrive sold iD on the notion that the process was too hard for hacker groups to reverse engineer, and it wasn't.
There was some replay protection as described in the article, but in the end the system was flawed enough that it didn't matter.
> A gamer could go to any of the hundreds of CompUSA/Computer City stores and buy the CD for $9.95.
I'm pretty sure I bought it for $6.66
Those photos of the inside of CompUSA are just stunning. Bright sodium lighting, not a dim dank cave (one of the benefits of CRTs over modern OLED screens and their color range from black to slightly-less-black). Colorful packaging everywhere (this was before fun was made illegal in 2012). Big signage, so you know where things are (the idea of a customer knowing where to find something without being diverted to an antivirus upsell by a "helpful" "associate" didn't bother MBA CEOs back then. Nowadays the idea would cause them them to spontaneously combust).
That box art is so cool. They don't make em like they used to!
I was wondering how it was economical to burn every CD with a different key. Turns out it was not!
> Upon contacting the operator, users were to also communicate a "SOURCE CODE". It played no part in generating the Unlock code. It may have been a way for the distributors to claim a transaction fee. Browsing eBay, I found many with names indicative of past/present retailers. 12-BSTBY BestBuy, 24-CCITY Computer City, 22-CUSA CompUSA, 88, 11-1111, 38-EB Electronic Boutique, 44-FTRSP Future Shop (Canada), 34-EGGH Egghead Software, and 56-MCTR Media Play/ Musicland.
Why do we think that CCITY is a code for Computer City and not Circuit City?
Presumably MCTR was MicroCenter as well.
The "incredibly expensive Intel Pentium" image link 404s.
Thank you for reporting the issue. Fixed.
$55 ($9.95 CD + $45 activation) in August 1996 = $116 today, per the BLS inflation adjustment calculator.
Gold was ~$400 back in 1996. Now it's ~11 times that.
An expensive game however you gib it.
There was a lot less competition in the digital entertainment industry back then. Nowadays you still have AAA games going for $80 in a much more crowded market.
This screw up ended up costing id Software millions in sales. With the Doom franchise they only included the first episode in the shareware version. But to save a buck I guess they decided this was better than mailing a second CD with the full game on it. Then doubling down on that greed by putting their entire game library on the disc. This is right about the time John Romero and John Carmack parted ways, can't help but wonder if this debacle played a role. And by all accounts Quake's development was painful, so imagine this happens after going through hell getting a game out the door.
If you were to do this today, how could you make it foolproof?
COD:BO7's single-player uses dedicated servers. Terrible, yes, but that's as foolproof as it gets when it comes to piracy. They also probably did this to combat all the 'services' offering to unlock camos/cosmetics that you'd normally have to grind for; players would give someone their credentials (probably stolen later and sold to cheaters) and the person on the other side would then use a hacked client to trick the backend into thinking they completed all the challenges. With a server-side campaign, that probably got much more time consuming, I reckon.
Every game separately encrypted then compressed into solid archives. Pressed in a way every game lands on different precise height of CD, padded before compression if necessary so it takes enough space on CD. Separate machine burning tiny holes in CD from the back, but not so tiny that CIRC could fixes them. At least one hole per file.
Install locates the damage and gives you code encoding locations. Over the phone You receive ECC code. No more magic keygens, pirates have to distribute whole multi megabyte games.
I respect the creativity, but this was a 22mb game. Nothing arj or rar wouldn’t handle. Or going on irc
there isn't a market for gamers buying games on CDs these days, so it's not really a problem that needs to be solved
The same problem exists with PC games released on DVDs.
there isn't a market for gamers buying games on DVDs these days. BlueRay the same.
Of course there are voice that do want real disks but sadly that time is over. Digital sales won.
There is, at least for collector's editions. I'll continue to support physical for as long as possible.
Some collectors editions do come without a game cd nowadays.
Well how does Steam do DRM?
I think most (all?) Steam games link steam_api.dll (or something like that) which somehow figures out the running Steam process and thus the Steam ID, and then does a HTTP API request to check that you own the title.
Like 15 years ago Steam launched CEG (Custom Executable Generation) - I remember reading a write-up a while ago, but each user had a customer launcher generated on install I think, and I think there was an extra segment put into the .exe file and that was put to use somehow, though I don't know if CEG is still a thing.
The ownership check you are referring to is a local check. There is another check (involving EncryptedAppTickets) which is not local, but far fewer games use that.
CEG is long dead/was broken immediately in a pragmatic, automatable way.
Its easier on a system that doesn't need to do broadcast distribution and with servers available.
Is it really fair to call these such algorithms "security through obscurity"?
By that logic, wouldn't real symmetric crypto also be considered the same, since it's still just math that similarly swaps the data with other bytes, based on an "obscured" password or key?
The obscured part was pretending the unlocker needs external information
I guess they could have done a public/private key pair in 1996, but doing that over a phone call was going to be painful. A simple password read over the phone and verified by the software was the best solution at the time.
Mind, this was also intended for a world where internet connectivity was not to be expected. The majority of PC owners were not online.
How would your scheme work? You use RSA to encrypt the entire thing? Or you use it to get a key for a block cipher and then use that?
I feel like either way, you'd need the key to be different for every CD, otherwise you could just share the shared key. But if the encryption of every CD is different, why not just share the block cipher key directly? They can have a list of CDs and the associated key(s).
What additional security or functionality does an RSA like step add here?
Which is true, so the distribution of the software cracker would have been fairly limited I would have thought.
Sneakernet was common also. Say you have a class of 30 kids, and one of them has access to Usenet to grab qcrack.zip. He'll put it on a floppy disk and share ;)
yeah, software travelled exactly the same way as rumours that such and such had a rib removed - remarkably effectively.
Also the same way we got viruses, long before anyone I knew had the Internet.
Haha, I remember that rumour being told to me by classmates back in the day. Great example.
Hells yeah qcrack.zip
BBS, zmodem
I was 12 years old and had qcrack on a floppy from a buddy. Everyone had it. Didn't even have to have internet.
I think it is fair, as all the released CDs are identical. So all the symmetric keys used to encrypt the data are also necessarily identical.
I think you right, however they left all the right clues to decrypt it in days. You really don't want to do that.
If I ask you to pass on a message from me to someone else, and the message is encrypted using a symmetric key known to me, and to the recipient, but not to you, there's no security-by-obscurity in place.
If I ask you to pass on a message to someone, and the message is encrypted using a symmetric key known only to me, and I expect the recipient to be able to decrypt it because he knows that the key is written on the back of the paper, that's security by obscurity. You're just as able to do the decryption as the intended recipient is; I'm just hoping that you won't realize that.
[comment withdrawn]
That was an image I found in Google Image search. I should have done a better job to research the provenance and I appreciate being called on it. I have updated the photo.
Kind of sounds like the low‑background steel [1] problem all over again.
I have a rather badly written tool [2] I made about 5–6 years ago. If you connect it to your Google Custom Search Engine (along with the appropriate CSE keys), you can quickly search, copy, and download images from before 2022. It also has a ban list so you can filter out Shutterstock, Alamy, and other heavily watermarked sources.
The app is a hot circle of garbage code, but it might still be useful as a starting point.
[1] https://en.wikipedia.org/wiki/Low-background_steel
[2] https://github.com/scpedicini/truman-show
The computers in the photo and the layout of the store and signage are also quite strange.
I fed the image into Gemini and it could not detect that it’s legit or AI, but in addition to the wrong store name, the other banner that’s supposed to day SOFTWARE says TWARE or something like that. It sure seems like AI to me.
Okay, I haunted CompUSA in my teen years and I purchased my first 386 system there. So the photograph included in this blog seems bona fide.
Except their logo or name was never "Compu USA". Where in the world did they get a store hung with signs that all say "Compu USA"? Is this some Mandela Effect that has me in its clutches? The business was definitely never named or branded that way. I did my Wikipedia and Bing Images research. There are no serious search hits. Gemini is giving some truly b.s. explanations that I do not believe.
Is this a concrete example of a USA store tolerating mis-branded signage... inside their own store? Or is it some third-world knockoff? The merch on the shelves, the bros shopping for themselves, the tie-wearing salesmen, they all seem totally legit. What kind of big-box computer store lets that happen to their signage?
I made a mistake when I researched photos for it and found "one" without sources. You are right this photo looks shady. I have replaced it (with source).
It’s not Mandela effect, it’s AI.