John Romero created a new Doom level in January, 23 years after the game’s release, and now, he’s cooked up another one, which is based off of the Command Control map in the original Doom’s first chapter.

Romero released the level on Twitter, saying it’s the “kind of classic FPS gameplay” he’s promising with his new FPS, Blackroom. In the level’s readme file, it’s called “a warm-up for [his] new classic FPS.” The level’s description read as follows:

“With the Toxin Refinery in the rearview, you make your way to Phobos Mission Control where the computers crunching the data from the Phobos Anomaly are located. You need to use them to gain access to the Phobos Lab, but remember hearing that the computers were tied into all areas of the installation and that you never knew when the environment around you would change. You need to keep your eyes alert to all movement–this place is not what it seems…”

E1M4: Command Control was originally created by Doom designer Tom Hall. Romero says, “It makes sense that if I was creating an entire E1 of just my work, this would be the level I would remake.” The readme file says It took him two weeks to complete the level in his spare time.

To play the map yourself, you’ll need to download it from Dropbox here. You then have to drag the .wad file to a .exe Doom file and warp to the level by typing IDCLEV14 while in the game. Romero suggests using ZDoom to play his new map.

Romero announced Blackroom yesterday. The new FPS will feature mechanics like rocket jumping, circle strafing, and shooting, but it will also put a large focus on narrative. You can read more about it right here.

Better late than never

When AMD showed off the new Fiji GPUs last year, they talked about four products:
R9 Fury X, (vanilla) R9 Fury, (miniature) R9 Nano, and a dual-GPU card that was eventually christened the Radeon Pro Duo. The first card showed up in June, followed by the Fury in August and the Nano in September. By all rights, the Pro Duo should have launched not long after the Nano, but that never happened, and the longer we waited, the more we wondered what was going on with the dual-GPU halo card. Today is the official launch of AMD’s Radeon Pro Duo, but it feels more than a little stillborn.

AMD informed us last week that they’re not sampling the Radeon Pro Duo to reviewers—or at least, not “traditional” reviewers. We’ll get to that in a moment, but all we have to go on right now are product images and a slide deck, the latter of which includes some benchmarks and a spec sheet. So we’re not reviewing the Pro Duo as such, but that doesn’t mean we don’t have plenty to discuss. You see, the Pro Duo is effectively CrossFire R9 Nano, on a single PCB, so armed with a couple of R9 Nano cards we’re able to get a pretty clear idea of how the Pro Duo would perform.

Dual-GPU cards have always been a niche market, and the Pro Duo takes that to a new level. The last dual-GPU card from AMD, the R9 295X2, was roughly equivalent to CrossFire R9 290X (with lowered clocks unless you overclocked), and it was priced at $1499, compared to $549 for individual 290X cards. It also launched six months after those cards, and the market was still languishing in 28nm land. Radeon Pro Duo targets that same price, $1499, but this time the R9 Nano has already dropped to $499. Perhaps more problematic is that we’re nine months after the Fury X, but we’re also about two months before the 14nm FinFET Polaris parts are rumored to arrive.

And that, in a nutshell, is why AMD isn’t sampling the card to reviewers—they can guess what we would say: “too little, too late, and too expensive.” They’re pretty much right. Other than occupying a single slot and using a single closed loop liquid cooler, it’s not any faster than a pair of R9 Nano cards. But AMD is throwing a bit of a curve ball here as they’re touting this as a professional card rather than just a gaming monster.

The marketing is all about content creation, TFLOPS, LiquidVR, and DX12; and the Pro Duo should do well in all those areas. To further muddy the waters, it appears AMD will also give owners of the Pro Duo a choice between gamer and professional drivers, so in a sense this is a workstation class card. This is AMD’s first VR Ready Creator offering as well, and AMD is partnering up with Crytek and their VR First initiative to help budding VR developers. If you’re willing to pay the price premium for a dual-GPU card and you need the specific features that AMD is offering, it might be worth the price premium.

What if you’re just a gamer?

While we don’t have an actual Radeon Pro Duo for testing, we do have a couple of R9 Nano cards. Performance should be the same, maybe with a small variance of a few percent. Considering you could buy a couple of R9 Nano cards and save $500, or even a pair of GTX 980 Ti cards while saving $300, though, the benchmarks are mostly a formality. This isn’t a card for gamers, particularly not with Polaris and Pascal right around the corner. Our test platform consists of an overclocked i7-5930K, 4x4GB DDR4-2667, two reference clocked GTX 980 Ti cards, and two R9 Nano cards.

For the games, we’re going to go whole hog here and including benchmarks from eleven different games. The games range from older titles that have had SLI and CrossFire profiles for ages to a few newer titles. We’ve tested using the latest AMD Crimson 16.4.2 Hotfix drivers and Nvidia’s 364.72 drivers, at 1920×1080 Ultra, 2560×1440 Ultra, and 3840×2160 and High or Ultra, depending on the game (The Division, GTAV, and Rise of the Tomb Raider were tested using their respective High presets). We’ve included both single GPU and dual GPU results to help shed some light on what to expect.

Starting with our 4K testing, dual Nano GPUs can come quite close to matching dual GTX 980 Ti, and even manage a few wins (in Far Cry Primal, GTAV, and Hitman: Absolution). Considering there’s a power draw advantage for AMD here (350W vs. 500W TDP), not to mention the memory disadvantage (4GB vs. 6GB per GPU), if we’re talking about AMD’s $1000 pair of cards and Nvidia’s $1200 pair of cards, both are great performers. A single 980 Ti is 15 percent faster than a single Nano, but in SLI/CF the gap at 4K is only three percent on average. This is due to overall much better scaling on AMD: CrossFire improves AMD’s performance by 85 percent, but SLI only increases the Nvidia performance by 66 percent.

Of course from a gaming perspective, paying $1500 for the same performance is a different matter. There’s also the matter of the minimum frame rates, where the 4GB VRAM takes its toll. This time, a single 980 Ti has 40 percent higher minimum frame rates on average, though in dual-GPU configurations the gap narrows to 19 percent. Interestingly, despite having less VRAM, 4K has been one of AMD’s strong positions.

Dropping to 2560×1440, things are more in favor of Nvidia overall. All of our test games are running above 60 fps on the 980 Ti SLI setup, and AMD comes close, though they fall well short in The Division—where CrossFire actually reduced performance for R9 Nano (which wasn’t the case at 4K). Considering 4K isn’t really all that common a use case, this is arguably a more important battleground. Average frame rates favor Nvidia’s GPU by around 15 percent overall, while minimum frame rates are over 30 percent higher.

1080p gaming, even at essentially maximum quality, really isn’t the target market of any of the graphics card setups here, though in general the results are pretty similar to 1440p.

As a reprise of what we saw last fall, this should all be pretty familiar territory. AMD’s R9 Nano is far more attractive now at $500 than the original $650, and it’s a sweet little card—though you do need to plan around the open air cooler if you’re using a small case, as venting heat into the chassis can be problematic. The Pro Duo does solve the issue of an open air cooler, but it’s also a much larger card and you now need a place for the radiator—not really a problem in any decent ATX case, naturally, but fitting it into a smaller can would require more planning and carefully selecting components. What the Pro Duo doesn’t do is cater to gamers in any way that we’d recommend going out and buying one—content developers, maybe, but not pure gamers.

That’s a shame, because as an engineering feat, the Radeon Pro Duo looks freaking awesome. Check out the compact cooling solution, and it’s hard not to be impressed; the HBM modules allow for a much smaller area than previous dual-GPU cards. It’s too bad AMD didn’t get this out the door earlier, when it could have made more of a splash. It’s a graphical tour de force for AMD’s Fiji, the most potent implementation we’re going to see, but sadly, it’s too late by half.

The Pro Duo isn’t bad, but unless you absolutely have to buy a new graphics card today—and you’re interested in LiquidVR, DX12, VR, and/or content creation—there’s no sense in rushing out to buy one. AMD may or may not release a high-end Polaris part in June, but Nvidia certainly has plans, if 
rumors are correct. Worst case, you can still get two R9 Nano cards—or even a pair of R9 Fury or R9 Fury X—for less money. That leaves the Pro Duo as a sweet looking card, but one that will likely have a very short shelf life. 

End-to-end encryption by default is quickly becoming the new standard for any communications app that claims to care about the privacy of those who use it. But not all encryption is created equal. And default doesn’t always mean default.

In just the month of April, both WhatsApp and Viber switched on that layer of protection from surveillance, which is designed to make it technically impossible for anyone to read the services’ messages other than the people in conversation: not eavesdropping hackers, not law enforcement, and not even the companies themselves. For the two companies’ massive userbases—more than 700 million phones run Viber, and more than a billion run WhatsApp—that switch seems to represent the start of a new era where strong encryption is no longer a privilege for geeks and paranoids, but an effortless mainstream tool.

But that effortlessness hasn’t arrived quite yet. In fact, despite the two companies touting their encryption as ready out-of-the-box, you actually need to flip a few subtle switches of your own to activate that level of security. From tweaking the apps’ settings to checking your backup configurations, there are a few quick steps you can take to make those apps’ end-to-end encryption significantly more secure. “Security is not magic, and using WhatsApp will not magically protect anyone from surveillance,” says Filippo Valsorda, a cryptographic engineer for Cloudflare who’s analyzed both WhatsApp and Viber’s crypto setups. “There are still things you need to be aware of to make sure you don’t undermine these apps’ end-to-end security…A lot of information is missing that the public needs to know.”

Tip 1: Enable Fingerprint Verification

In WhatsApp’s case, Valsorda points out there’s a nagging problem of authentication: For encryption to guarantee that only an intended recipient can decrypt a message, that recipient needs to prove they’re who they say they are. WhatsApp and other end-to-end encrypted messaging tools let communicators check each others’ “key fingerprints”—an abbreviated version of a unique key that WhatsApp stores on the phone to prove a person’s identity.

But in its default state, WhatsApp doesn’t alert a sender when the key fingerprint of a recipient has changed. A new fingerprint could merely mean that the recipient has started using a new phone or deleted and reinstalled the app. Or it could mean something more troubling: that a “man-in-the-middle”—such as a law enforcement agency with a wiretap order forcing WhatsApp’s cooperation—has inserted himself, and is intercepting and decrypting every message before passing it on to the intended recipient.

Luckily, WhatsApp offers a “security notifications” feature—not on by default—that automatically remembers all of your contacts’ fingerprints for you and alerts you if a fingerprint changes. To turn it on, flip the switch on the Security page under Accounts in the app’s settings.”Without that setting on, the fingerprint can change at any time, and the phone will say sure, I’ll use this new key,” says Valsorda.

“But if that setting is on, an attacker can’t decide at some point that you’re an interesting person and start intercepting from then on, because a warning will appear.”

Viber handles that key fingerprint verification with a different process that requires its own sort of manual verification. As the company explains in its security FAQ, a contact is only considered “verified” after you’ve called them through Viber’s voice-calling feature, both verified that you’re talking to the person you think you’re talking to, and then tapped a lock icon during the call. From then on, that person’s messages will appear in green. If that color changes to red, it’s an automatic warning that their key fingerprint has changed. But until the person is verified, their fingerprint is considered unverified and can change without any such alert.

Tip 2: Disable Cloud Backups

Beyond key fingerprints, anyone who backs up their data may face an even more glaring issue: those backups often aren’t encrypted—or at least not using an encryption system for which only you control the key. Both Whatsapp and Viber messages, for all their fancy end-to-end encryption, have that protection stripped away when they’re backed up to Apple’s iCloud servers or Google Drive. And that leaves your messages open to all the usual risks of exposure to hackers, to Apple or Google themselves, or to any government that can force those companies to turn over the data. “If you have an app that backs up to iCloud, that’s for the purpose of restoring that content to another device…End-to-end [encryption] suggests no other device can read those messages,” says iOS forensics consultant Jonathan Zdziarski. “To me, those two terms are mutually exclusive.”

On the iPhone, those backups can be easily turned off for specific apps under the Backup Options menu in settings. For iPhone owners using WhatsApp, which actually backs itself up two ways, you need to take an extra step: disable backups within the app itself under “Chat Backup” in the Chats menu in settings. Android owners can avoid the problem by not setting up Google Drive backups from WhatsApp in the first place. And the same advice holds for third-party cloud backups like Dropbox. If you want to keep your messages fully end-to-end encrypted, never sync WhatsApp or Viber with a cloud backup program.

The downside, of course, is that turning off backups means that messages can’t be recovered if your phone is destroyed or lost. With true end-to-end encryption, that’s almost considered a feature. Cryptographer Matthew Green has written about what he calls the Mud Puddle Test: If you drop your phone in a mud puddle, then slip in that puddle and crack your head, forgetting all your passwords, can you still recover your data? If you can—say, by using Apple’s password recovery feature to access an iCloud backup—the data wasn’t truly encrypted in the first place. The encrypted messaging app Signal, recommended by Edward Snowden and widely considered the most secure option for encrypted messaging, passes that test. If a phone running Signal is dropped in a mud puddle and destroyed, the messages are simply gone.

Not everyone is willing to risk losing their messages in exchange for the security of true end-to-end encryption. But that’s the tradeoff that fully secured end-to-end encryption requires. And those seeking privacy should know as much, rather than be lulled into a false sense of security by companies’ promises of of protection that don’t include caveats and edge-cases. “Chances are you’re making a compromise by backing up to the cloud,” says Zdziarski. “That has its place, and it’s useful. But the user needs to be aware that by doing that they’re probably exposing their data.”

Go Back to Top. Skip To: Start of Article.