Jump to content
Search In
  • More options...
Find results that contain...
Find results in...


  • Content Count

  • Joined

  • Last visited


About kuhnertdm

  • Title

Recent Profile Visitors

The recent visitors block is disabled and is not being shown to other users.

  1. Original Source: https://www.wsj.com/articles/verizon-looks-to-unload-tumblr-blogging-site-11556823135 Non-Paywalled Secondary Source: https://techcrunch.com/2019/05/02/verizon-reportedly-seeking-to-sell-tumblr/ Story: In 2013, Yahoo bought the Tumblr platform for over $1 billion. In 2017, this passed through to Verizon (and shortly afterwards to the "Verizon Media Group" brand) when Verizon bought Yahoo. Now, the Wall Street Journal is reporting from unnamed sources that Verizon is seeking to sell Tumblr off. The platform was originally marketed as a "blogging" application when it was created in 2007, though it didn't really stay in that niche for long, evolving into its own format based very heavily around reposting media and adding a comment, creating comment chains through sharing. Verizon had struggled for a while to properly monetize the platform. Last year, banned all adult content (a significant portion of the media on the platform) for a variety of potential reasons, including increasing the value of their ad spots (the most likely, similarly to Youtube's "adpocalypse"), trying to rebrand the platform as more family-friendly, and trying to deal with concerns of the iOS app store threatening to remove the app over issues of child porn making its way to the app. This significantly cut traffic on the platform. Those who only used Tumblr for the adult content suddenly had no reason to use it, and those who used it for other purposes suddenly saw a lot of non-adult content getting caught by the detection algorithms anyway. It seems that Verizon is now giving up the attempt to make the platform profitable, and is instead just trying to cut their losses. Opinion: I had originally thought that the porn ban had accomplished its purpose, even if the traffic got cut. It is possible for a proposition like this to be worth it if the increased ad spot value ends up creating a net increase in revenue despite fewer numbers of users (i.e. fewer users seeing much more valuable ads can still make more money). It also very clearly accomplished the other goals listed above. However, it looks like the traffic loss was vastly underestimated in the tradeoff.
  2. Source: https://techcrunch.com/2019/05/01/tesla-sued-in-wrongful-death-lawsuit-that-alleges-autopilot-caused-crash/ Story: Apple engineer Walter Huang was killed after his Tesla Model X accelerated into and crashed into a highway median, and now his family is suing Tesla for a variety of allegations including "product liability, defective product design, failure to warn, breach of warranty, intentional and negligent misrepresentation and false advertising". They are also naming the California DoT in the suit, as they failed to replace a crash attenuator guard on the median after a previous crash there, which contributed to the damage on this crash. After news of the crash surfaced back in March of 2018, Tesla was quick to release a public statement blaming Huang for the crash. They were a party to the National Transportation Safety Board investigation of the crash, and released a report of the incident without approval from the NTSB. This led to the NTSB voicing public disapproval of Tesla's actions, resulting in Musk attacking the safety board on Twitter. Subsequently, Tesla stated publicly that they withdrew from being a party in the investigation, but the NTSB later released a statement that this was not true, and they were actually removed from the investigation by the NTSB following their premature release of the report, in an effort to blame Huang. The report released by Tesla indicates a few things about the crash. First, the incident occurred in literal broad daylight (used by Tesla as evidence that Huang was not paying attention, but could also be used as evidence that even in ideal conditions, autopilot is dangerous). Secondly, Tesla brought attention to the fact that the vehicle displayed multiple warnings for Huang having his hands off of the wheel for too long. The report shows that every one of these warnings happened more than 15 minutes before the crash. It also shows that Huang's hands were not detected as "on the wheel" for 34 of the 60 seconds before impact, and that there was no manual action taken to avoid the crash. Opinion: It seems that the suit (if not settled, which I hope it won't be) is very quickly going to shift its focus to Tesla's advertising of autopilot functionality in its cars. Tesla's defense here is almost guaranteed to be "If you don't have perfect attention to the road and have your hands on the wheel at all times (Note: Defeating the purpose of autopilot), then we're not responsible if autopilot kills you". Hopefully this can be the case to shut that argument down, as Tesla's advertising thoroughly touts this as something that handles steering/braking/acceleration/etc for you, even down to the literal name of "autopilot" for a system that cannot be trusted to pilot you automatically, without killing you. I think the family's lawyer had the key line here: "Mrs. Huang lost her husband, and two children lost their father because Tesla is beta testing its Autopilot software on live drivers."
  3. If for some reason they actually do follow through, then great, that's good for the internet as a whole. However, I have no expectation whatsoever that that will happen. That's literally just Facebook saying "We are resolving to make business decisions that will dramatically cut our revenue. Trust us." That said, it could potentially be possible if they are in danger of massive revenue-based fines* (similar to what GDPR does), or even criminal action being taken (less likely), that they would basically be forced to do this, and are making it a big spectacle to look good in the media for once. *Revenue-based fines specifically, as any flat fine would be completely ineffective at curbing the behavior.
  4. Not exactly. Securities laws themselves have been pretty set in stone for a long time, and Tesla pretty much close the potential for argument of "but it's his personal account" when they routinely tell investors that Musk's account is an official source of Tesla business statements, along with the actual Tesla account. That's why that argument wasn't brought up at the hearing at all, and instead they decided to attack "what constitutes material information".
  5. Source for background info: https://www.cnbc.com/2018/09/17/telsa-ceo-elon-musk-sued-by-thai-rescue-diver-vernon-unsworth.html Source for motion to dismiss: https://www.theguardian.com/technology/2018/dec/27/elon-musk-pedo-diver-vernon-unsworth-defamation-lawsuit-dismiss Source (From one of Unsworth's lawyers): Background: Back when one of the biggest news stories around was the story of a youth soccer team being trapped in an underwater cave in Thailand, Elon Musk publicly decided that he was going to attempt to have SpaceX engineers design a mini-submarine to aid in the rescue efforts. They built the apparatus (which ended up being a simple, small cylindrical container that may or may not have actually been able to fit into the tunnels the kids needed to go through - this is still under debate), and when they showed up to assist, the actual rescue team declined their assistance. When commenting on this on Twitter, Elon Musk called one of the divers, Vernon Unsworth, a "pedo guy". Musk later deleted all the tweets and tweeted a public apology to Unsworth for his statements. Then, in response to someone on Twitter asking him about the situation, he changed his mind and doubled down by saying that it was "strange" that Unsworth had not sued him yet (Note: By that time, Musk had actually already received initial correspondence from Unsworth's lawyers). He then later tripled down in an online correspondence with Buzzfeed, in which he called Unsworth a "child rapist". He claimed in that same email that that remark was "off the record" (Note: This is not how "off the record" statements work - Both sides must agree before the remark is made that whatever the person is about to say is off the record). In the same email, Musk claimed without evidence or reasoning that the reason Unsworth was in Thailand was to find a "child bride" (Note: Unsworth has been married to 40-year-old Woranan Ratrawiphukkun for seven years now, who has publicly spoken out against Musk for this remark). By this point, Unsworth's lawyers had previously contacted Musk to request an agreement outside of court, but now that Musk has seemed adamant on not caring and continuing to double down, they have chosen to sue. Story: When the suit was first filed, Musk immediately attempted to have the case dismissed, stating that because the remarks were made on Twitter, one should reasonably expect that what they are reading is an opinion, and not a fact. His lawyers also claimed that his remarks are protected by the first amendment (Note: The first amendment does not apply to cases of defamation). In a public comment on this, Unsworth's lawyer L Lin Wood stated that this reasoning is "novel but inaccurate", as "well-established law" doesn't work like that at all. On Friday, Wood announced that Musk's motion to dismiss the case was denied, which means that now, Musk must go to court to defend himself from the defamation charges. Opinion: Very good that the judge isn't having any of this. And honestly, once Musk actually does go to court, I really hope that Wood doesn't allow him to settle. This is the most textbook case of defamation that I can think of, and really, the only thing that Musk had going for him (the statements were made online so they shouldn't count, somehow) has now been taken away from him. If there is no settlement, then it's almost guaranteed that Unsworth will win the case, and Musk could finally receive some actual punishment for the way he's conducted himself publicly for years.
  6. Source: https://www.cnn.com/2019/04/26/tech/elon-musk-sec-settlement/index.html Source for full list of material information: https://techcrunch.com/2019/04/26/elon-musk-sec-agree-to-guidelines-on-twitter-use/ Background: Elon Musk was brought to court by the SEC last year for securities fraud after he tweeted an untrue statement that Tesla was willing and able to go private at $420 per share, which caused stock prices for Tesla to jump instantly. They settled with an agreement that: Musk and Tesla would pay $20 million each in fines which would go to investors harmed by the fraudulent statement Musk would step down as chairman of the board at Tesla for at least 3 years, but could remain CEO Tesla would add two new independent members to its board of directors Tesla would implement a system where any Tweet that Musk makes that could reasonably be considered "material" to investors must be reviewed and approved by a securities lawyer The last point of the agreement was supposedly broken earlier this year when Musk tweeted an incorrect projection of Tesla's production in 2019 (The tweeted number was 25% higher than the actual agreed-upon projection made to investors by Tesla). Tesla admitted upon inquiry by the SEC that their securities lawyer was not asked to review the tweet, but later saw it and instructed Musk to tweet a correction. The SEC then asked federal judge Alison Nathan to hold Musk and Tesla in contempt of court for knowingly breaking their settlement agreement. Story: The contempt hearing itself has now been settled. During the hearing, Musk and Tesla's argument was that the original settlement did not define what was "material to investors", and as such, Musk himself should be the person who decides what is material for legal purposes (effectively meaning that nothing would ever have to go through the lawyer, no matter what it said). Judge Nathan enforced that Musk and the SEC should amend the agreement so that in the future, Musk would not be able to use this loophole to tweet whatever information he wants (even if it could reasonably be considered material). The agreement was amended to explicitly define "material information" as: any information about the company’s financial condition or guidance, potential or proposed mergers, acquisitions or joint ventures, production numbers or sales or delivery number (actual, forecasted, or projected), new or proposed business lines that are unrelated to then-existing business lines (presently includes vehicles, transportation, and sustainable energy products); projection, forecast, or estimate numbers regarding Tesla’s business that have not been previously published in official company guidance events regarding the company’s securities (including Musk’s acquisition or disposition of shares) nonpublic legal or regulatory findings or decisions; any event requiring the filing of a Form 8-K such as a change in control or a change in the company’s directors; any principal executive officer, president, principal financial officer, principal accounting officer, principal operating officer, or any person performing similar functions No further action will be taken towards Musk or Tesla as a result of this contempt hearing, but it's safe to assume that now that this loophole is closed, the next time Musk tweets incorrect info to boost the stock prices, there will be actual action taken. Opinion: I'm personally really disappointed in this decision, especially considering that: Musk's lawyer explicitly admitted during the hearing that, under their interpretation, there was literally nothing that Musk could do to break that part of the agreement The information that was tweeted out actually falls into the agreed-upon list of material information as the inflated projection was not published in official company guidance That said, it's good that now there's essentially no room for Musk to escape if he tries to commit fraud in this way again.
  7. TLDW: Apple "New Airpods" (Referred to here as "Airpods 2" - Apple is taking a cue from Nintendo here I guess) Buds themselves look/feel identical to the old ones Released in conjunction with new wireless charging-capable charging case (which still has a lightning port for if you don't have a wireless charging pad) that also looks a bit nicer than the old one Very slight improvements Battery time during calls increased from 2h to 3h Uses Bluetooth 5 instead of 4.2 Able to activate Siri by voice now instead of just double-tapping 1-2s faster device switching 30% improvement in audio latency (though this was not noticeable to the reviewer) Apple did nothing to address concerns of different ear shapes leading to buds falling out or bad audio quality for many people (no replaceable/adjustable tips - one shape, and that's it) No new types of gestures - It only allows for double tap, though the double tap gesture can be customized per ear to a variety of different functions. Telling Siri is still the only way to change the volume; you can't do it with gestures. Very big problems on Android, including volume randomly switching to max upon switching to a new song No manual way to update the firmware (Official instructions are "Plug them into an Apple device and wait", which did not work for them after hours of letting them sit) Recommended if you know that they will fit your ears and are using an Apple phone. Not recommended for Android. Comments that Apple is seriously going to need to put more effort into these if they don't want third parties to take over.
  8. Sounds like it's your email account that's insecure then. Change your password there, and maybe consider setting up 2FA on one or the other
  9. Update was pushed yesterday, so if you've updated since yesterday you should have the fix.
  10. To clarify on the previous response, this is the typical process: To register a new user Generate a random salt (This doesn't need to be crypto-secure randomness, just whatever's easiest to generate a random number.) Concatenate the salt to the end of the user's password and hash that concatenated value (Do NOT implement this yourself, use the crypto library you mentioned, and if you're able to specify an algo, something like bcrypt is best. Do NOT use MD5 or another algo used for checksums and not for encryption.) Store the following values in a file for lookup later: User ID/Username/some identifier Plaintext salt (This is why we didn't need it to be crypto-secure random) Hashed value you just made To auth a returning user Look up the values you saved upon registration Concatenate the saved salt to the end of the supplied password, and hash that Compare the hash value to the saved hash value. If they match, user is auth'd
  11. Source: https://techcrunch.com/2019/04/16/ea-origin-bug-exposed-hackers/ Story: EA has released an update for their "Origin" game launcher on Windows, fixing a vulnerability that could allow for arbitrary code execution at the privilege level of the currently logged-in user. The exploit must be triggered by following a maliciously constructed origin:// link from within a web browser. That type of link is typically used to launch Origin on the user's computer and perform an action automatically (for example, to navigate to a game on the store, or to launch a game from the app). However, researchers found that these links could also trigger Origin to run arbitrary code not related to the Origin service at all. The proof of concept launched the calculator app, which is a standard for proving that ACE can be done with an exploit. However, the potential negative effects of this kind of vulnerability basically extend to taking full control over the user's computer, as it could run arbitrary commands through Powershell, and if the user is logged in as a local admin (which the vast majority of gamers are on their own machines), those commands can be run as admin. To emphasize, the exploit can only be triggered by following a maliciously-crafted link, but that doesn't necessarily mean it's always user-triggered. If a site is already compromised with malware that can automatically redirect a web page, then this can all be done without the user knowing. Opinion: If you have Origin installed (even if you don't use it), you should update, as this exploit can happen even while Origin is not running, and without the user performing any action to trigger it.
  12. Every article about this event doesn't mention whether it was sandboxed, because the correct assumption is that it is. Saying that the articles indicate a problem is basically just assuming that they fucked up, because they didn't explicitly state that they didn't. If you read a local news article about someone getting arrested, you're not going to just assume that the officer randomly shot a cat because the article didn't say "no cats were randomly shot, per protocol".
  13. Very bad and misleading headline. It was done specifically for analysis purposes, and though it doesn't explicitly state as such, we can assume that it was done on a sandboxed machine. That's a basic procedure, and without plugging it into any machine ever, you can't get the full picture for forensics. It's really sad how every tech news site right now is chomping at the bit to say "LOOK THEY DID THIS THING THAT'S BAD AT THE VERY MOST SURFACE LEVEL!!!" when in context it's business as usual.
  14. TLDW: Gigabyte Aero 15 x9 "AI Laptop" (Gaming Laptop) Specs: Intel Core i9-8950HK CPU (Overclockable) Nvidia RTX 2070 Max-Q GPU 16GB DDR4 RAM 1TB NVMe SSD 94Wh battery I/O: 3x USB 3.1 type-A 1x HDMI 2.0 1x RJ45 (Ethernet) 1x DP 1.4 USB 3.1 Type-C 1x Thunderbolt 3 USB 3.1 Type-C UHS-II Card Reader Combo Audio Jack Main selling point: "AI" optimization for games 144Hz 15" 1080p display OR 60Hz 15" 4K display (Option) $2400 for their recommended options Upsides: Good visual/structural design Lots of I/O Tons of horsepower in a really slim chassis GPU performs great for gaming, CPU is decent (see thermal throttling note below) Reasonable noise level Great display, even good enough for pro photo/video work Great trackpad (2nd favorite for the reviewer among all gaming laptops) Great battery, still beats the competition but not by as much anymore now that the competition is getting better Easily upgradable RAM/storage/battery Considerably cheaper than comparable competition with double the storage Downsides: Very bad thermal throttling, crippling the power of the CPU (i.e. literally no reason to not buy the i7 model because it will give you the same performance and be much cheaper) AI feature does not improve performance at all (against the implication of the marketing). It basically amounts to changing the optimization settings automatically for you (essentially turning your "gaming mode" on when you're gaming, and stuff like that) No G-Sync Bad webcam, bad mic, and AWFUL webcam positioning (also doesn't support Windows Hello) Various nitpicks on the keyboard, though it works well enough (gives it a "B+" because the competition is still better) Conclusion: Very strongly recommended for those looking for a good gaming laptop. Recommended except in the case where money is no object, in which case the reviewer recommends the Alienware M15 or Razer Blade 15 instead. Recommendation is made on the merits performance and convenience-wise and basically ignores the AI as a feature, as it doesn't improve performance like they say it does, and pretty much everything it provides can be done manually on comparable laptops anyway.