Just remember... with Clomid, more is not necessarily better. A small dose (like 25mg every other day) will moderately boost testosterone & leave you looking & feeling GREAT. On the other hand, a BIG dose (like 50mg/day) will boost testosterone a lot, but will ALSO boost estradiol & will probably make you feel like shit. Definitely give Clomid a try first, but if 25mg every other day doesn't do enough after 3-6 months (it takes time to work), bite the bullet and get testosterone instead.
I almost bought 3 UniFi AC-pro units, until I discovered that they eliminated the ONE GODDAMN FEATURE from their AC models that made their hardware unique: Zero-Handoff.
Yeah, ZH was bad for networks with lots of users spread across a large area, but it was PERFECT for making 2.4GHz wifi usable by a small group of people in congested urban areas [by allowing you to put a 2.4ghz AP in every room with transparent hand-offs that ACTUALLY WORKED with Android devices & shout over your neighbors (~97% of Android devices STILL have fucked-up wpa-supplicants that won't switch APs unless the AP literally kicks them off.)
Wireless HDMI. And maybe some future standard for PCIe and Thunderbolt over TCP/IP (the way we now have semi-standards for transporting USB, SCSI, and FireWire over TCP/IP (like almost every networked printserver appliance now sold).
At some point, it probably WILL become cheaper & easier to transmit insanely fast data without wires. At least, if you have fiber to each room, feeding a 60GHz access point with 20 feet of cat6 doing 10-gig ethernet. At that point, the wireless network is basically short-range de-facto open-air fiber [without the actual fiber].
TSA confiscates nail clippers from pilots, too. The fact that there's a literal AXE hanging behind them in the cockpit (so they can smash the window and escape if the plane crashes and they somehow manage to survive long enough for the axe to be useful) has no effect on TSA's logic.
I'd be more worried that they'd install NSA-grade bootloader-compromising malware capable of surviving anything short of JTAG-reflashing everything from the motherboard BIOS to the hard drive, videocard, and network card firmware, and turn my kick-ass laptop into one that mysteriously crashes for no apparent reason thereafter, even after I've reinstalled Windows multiple times (without even getting into the fact that it would be permanently compromised from a privacy and security standpoint). Think: Sony rootkit on steroids, with the nearly-unlimited of the US government and support from the legal system behind it (for the few who don't know, Sony's rootkit was distributed as a file that auto-ran if you inserted certain audio CDs to play them on your computer. It literally REFLASHED YOUR DRIVE'S FIRMWARE to disable functions used by ripping software).
The question isn't whether the NSA has malware like that. They absolutely do. Google "Advanced Persistent Threat" ("APT"), and know that it's common knowledge that the US, Russia, Britain, China, and Israel (plus countless more) ALL have state espionage agencies with the resources to develop and deploy APTs... and they actively do it every single day.
The NSA is full of self-perceived super-patriots who've willingly sacrificed every shred of their own privacy, and see nothing wrong with inflicting large-scale collateral damage to American citizens' computer hardware in the holy name of protecting the American homeland from any threat... major or minor, real or perceived. To their mindset, if deploying malware to the laptops of 14 million American citizens crossing the border in some given year causes Windows (or any network hardware that might be subsequently used by those laptops) to occasionally crash for no apparent reason thereafter, but enables DHS to prevent a single terrorist attack, it's 100% worth it, and as far as they're concerned, anyone who thinks otherwise is an evil commie terrorist-loving scumbag who hates America.
This story is insane. Prototypes of things involving emerging technology are NEVER, EVER, **EVER** tiny & compact.
In the early 1980s, pre-Commodore Amiga showed off their new computer's prototype at Comdex. It was a rack the size of a small refrigerator stuffed with handmade (wire-wrapped) logic boards. Two years later, it was an attractive-looking desktop computer with nifty open space underneath that was big enough to tuck the keyboard into.
The first version of Android was developed for a device that was a "phone" only in the sense that it could be used to make and receive phone calls, but was REALLY several cubic feet of prototype boards connected with ribbon cables and LITERAL duct tape.
It would be a HUGE mistake for MagicLeap to prematurely commit to a controller design just for the sake of early miniaturization. I'd rather see them implement the controller as an 802.11ad-connected semi-dumb remote frame buffer, and offload the back-end heavy lifting to a desktop PC that's as big as it needs to be to do its job and impress everyone.
The fact is, landfills around the world are littered with the corpses of prematurely-optimized hardware that ended up being inadequate for their intended purpose. That's why first-gen routers usually have more ram, faster processors, and better chipsets than second-gen routers... the first-gen ones are slightly over-engineered to give them headroom to handle more advanced capabilities, while the second-gen ones are pruned back to the bare minimum specs capable of running the first-gen model's firmware 9-15 months after release.
Yeah, just TRY enabling QoS on a consumer-grade AP/Router. It'll KILL your throughput. Why? Because consumer-grade hardware doesn't have sufficiently-fast CPUs to actually inspect network traffic at gigabit speeds and make intelligent traffic-shaping decisions... they just implement "QoS" by arbitrarily limiting the bitrate from any one device to some fraction of what it thinks is the total link rate, exclude traffic on ports used by popular VoIP services, and call it a day.
With QoS enabled, I couldn't get more than ~8mbps to speedtest.net from any device on my LAN. With QoS disabled, I get 50mbps+ without breaking a sweat.
> They've become a smaller fish in a much larger pond.
More like a slightly-larger fish in an exponentially-larger ocean. On its least-profitable month during the Great Recession, Microsoft made more money than they did the entire year after Windows 95 came out. Most investors would be *delighted* to experience Microsoft-style failure.
Most of Microsoft's wounds have been either self-inflicted, or more like allergic reactions to imaginary problems -- dumping Windows Mobile *right* at the point when its capabilities stomped all over iOS and Android & it finally started to be "not ugly", self-mutilating Windows while consumed by "Tablet Fever", etc.
A few years ago, MSI *had* a line of modular "whitebox" laptops. So did Clevo. They've become rare, because discrete graphics are almost the only thing *left* to vary... and engineering the cooling to be adequate for high-powered GPUs without going totally overboard for low-powered GPUs is almost impossible.
The thing is, a modern word processor now has capabilities that used to require a desktop publishing app (remember PageMaker?). When you get down to it, WordPerfect 6 and earlier were basically HTML editors, but with their own proprietary tokenization scheme instead of anything SGML-derived).
The big question: can it at least do 5GHz 802.11n, or is it 2.4GHz-only? Most n-only implementations sold NOW are n-only BECAUSE they're 2.4GHz-only (by definition, 802.11ac REQUIRES 5GHz).
In single-family suburbia it might not matter (much), but in most urban residential areas, 2.4GHz wifi has become almost unusable.
I wish the FCC would buy back the upper half of the wi-fi channel centered on channel 14 via eminent domain, then allow wi-fi to use it ONLY as a 20MHz channel limited to ~5-20mW EIRP (or 1-4mW amp power, to reduce certification costs) to give urban users AT LEAST a single channel that's good enough for same-room use, but sufficiently power-limited for a single sheet of drywall to attenuate it by 20-50dB.
In 2006 there were amps that could SWITCH HDMI, but very few that could sit between the source and TV, extract the audio, AND spoof a 5.1-capable EDID. Go to avsforums.com & read the hundreds of angry forum posts from ~2008-2010 when Blu-Ray became real, and most people with "HDMI" receivers discovered that they had broken implementations that would NEVER be allowed to work as advertised.
Then, the HDMI people turned around & fucked everyone a SECOND time with HDCP 2.2. It's madness. They literally expect people to keep throwing away perfectly good amps every 2-4 years just to continue being allowed to have surround sound.
And I can assure you, prior to ~2013, almost NO amps supported 5.1 LPCM. They could do 7.1 DD+, 7.1 DTS+, and 2.0 LPCM, but NOT 5.1 LPCM.
Nintendo was too fucking cheap to license Dolby surround, so if your amp didn't have working HDMI audio input (at launch, most of the few HT receivers that supported HDMI had broken implementations) AND support 5.1 LPCM, no surround sound for you. It couldn't even fall back to fucking ProLogic. The goddamn GAMECUBE had better surround-sound support (via ProLogic) than 98% of Wii-U owners ever got to enjoy.
To wit: in 2012, most home theater amps only supported DDS and DTS via S/PDIF. Nintendo decided to go with a standard that most amps (even ones that were only a year or two old) couldn't use... then, or EVER.
Lack of DD5.1 is a major reason why I always get the xb360 version of any game even though I also have a (generally unloved) Wii-U... graphically, Wii-U is no better than xb360, and the 360 has DD 5.1 surround sound & better controllers.
The xb360 is ALSO why I'm able to enjoy DD 5.1 surround with Netflix & Amazon... it's the ONLY platform allows you to enjoy surround sound with streaming video services if your amp lacks hdmi and DD7.1+
Oh, and let's not forget that Wii-U has a Blu-Ray drive, but can't actually PLAY Blu-Ray movies because Nintendo was too cheap to license the nessary IP.
Mass-production vs one-off customization. Even with robots and CAM to handle the grinding, it still takes more time and effort to grind two lens surfaces instead of just one... and more effort to calculate custom lenses instead of blindly grinding another standardized design on the other side.
That said, I think that competition from cheap online labs will force traditional retailers to reduce the cost of freeform lenses and market them more aggressively. If you can take your prescription and buy a mediocre pair of glasses for $15 online, why would you pay $50-100 for an equally-mediocre pair? On the other hand, if a mediocre pair is $15, but a freeform pair is $200, and the freeform pair is sufficiently better to feel like a HUGE improvement, people will pay the higher cost for the quality pair (even IF they buy a cheap mediocre pair online to keep as a spare pair). By (US) law, your "prescription" (sph, cyl, axis, prism) have to be provided on demand so you can use it elsewhere, but the measurements needed to make custom lenses for a pair of frames are still considered proprietary (and are pretty damn hard to measure without the proper equipment), so they aren't required to share them with you. Ten years from now, mediocre online glasses will be regarded the way drugstore reading glasses are now... something that's better than nothing if you're poor or need a spare, but blatantly inferior to "real" glasses.
As far as SV lenses for astigmatism go, the main advantage of two-sided freeform fully-custom over single-sided freeform semi-custom is aesthetics... with control over BOTH surfaces, you can neutralize things like magnification/minification (so your eyes don't look larger or smaller to observers seeing them through the lens), and adjust the base curve near the periphery to allow thinner lenses. If you only have full freedom over one surface, you can optimize for optics at the expense of thicker lenses, or thinner lenses at the cost of optic fidelity, and your ability to neutralize out magnification/minification will be severely constrained.
From what I've been read, single-surface semi-custom freeform is a HUGE step forward from single-surface lenses ground with traditional standard curves, but the difference between single- and double-surface freeform (for SV astigmatism) is more like the difference between 480p24 from a DVD and 720p24 from Blu-Ray... it's there, but it's not nearly as dramatic as the difference between nominal-480i60 from VHS and 480p24 from DVD.
I believe lenses marketed as "aspheric" are basically designed the way freeform lenses are (with raytracing), but are mass-produced in only a few permutations of sph+cyl... and usually, optimized for thinness over optics. Personally, I wouldn't bother with them... they cost almost as much as single-surface freeform lenses, and don't provide nearly as much optical improvement.
Not to mention, the additional anchoring requirements needed to survive an earthquake. In Florida, you can get away with pilings that passively rest on top of shallow bedrock. In California, the pilings have to be anchored to the bedrock through shock absorbers. Otherwise, it could fall over like a top-heavy stack of books.
Until fairly recently, Florida highways had almost NO explicit design features to resist seismic activity. They were LITERALLY built on pilings resting passively on bedrock that were held in place by nothing more than friction & gravity at the bottom, and the bridge deck's beams at the top. It wasn't a seismic failure per se, but a freeway bridge under construction in Miami (97th Avenue over SR836) actually collapsed during Hurricane Wilma... the pilings were up, and the beams were laid on top, but the contractor didn't bother to temporarily anchor & brace them, and the hurricane hit before the deck's concrete was poured. The area was flooded (providing lubrication), the hurricane spawned a tornado, and the combination pushed a beam past its tipping point, ultimately taking down the other beams & columns around it. As a result, contractors are now required to brace the beams with temporary supports until the deck gets poured. My point is, if mere WIND (lubricated by water) could cause a structural failure, just imagine what an underground explosion at a gas station next to a freeway overpass that was strong enough to make the earth visibly ripple could do to a road built like that. In contrast, when I-595 was rebuilt in Fort Lauderdale, the new ramps had additional bracing every few columns to keep the road deck from sliding from side to side if something made the ground vibrate.
The current gold-standard of long-term correction is probably a custom-ground scleral lens. Dr. Gemoules in Dallas fits lenses using the hardware NORMALLY used for laser surgery... he scans both eyes in 3d and orders the first (few) set(s) with the eye's precise shape ground into the back. Once the back side is optimal, he does a wavefront scan while you're wearing those lenses and uses the data to raytrace the ideal front curve.
He's expensive compared to Walmart or Lenscrafters, but often cheaper than what some OTHER doctors charge for less-optimized scleral contacts.
Results vary, of course, but he's had patients walk in who were almost blind without glasses go home with 20/15 (or better) vision.
The catch? You have to physically go to Dallas & spend a week there. He's the only doctor who does it (not sure how much is unwillingness to partner with doctors elsewhe, and how much is due to FDA regulation of officially-experimental treatments... it's possible that current patients are officially & technically part of an ongoing clinical trial). But I'd say about half the things written about him have been "wildly enthusiastic", and only a few have been strongly negative (and most of THEM seem to be just unfortunate patients whose problems simply outstrip the capabilities of ANY currently-existing technology).
Google for "Laserfit AND Gemoules" (sans quotes)
Disclaimer: I'm not a patient of his... but I sure as hell WOULD be if I could afford it. If there's a better alternative available from ANYONE today, I have yet to read about it.
Fyi, if you've had laser surgery, your "glasses" prescription might be minor, but your "contact" prescription will probably be similar in strength to your pre-laser prescription. Why? GP scleral lenses optically fuse with the cornea via tears (watery, not injury), and mostly neutralize whatever lens was burned into the cornea by the laser. Other contacts are similar, but it's the most dramatic with scleral RGP lenses (soft lenses might bend slightly to conform to an irregular cornea).
Also, scleral RGP lenses LOOK scary-huge, but in practice they're more comfortable & stable because they rest on the white part (little/no nerve sensation) instead of the iris/pupil/cornea (lots of nerve sensation).
Except, it's only "peripheral vision" when you're looking straight ahead. The moment you look to one side by moving your eyes, it ceases to be "peripheral", and being in sharp focus becomes VERY important.
It's because 99.9% of progressive lenses have a semi-custom lens ground into one side, but a standardized lens ground into the other, with no regard for getting the optics right anywhere besides the center vertical axis.
It IS possible to make better glasses using raytracing (to design) and custom grinding (to implement) on both sides to "get it right" along the horizontal axis, too... but those lenses are expensive, and you'll only see a dramatic improvement if
1) the optician gets all the measurements precisely right (they rarely do, because full custom lenses are expensive, few stores sell more than one or two pairs a year, and they require more measurements than "normal" lenses that most techs don't fully understand);
2) The glasses are meticulously adjusted for proper alignment... and kept in proper alignment with frequent adjustments.
Realistically, you're looking at glasses that cost about $600-1,000/pair PLUS the frame cost.
The tech was developed for custom progressive lenses, but it can also be used to make better single-vision lenses for people with astigmatism. Normal glasses have a standardized sphere + base curve manufactured into one side, and a non-optimized cylindrical lens ground onto the other side (usually, without regard for lens angle or lens distance FROM pupil). I believe full-custom single-vision lenses run about $500-800 more than "regular" (non-optimized) ones.
The magic word to say & demand when asking about such lenses is "freeform" (often, used in conjunction with "custom" and/or "high-definition"). Just be aware that the optician's skill & experience fitting freeform lenses is ENORMOUSLY important. Even freeform custom lenses can suck if the optician is careless with his/her measurements. And demand the specific word "freeform" -- unlike other marketing terms, "freeform" has a very specific meaning in the industry. Not all lenses advertised as "hi-def" or "custom" are literally "freeform". At least one brand that's "semi-custom" exists that grinds a customized & optimized cyl onto one side of a lens manufactured with a standard sph+base curve on the other. For SV astigmatism, semi-custom might be good enough... but for progressive, you really want full-blown two-surface freeform custom lenses.
The South Bay area actually has a SHITLOAD of undeveloped land... it's just that most of it happens to be mud flats adjacent to SF Bay. The catch is, you can't "pull a Florida" and turn that swampland into prime waterfront real estate by dredging out finger canals and using the fill dirt to raise the surrounding terrain, because the first major earthquake will cause all that fill dirt to settle & pack itself down, most likely destroying any single-family home built on top of it.
You COULD, however, build skyscrapers on that land, because skyscrapers sit on pilings that go all the way down to the bedrock. The surrounding land (if fill dirt from dredging) would still settle during a major earthquake, but the skyscrapers themselves would be fine.
The catch is, skyscrapers aren't cheap to build... especially if every one of them has to be an architecturally-spectacular work of art to have any chance in HELL of getting approved by voters. And right now, there's almost zero market demand for skyscrapers in the mid-bay area, because people who could AFFORD a condo in a new bayfront skyscraper would rather live in a mountainside mansion, and people who'd WANT to live in a bayfront condo can't afford to buy one.
Raising wages will initially just create inflation, but beyond a certain point, you'll have SO MUCH MONEY getting thrown around relative to construction costs, developers will say "fuck it" and actually TRY getting skyscraper projects approved, because land costs will be so staggering relative to new construction costs. Or they'll start to do things that would otherwise be cost-prohibitive, like buying every single-family home in a neighborhood (taking the loudest NIMBYs out of the equation) so they can demolish the whole neighborhood and replace the single-family homes with 3-4-story townhomes and 5-10 story apartments & condos.
> Digital music uses 10 times the nyquist rate; better than a human ear can hear and distinguish
192,000hz sample rates are about 10x, but 44,100hz is barely more than the Nyquist minimum.
Nyquist is a MINIMUM. The only thing Nyquist GUARANTEES is that a 44.1khz sample rate is INADEQUATE for reproducing frequencies above 22.05khz. It makes no claim that it's actually good enough for flawless reproduction.
Also, 192khz sample rates are really more like 96khz if you're attempting to mix two streams in realtime due to clock jitter (unless all live streams share an accurate clock signal that compensates for cable length and the speed of light, the bytes won't precisely align).
That said, 24-bit/channel at 192khz is definitely past any point where double-blind listeners could distinguish between live (but non-class-D-amplified) and digital.
Newer technologies DO have a few quantifiable advantantages over CRTs: subpixel-addressability, and no convergence error.
NOBODY circa 2000 thought LCD would EVER achieve the dominance it has today -- almost everyone in the display industry was betting on FED as "the next big thing" (FED is basically a CRT that's directly-illuminated at the subpixel level by an array of solid-state electron emitters sitting behind them). The problem with FED was dead pixels... they were as seemingly-unavoidable with FED as they were with TFT. Ultimately, LCD manufacturers solved the problem first. The reality, though, is that the same technology that now gives us LCDs without dead/stuck pixels could give us FED, too. FED has one problem, though: it uses as much power AS a CRT, so it's not really viable for laptops, phones, etc. OLED basically gives us everything FED could, with the added bonus of (relatively) low power consumption. Even in the early 90s, FED's creators knew LED was the future... it's just that back then, blue LEDs were still a lab & prototype novelty, so they came up with the idea of using a LED-like electron emitter to excite blue phosphors.
Regardless of whether FED still has a future or we go straight to OLED, LCDs *suck* compared to both.
Mechanically, they were similar to the generation or two of turn-of-the-century ultra- ultra- ultra-high-end car CD players that minimized skipping by mechanical means (vibration-dampening, etc) instead of just letting it skip, but reading the disc 4-16x faster & buffering enough to keep the output pipeline filled during the recovery. Which, incidentally, sounded WAY better than the ones that relied on buffering since audio cds have no concept of tracks or sectors, so figuring out precisely WHERE to continue after a skip event is problematic (redbook audio has no FEC, so you could potentially read the same chunk 5 times & get 2-5 chunks that differ by at least a single byte. That's why DAE apps like Exact Audio Copy read everything multiple times, try to line up the chunks, and determine which byte is the right value, and which ones were just read errors).
Slashdot Mirror
Just remember... with Clomid, more is not necessarily better. A small dose (like 25mg every other day) will moderately boost testosterone & leave you looking & feeling GREAT. On the other hand, a BIG dose (like 50mg/day) will boost testosterone a lot, but will ALSO boost estradiol & will probably make you feel like shit. Definitely give Clomid a try first, but if 25mg every other day doesn't do enough after 3-6 months (it takes time to work), bite the bullet and get testosterone instead.
Pull-ups are a bad metric, because a healthy, muscular woman with large breasts is carrying the added-weight equivalent of two pot bellies.
I almost bought 3 UniFi AC-pro units, until I discovered that they eliminated the ONE GODDAMN FEATURE from their AC models that made their hardware unique: Zero-Handoff.
Yeah, ZH was bad for networks with lots of users spread across a large area, but it was PERFECT for making 2.4GHz wifi usable by a small group of people in congested urban areas [by allowing you to put a 2.4ghz AP in every room with transparent hand-offs that ACTUALLY WORKED with Android devices & shout over your neighbors (~97% of Android devices STILL have fucked-up wpa-supplicants that won't switch APs unless the AP literally kicks them off.)
Wireless HDMI. And maybe some future standard for PCIe and Thunderbolt over TCP/IP (the way we now have semi-standards for transporting USB, SCSI, and FireWire over TCP/IP (like almost every networked printserver appliance now sold).
At some point, it probably WILL become cheaper & easier to transmit insanely fast data without wires. At least, if you have fiber to each room, feeding a 60GHz access point with 20 feet of cat6 doing 10-gig ethernet. At that point, the wireless network is basically short-range de-facto open-air fiber [without the actual fiber].
TSA confiscates nail clippers from pilots, too. The fact that there's a literal AXE hanging behind them in the cockpit (so they can smash the window and escape if the plane crashes and they somehow manage to survive long enough for the axe to be useful) has no effect on TSA's logic.
I'd be more worried that they'd install NSA-grade bootloader-compromising malware capable of surviving anything short of JTAG-reflashing everything from the motherboard BIOS to the hard drive, videocard, and network card firmware, and turn my kick-ass laptop into one that mysteriously crashes for no apparent reason thereafter, even after I've reinstalled Windows multiple times (without even getting into the fact that it would be permanently compromised from a privacy and security standpoint). Think: Sony rootkit on steroids, with the nearly-unlimited of the US government and support from the legal system behind it (for the few who don't know, Sony's rootkit was distributed as a file that auto-ran if you inserted certain audio CDs to play them on your computer. It literally REFLASHED YOUR DRIVE'S FIRMWARE to disable functions used by ripping software).
The question isn't whether the NSA has malware like that. They absolutely do. Google "Advanced Persistent Threat" ("APT"), and know that it's common knowledge that the US, Russia, Britain, China, and Israel (plus countless more) ALL have state espionage agencies with the resources to develop and deploy APTs... and they actively do it every single day.
The NSA is full of self-perceived super-patriots who've willingly sacrificed every shred of their own privacy, and see nothing wrong with inflicting large-scale collateral damage to American citizens' computer hardware in the holy name of protecting the American homeland from any threat... major or minor, real or perceived. To their mindset, if deploying malware to the laptops of 14 million American citizens crossing the border in some given year causes Windows (or any network hardware that might be subsequently used by those laptops) to occasionally crash for no apparent reason thereafter, but enables DHS to prevent a single terrorist attack, it's 100% worth it, and as far as they're concerned, anyone who thinks otherwise is an evil commie terrorist-loving scumbag who hates America.
This story is insane. Prototypes of things involving emerging technology are NEVER, EVER, **EVER** tiny & compact.
In the early 1980s, pre-Commodore Amiga showed off their new computer's prototype at Comdex. It was a rack the size of a small refrigerator stuffed with handmade (wire-wrapped) logic boards. Two years later, it was an attractive-looking desktop computer with nifty open space underneath that was big enough to tuck the keyboard into.
The first version of Android was developed for a device that was a "phone" only in the sense that it could be used to make and receive phone calls, but was REALLY several cubic feet of prototype boards connected with ribbon cables and LITERAL duct tape.
It would be a HUGE mistake for MagicLeap to prematurely commit to a controller design just for the sake of early miniaturization. I'd rather see them implement the controller as an 802.11ad-connected semi-dumb remote frame buffer, and offload the back-end heavy lifting to a desktop PC that's as big as it needs to be to do its job and impress everyone.
The fact is, landfills around the world are littered with the corpses of prematurely-optimized hardware that ended up being inadequate for their intended purpose. That's why first-gen routers usually have more ram, faster processors, and better chipsets than second-gen routers... the first-gen ones are slightly over-engineered to give them headroom to handle more advanced capabilities, while the second-gen ones are pruned back to the bare minimum specs capable of running the first-gen model's firmware 9-15 months after release.
There's even a name for it -- "Dark UI Patterns"
http://darkpatterns.org/ is an entire website about them.
Yeah, just TRY enabling QoS on a consumer-grade AP/Router. It'll KILL your throughput. Why? Because consumer-grade hardware doesn't have sufficiently-fast CPUs to actually inspect network traffic at gigabit speeds and make intelligent traffic-shaping decisions... they just implement "QoS" by arbitrarily limiting the bitrate from any one device to some fraction of what it thinks is the total link rate, exclude traffic on ports used by popular VoIP services, and call it a day.
With QoS enabled, I couldn't get more than ~8mbps to speedtest.net from any device on my LAN. With QoS disabled, I get 50mbps+ without breaking a sweat.
No, Windows 8 was explosive, watery diarrhea. Windows 10 is more like soupy pellets mixed with corn.
> They've become a smaller fish in a much larger pond.
More like a slightly-larger fish in an exponentially-larger ocean. On its least-profitable month during the Great Recession, Microsoft made more money than they did the entire year after Windows 95 came out. Most investors would be *delighted* to experience Microsoft-style failure.
Most of Microsoft's wounds have been either self-inflicted, or more like allergic reactions to imaginary problems -- dumping Windows Mobile *right* at the point when its capabilities stomped all over iOS and Android & it finally started to be "not ugly", self-mutilating Windows while consumed by "Tablet Fever", etc.
A few years ago, MSI *had* a line of modular "whitebox" laptops. So did Clevo. They've become rare, because discrete graphics are almost the only thing *left* to vary... and engineering the cooling to be adequate for high-powered GPUs without going totally overboard for low-powered GPUs is almost impossible.
The thing is, a modern word processor now has capabilities that used to require a desktop publishing app (remember PageMaker?). When you get down to it, WordPerfect 6 and earlier were basically HTML editors, but with their own proprietary tokenization scheme instead of anything SGML-derived).
The big question: can it at least do 5GHz 802.11n, or is it 2.4GHz-only? Most n-only implementations sold NOW are n-only BECAUSE they're 2.4GHz-only (by definition, 802.11ac REQUIRES 5GHz).
In single-family suburbia it might not matter (much), but in most urban residential areas, 2.4GHz wifi has become almost unusable.
I wish the FCC would buy back the upper half of the wi-fi channel centered on channel 14 via eminent domain, then allow wi-fi to use it ONLY as a 20MHz channel limited to ~5-20mW EIRP (or 1-4mW amp power, to reduce certification costs) to give urban users AT LEAST a single channel that's good enough for same-room use, but sufficiently power-limited for a single sheet of drywall to attenuate it by 20-50dB.
In 2006 there were amps that could SWITCH HDMI, but very few that could sit between the source and TV, extract the audio, AND spoof a 5.1-capable EDID. Go to avsforums.com & read the hundreds of angry forum posts from ~2008-2010 when Blu-Ray became real, and most people with "HDMI" receivers discovered that they had broken implementations that would NEVER be allowed to work as advertised.
Then, the HDMI people turned around & fucked everyone a SECOND time with HDCP 2.2. It's madness. They literally expect people to keep throwing away perfectly good amps every 2-4 years just to continue being allowed to have surround sound.
And I can assure you, prior to ~2013, almost NO amps supported 5.1 LPCM. They could do 7.1 DD+, 7.1 DTS+, and 2.0 LPCM, but NOT 5.1 LPCM.
Even NOW, support for 5.1 LPCM is a crapshoot.
Nintendo was too fucking cheap to license Dolby surround, so if your amp didn't have working HDMI audio input (at launch, most of the few HT receivers that supported HDMI had broken implementations) AND support 5.1 LPCM, no surround sound for you. It couldn't even fall back to fucking ProLogic. The goddamn GAMECUBE had better surround-sound support (via ProLogic) than 98% of Wii-U owners ever got to enjoy.
To wit: in 2012, most home theater amps only supported DDS and DTS via S/PDIF. Nintendo decided to go with a standard that most amps (even ones that were only a year or two old) couldn't use... then, or EVER.
Lack of DD5.1 is a major reason why I always get the xb360 version of any game even though I also have a (generally unloved) Wii-U... graphically, Wii-U is no better than xb360, and the 360 has DD 5.1 surround sound & better controllers.
The xb360 is ALSO why I'm able to enjoy DD 5.1 surround with Netflix & Amazon... it's the ONLY platform allows you to enjoy surround sound with streaming video services if your amp lacks hdmi and DD7.1+
Oh, and let's not forget that Wii-U has a Blu-Ray drive, but can't actually PLAY Blu-Ray movies because Nintendo was too cheap to license the nessary IP.
Mass-production vs one-off customization. Even with robots and CAM to handle the grinding, it still takes more time and effort to grind two lens surfaces instead of just one... and more effort to calculate custom lenses instead of blindly grinding another standardized design on the other side.
That said, I think that competition from cheap online labs will force traditional retailers to reduce the cost of freeform lenses and market them more aggressively. If you can take your prescription and buy a mediocre pair of glasses for $15 online, why would you pay $50-100 for an equally-mediocre pair? On the other hand, if a mediocre pair is $15, but a freeform pair is $200, and the freeform pair is sufficiently better to feel like a HUGE improvement, people will pay the higher cost for the quality pair (even IF they buy a cheap mediocre pair online to keep as a spare pair). By (US) law, your "prescription" (sph, cyl, axis, prism) have to be provided on demand so you can use it elsewhere, but the measurements needed to make custom lenses for a pair of frames are still considered proprietary (and are pretty damn hard to measure without the proper equipment), so they aren't required to share them with you. Ten years from now, mediocre online glasses will be regarded the way drugstore reading glasses are now... something that's better than nothing if you're poor or need a spare, but blatantly inferior to "real" glasses.
As far as SV lenses for astigmatism go, the main advantage of two-sided freeform fully-custom over single-sided freeform semi-custom is aesthetics... with control over BOTH surfaces, you can neutralize things like magnification/minification (so your eyes don't look larger or smaller to observers seeing them through the lens), and adjust the base curve near the periphery to allow thinner lenses. If you only have full freedom over one surface, you can optimize for optics at the expense of thicker lenses, or thinner lenses at the cost of optic fidelity, and your ability to neutralize out magnification/minification will be severely constrained.
From what I've been read, single-surface semi-custom freeform is a HUGE step forward from single-surface lenses ground with traditional standard curves, but the difference between single- and double-surface freeform (for SV astigmatism) is more like the difference between 480p24 from a DVD and 720p24 from Blu-Ray... it's there, but it's not nearly as dramatic as the difference between nominal-480i60 from VHS and 480p24 from DVD.
I believe lenses marketed as "aspheric" are basically designed the way freeform lenses are (with raytracing), but are mass-produced in only a few permutations of sph+cyl... and usually, optimized for thinness over optics. Personally, I wouldn't bother with them... they cost almost as much as single-surface freeform lenses, and don't provide nearly as much optical improvement.
Not to mention, the additional anchoring requirements needed to survive an earthquake. In Florida, you can get away with pilings that passively rest on top of shallow bedrock. In California, the pilings have to be anchored to the bedrock through shock absorbers. Otherwise, it could fall over like a top-heavy stack of books.
Until fairly recently, Florida highways had almost NO explicit design features to resist seismic activity. They were LITERALLY built on pilings resting passively on bedrock that were held in place by nothing more than friction & gravity at the bottom, and the bridge deck's beams at the top. It wasn't a seismic failure per se, but a freeway bridge under construction in Miami (97th Avenue over SR836) actually collapsed during Hurricane Wilma... the pilings were up, and the beams were laid on top, but the contractor didn't bother to temporarily anchor & brace them, and the hurricane hit before the deck's concrete was poured. The area was flooded (providing lubrication), the hurricane spawned a tornado, and the combination pushed a beam past its tipping point, ultimately taking down the other beams & columns around it. As a result, contractors are now required to brace the beams with temporary supports until the deck gets poured. My point is, if mere WIND (lubricated by water) could cause a structural failure, just imagine what an underground explosion at a gas station next to a freeway overpass that was strong enough to make the earth visibly ripple could do to a road built like that. In contrast, when I-595 was rebuilt in Fort Lauderdale, the new ramps had additional bracing every few columns to keep the road deck from sliding from side to side if something made the ground vibrate.
The current gold-standard of long-term correction is probably a custom-ground scleral lens. Dr. Gemoules in Dallas fits lenses using the hardware NORMALLY used for laser surgery... he scans both eyes in 3d and orders the first (few) set(s) with the eye's precise shape ground into the back. Once the back side is optimal, he does a wavefront scan while you're wearing those lenses and uses the data to raytrace the ideal front curve.
He's expensive compared to Walmart or Lenscrafters, but often cheaper than what some OTHER doctors charge for less-optimized scleral contacts.
Results vary, of course, but he's had patients walk in who were almost blind without glasses go home with 20/15 (or better) vision.
The catch? You have to physically go to Dallas & spend a week there. He's the only doctor who does it (not sure how much is unwillingness to partner with doctors elsewhe, and how much is due to FDA regulation of officially-experimental treatments... it's possible that current patients are officially & technically part of an ongoing clinical trial). But I'd say about half the things written about him have been "wildly enthusiastic", and only a few have been strongly negative (and most of THEM seem to be just unfortunate patients whose problems simply outstrip the capabilities of ANY currently-existing technology).
Google for "Laserfit AND Gemoules" (sans quotes)
Disclaimer: I'm not a patient of his... but I sure as hell WOULD be if I could afford it. If there's a better alternative available from ANYONE today, I have yet to read about it.
Fyi, if you've had laser surgery, your "glasses" prescription might be minor, but your "contact" prescription will probably be similar in strength to your pre-laser prescription. Why? GP scleral lenses optically fuse with the cornea via tears (watery, not injury), and mostly neutralize whatever lens was burned into the cornea by the laser. Other contacts are similar, but it's the most dramatic with scleral RGP lenses (soft lenses might bend slightly to conform to an irregular cornea).
Also, scleral RGP lenses LOOK scary-huge, but in practice they're more comfortable & stable because they rest on the white part (little/no nerve sensation) instead of the iris/pupil/cornea (lots of nerve sensation).
Except, it's only "peripheral vision" when you're looking straight ahead. The moment you look to one side by moving your eyes, it ceases to be "peripheral", and being in sharp focus becomes VERY important.
It's because 99.9% of progressive lenses have a semi-custom lens ground into one side, but a standardized lens ground into the other, with no regard for getting the optics right anywhere besides the center vertical axis.
It IS possible to make better glasses using raytracing (to design) and custom grinding (to implement) on both sides to "get it right" along the horizontal axis, too... but those lenses are expensive, and you'll only see a dramatic improvement if
1) the optician gets all the measurements precisely right (they rarely do, because full custom lenses are expensive, few stores sell more than one or two pairs a year, and they require more measurements than "normal" lenses that most techs don't fully understand);
2) The glasses are meticulously adjusted for proper alignment... and kept in proper alignment with frequent adjustments.
Realistically, you're looking at glasses that cost about $600-1,000/pair PLUS the frame cost.
The tech was developed for custom progressive lenses, but it can also be used to make better single-vision lenses for people with astigmatism. Normal glasses have a standardized sphere + base curve manufactured into one side, and a non-optimized cylindrical lens ground onto the other side (usually, without regard for lens angle or lens distance FROM pupil). I believe full-custom single-vision lenses run about $500-800 more than "regular" (non-optimized) ones.
The magic word to say & demand when asking about such lenses is "freeform" (often, used in conjunction with "custom" and/or "high-definition"). Just be aware that the optician's skill & experience fitting freeform lenses is ENORMOUSLY important. Even freeform custom lenses can suck if the optician is careless with his/her measurements. And demand the specific word "freeform" -- unlike other marketing terms, "freeform" has a very specific meaning in the industry. Not all lenses advertised as "hi-def" or "custom" are literally "freeform". At least one brand that's "semi-custom" exists that grinds a customized & optimized cyl onto one side of a lens manufactured with a standard sph+base curve on the other. For SV astigmatism, semi-custom might be good enough... but for progressive, you really want full-blown two-surface freeform custom lenses.
Not necessarily.
The South Bay area actually has a SHITLOAD of undeveloped land... it's just that most of it happens to be mud flats adjacent to SF Bay. The catch is, you can't "pull a Florida" and turn that swampland into prime waterfront real estate by dredging out finger canals and using the fill dirt to raise the surrounding terrain, because the first major earthquake will cause all that fill dirt to settle & pack itself down, most likely destroying any single-family home built on top of it.
You COULD, however, build skyscrapers on that land, because skyscrapers sit on pilings that go all the way down to the bedrock. The surrounding land (if fill dirt from dredging) would still settle during a major earthquake, but the skyscrapers themselves would be fine.
The catch is, skyscrapers aren't cheap to build... especially if every one of them has to be an architecturally-spectacular work of art to have any chance in HELL of getting approved by voters. And right now, there's almost zero market demand for skyscrapers in the mid-bay area, because people who could AFFORD a condo in a new bayfront skyscraper would rather live in a mountainside mansion, and people who'd WANT to live in a bayfront condo can't afford to buy one.
Raising wages will initially just create inflation, but beyond a certain point, you'll have SO MUCH MONEY getting thrown around relative to construction costs, developers will say "fuck it" and actually TRY getting skyscraper projects approved, because land costs will be so staggering relative to new construction costs. Or they'll start to do things that would otherwise be cost-prohibitive, like buying every single-family home in a neighborhood (taking the loudest NIMBYs out of the equation) so they can demolish the whole neighborhood and replace the single-family homes with 3-4-story townhomes and 5-10 story apartments & condos.
> Digital music uses 10 times the nyquist rate; better than a human ear can hear and distinguish
192,000hz sample rates are about 10x, but 44,100hz is barely more than the Nyquist minimum.
Nyquist is a MINIMUM. The only thing Nyquist GUARANTEES is that a 44.1khz sample rate is INADEQUATE for reproducing frequencies above 22.05khz. It makes no claim that it's actually good enough for flawless reproduction.
Also, 192khz sample rates are really more like 96khz if you're attempting to mix two streams in realtime due to clock jitter (unless all live streams share an accurate clock signal that compensates for cable length and the speed of light, the bytes won't precisely align).
That said, 24-bit/channel at 192khz is definitely past any point where double-blind listeners could distinguish between live (but non-class-D-amplified) and digital.
Newer technologies DO have a few quantifiable advantantages over CRTs: subpixel-addressability, and no convergence error.
NOBODY circa 2000 thought LCD would EVER achieve the dominance it has today -- almost everyone in the display industry was betting on FED as "the next big thing" (FED is basically a CRT that's directly-illuminated at the subpixel level by an array of solid-state electron emitters sitting behind them). The problem with FED was dead pixels... they were as seemingly-unavoidable with FED as they were with TFT. Ultimately, LCD manufacturers solved the problem first. The reality, though, is that the same technology that now gives us LCDs without dead/stuck pixels could give us FED, too. FED has one problem, though: it uses as much power AS a CRT, so it's not really viable for laptops, phones, etc. OLED basically gives us everything FED could, with the added bonus of (relatively) low power consumption. Even in the early 90s, FED's creators knew LED was the future... it's just that back then, blue LEDs were still a lab & prototype novelty, so they came up with the idea of using a LED-like electron emitter to excite blue phosphors.
Regardless of whether FED still has a future or we go straight to OLED, LCDs *suck* compared to both.
Mechanically, they were similar to the generation or two of turn-of-the-century ultra- ultra- ultra-high-end car CD players that minimized skipping by mechanical means (vibration-dampening, etc) instead of just letting it skip, but reading the disc 4-16x faster & buffering enough to keep the output pipeline filled during the recovery. Which, incidentally, sounded WAY better than the ones that relied on buffering since audio cds have no concept of tracks or sectors, so figuring out precisely WHERE to continue after a skip event is problematic (redbook audio has no FEC, so you could potentially read the same chunk 5 times & get 2-5 chunks that differ by at least a single byte. That's why DAE apps like Exact Audio Copy read everything multiple times, try to line up the chunks, and determine which byte is the right value, and which ones were just read errors).