I did that, and in response they killed the product line.
I have a Nexus 5. I skipped the 5X because it didn't seem like a significant upgrade. Then they killed off the Nexus line for the twice-as-pricey Pixel, now on its second iteration. Now I actually need to upgrade, and I can't find any good 5", ~$300 phones with minimal OS changes and good long-term support. The Nokia 5.1 looks like just what I need, but it hasn't hit the US market yet.
Citation needed on the "would help us economically". AFAIHR, it would basically be as effective as handing a few hundred billion dollars to a couple construction conglomerates - once the construction is done, how does the wall itself improve the economy? It produces no value - no products or services that are needed or desired. And if we're going to stimulate the economy by throwing money at projects, how about ones that actually provide value - overhaul the highways and bridges (lower transportation costs = lower economic friction), or build some protection against natural disasters (a penny of prevention is worth a dollar of repair).
I'm also calling false dilemma on your "either we get a wall or we get unrestricted immigration". Nobody is arguing in favor of unrestricted immigration, and the status quo is in fact heavily restricted immigration. The best argument against the wall is that it already exists in the places where the cost/benefit makes sense to have a wall. Our net immigration with Mexico itself is negative - more people now emigrate to Mexico than immigrate from. The total flux is only positive because of immigrants from central/south America via Mexico. If your sole concern is preventing illegal immigration to the US, the best place to build the wall would be on the borders between Mexico and Guatemala, and Mexico and Belize. ("Mexico is going to build a wall, and we're gonna pay for it!", as it were)
The general argument being made by "us liberals" is: 1) Building a wall is a fuckton of money and a not insubstantial amount of environmental damage for negligible impact on immigration 2) We should not violate our own laws regarding due process, search and seizure in an attempt to enforce immigration laws 3) We really, really should not violate international humanitarian law while enforcing immigration (the "detention camps" are definitely breaking a few of these). 4) Illegal immigrants who have a valid claim for asylum should not be deported unless necessary - it is generally unethical to deport someone if it will result in their death at the hands of an angry dictator or warlord.
Additionally, I would present a further argument, which is not generally discussed among liberals so I'm not sure if it's popular or not: 5) Abuse of legal visas (H-1B etc.) causes economic problems on par with or greater than those caused by illegal entry, and a wall does jack shit to stop it.
Depending on the other power sources on the grid, it can make sense to pump.
Nuclear plants don't like to vary their output. They function best when outputting a constant amount of power 24/7. But power demand is roughly sinusoidal - a peak during the afternoon, and a trough in the early morning. So often they let the reservoir drain during the peak hours, and then use the excess power overnight to pump it back up. I've vacationed at an artificial lake that did this - the water level cycled by about two feet over the day, sort of like an artificial tide.
Solar plants vary their output daily. They produce a ton of power during the day, then none at night, unless they're one of those weird thermo-solar things. So you use the excess power during the day to pump water up, then let it drain overnight when the solar plants are dark (pun absolutely intended).
Wind and solar plants can overproduce. There have been times where the regional spot price of electricity went negative - the grid would pay you to use electricity, because they have so much they'll start going out of spec if the load doesn't climb to match. Even without going negative (that's usually the result of weird financial/political incentives), it makes sense to store excess power, then release it when the supply is low and demand is high, especially when the infrastructure to do so already exists.
A modern "core" has several "execution units". Unlike very early x86, where it was divided between ALU and FPU, these are divided more finely and evenly - one might do integer math, vector shifts, and branches, while another might do integer math, vector logic, and data stores. There's usually redundancy on common instruction types (eg. Haswell has three that can do address stores, but only one can do divides).
In a single thread, this is used for superscalar execution. If you have code something like "a = b / c; d = e * f;", both instructions can be run in parallel since neither depends on the other. This also hides the cost of x86's more complicated addressing modes - computing the address gets dispatched to an execution unit just like a normal multiply/add, and the result just gets sent to the store unit.
But sometimes a thread has lots of dependencies, or does mainly a single type of operation. Maybe it's crunching through a bunch of multiply-adds. Rather than let the remainder of the core sit idle, you can run another thread, or even another process, on it. If this second one mainly hits a different EU - say, it's doing a lot of shifting and bit-twiddling - you can get a 100% speedup.
You rarely get so much of a boost in practice. A worker-thread type of program, splitting a parallel task across cores, will generally be using the same execution units in each thread. And SMT doesn't help if you're bottlenecked on something besides execution - well-optimized code, as often as not, is limited by memory throughput rather than execution.
The other boost comes from covering memory latency. If one thread hits a load that isn't in L1 cache, it will stall while the load is served. If it's in L2 cache, that's not too long - a dozen cycles or so. If you're going out to main memory, you're looking at a few hundred, maybe a few thousand cycles of NOPs - so why not switch to another thread, that has all it's data in L1 cache already? Modern x86 processors have pretty low memory latency compared to other architectures, so two threads is generally the most you'd find useful for this, but other systems with harsher memory latency will go even wider - the latter-day SPARCs do eight threads per core, and some parts of a GPU will operate in the hundreds. This is why some non-superscalar architectures will still have multiple threads per core - it's only ever actually running one instruction, but it will rarely be running zero.
I tried desktop RSS feed readers many years ago, back in the Firefox 3.0 era, when the RSS update happened in the main UI thread and would draw your browser down to an absolute crawl. Nothing I found worked the way I actually use RSS feeds.
I mainly use Live Bookmarks for webcomics. I have around a hundred I follow. Every day, I roll down a big folder on my bookmarks bar that has a Live Bookmark "folder" for each of them, and read any unread entries, one at a time. This does take a while but most of that time is spent reading, not on the RSS process because I can just hover over an LB and see if I need to read anything. I've manually sorted them into groups by how often they're supposed to update - daily/5-a-week/3-a-week/weekly/etc. which speeds the process up even more. And the few webcomics that don't have a working RSS feed fit into the process nicely - they happen to have mostly-consistent schedules so I just remember "hey, it's wednesday, ___ should have updated". It's also not uncommon for a webcomic to go on a lengthy hiatus. I have a subfolder of "maybe dead?" LBs which I check less often.
I really do not want to miss even one update (most of the ones I read are story-based), which RSS readers designed for skimming a news feed generally don't focus on. In particular, ones that just shovel everything into one ginormous stream won't work for me. Live Bookmarks follows my browser history so it shows which entries I have and have not read, so even if I skip a few days I won't be unable to catch back up.
As extreme cases, it also needs to be able to deal with an RSS feed with several hundred entries in it (one comic includes its entire decade-plus archive into the feed), several feeds that include entries I don't want to read (news/blog entries), and one that goes in reverse order for whatever fucking reason (newest item at bottom).
I read from multiple devices, so Firefox syncing history between them mostly keeps that in line (this is maybe 50/50 whether it actually works or not, because Firefox doesn't sync redirect pages and like half the RSS feeds I read don't give canonical URLs).
Do you (or anyone else) have any desktop RSS feed reader that works for my use case as well as, or better than, Firefox Live Bookmarks?
New Shepard is a testbed for their orbital tech, which will be the New Glenn rocket. Specifically, where New Shepard uses a single BE-3 engine, New Glenn's second stage will use two BE-3 engines. The BE-3 engine is designed more for the orbital role than the suborbital one - hydrolox engines have very high Isp but low propellant density, making them much better suited for high altitude flight. The BE-4 engines that will power New Glenn's first stage use a methalox chemistry, and have been built and test-fired (they're also a leading contender for ULA's new Vulcan rocket, although due to typical government fuckery, Aerojet is somehow still in the running).
There's also the whole reusability thing. New Shepard is, in many ways, an equivalent of SpaceX's Grasshopper rocket - which was a suborbital-only test article, which made numerous test flights as they figured out the basics of vertical landing.
While there is indeed a very large gap between suborbital and orbital flights, it is not so large as you seem to think.
No decision either way has been made. The Aerospace Safety Advisory Panel has listed several hazards that need to be mitigated, whether by changing loading order or through other means (and, speaking informally, members of the panel suggested other methods were completely possible). A decision is "expected shortly", but that was two months ago and I couldn't find anything newer, so either "shortly" doesn't mean what it normally does, or the decision was not made widely known.
Blow does not care that his language sees wide adoption. The only user it *needs* to satisfy is himself - if nobody else uses it, that's their problem, not is. He's not trying to kill C++, he's trying to leave it. As a bizarrely wealthy*, completely independent developer, he's in a position to do that.
As a completely independent dev, I am watching Jai with interest. C++ is a bad language - it makes things that ought to be easy, very difficult, and things that ought to be difficult, easy. I am looking at several other C++ replacements, particularly Rust, but Jai looks the most appealing to me.
You are correct with regards to the larger industry - companies with large existing codebases cannot simply rewrite everything on a whim, and the C++ committee is limited in how much they can improve, because they cannot break backwards compatibility. But that is no reason why new projects must be in C++, nor why new languages cannot succeed.
* I have played both Braid and The Witness. They were okay. A bit hollowly pretentious. I don't see why they're put on such high pedestals, or how they earned him so much money. I don't begrudge him his success, but I am a bit mystified by it.
Is it really? Soyuz-2 has had quite a few failures (four complete, three partial, out of 77 launches, and not clustered towards the start as you normally see). Soyuz-FG (the one currently used for crewed flights) may have made 66 successful launches in a row, but they're clearly having quality issues with Soyuz-2.
I'm wondering when Soyuz will end. It's a horribly inefficient design by modern standards, even with the updates they've been doing. No insult to Korolyov, it was a great rocket for its time... but it's time is long past.
Soyuz's payload fits in between Angara 1 and Angara 5, which is probably why they aren't yet planning to discontinue Soyuz. Like Falcon and Delta IV, Angara is built around a small common design, which can be used as a side-mounted booster for heavier payloads, except in their case, they're strapping four boosters around the central core instead of one, to make the heavy Angara 5 which is replacing Proton.
I see an opening for a two-booster Angara 3. I think it would end up being somewhere between 150% and 200% the lifting capacity of Soyuz, which makes it less than ideal as a drop-in replacement, but should be serviceable as a lineup replacement.
Of course, the continued flight of any Russian rocket (for anything but Russian military/intelligence payloads) kind of depends on them getting some form of reusability. They designed a folding-wing, horizontal-landing version of the Angara URM, but apparently they don't have the funding to actually build it.
Often, management is relatively clueless about their own operations. They don't always know who the critical members of their team are.
A few years back, I was pushed out of a company because the chief of operations thought I was their weakest developer, and that the actual weakest dev was their rockstar. I was heavily specialized in database and backend coding, which management didn't see as important because they never got client complaints about it. Rather than seeing that as "this dev writes solid code", they saw it as "this dev is slower than Rockstar". Now, Rockstar would push code to production as soon as it compiled, only getting away with it because his main task was endless revisions on a social media network with maybe a dozen users that weren't funding the project, and he constantly relied on other devs to do anything outside his limited skillset (I wrote basically every database query for that project, despite never being assigned to it).
Had there been a "jury of my peers", I'd have been completely vindicated. But, they weren't the type to listen to their peons, let alone ask their opinion. So, out the door I went. And half the company followed in the next year.
Now, in my case, there was no higher level of management, and few people will set up a system to protect their company from their own mistakes. Amazon, however, is big enough to have several layers of management between the ground level and the board room. They can benefit from a mechanism to protect from bad frontline managers' staffing decisions. This is pretty unconventional, and I can see some potential defects in their specific implementation, but the principle seems sound.
I have often seen Moore's Law formulated as transistors per unit cost, which I think is a useful measurement. If we hit a wall on feature size, this could still allow continually improved performance (and to a lesser extent, performance per watt).
There are still programmers who optimize at that level - and then go one further, by pipelining in such a way that the core can execute both threads at as close to full speed as possible. Usually this ends when you're processing data as fast as the L1 cache can prefetch it - with SIMD instructions, you can hit 32 bytes/clock/thread (two 16-byte operations in one clock), while the L1 cache on the current-gen processors can read 64 bytes/clock/core.
This isn't done on every program, or even most programs, and nobody's optimizing their entire codebase to this level, but for stuff like compressors/decompressors, or codecs, yeah, there's still programmers who will optimize all the way down to the metal.
TFA includes a claim that the fix for this hole will actually improve performance. I don't see how that's possible - the problem is caused by a feature that improves performance, and the fix, as I understand it, is simply to not use that feature. Is there something I'm missing?
Rare seems to still be doing okay. They haven't been cranking out hits as reliably as they did in the 90s, but neither have the devs who left Rare to start separate companies (Free Radical/Crytek UK had flubs like Haze, and Playtonic's Yooka-Laylee was a mixed bag).
Swatting is like an auto-immune disease, of a nation instead of a person. The parts that are supposed to protect us (immune system/police system) have become so aggressive and powerful that they can be easily tricked into acting against us.
When a person has an auto-immune disease, we treat them by both suppressing the immune system (bringing it back down to normal, safe levels) and by eliminating any foreign bodies that are triggering the response. When a nation has it, I think it is sensible to do the same - demilitarize the police force, improve training, make it so that fake police calls don't regularly end in dead innocents, but also go after the bad actors who are trying to take advantage of an over-aggressive police response.
Signing up for Comcast at a new apartment. Selected "use my own modem", because I still have the DOCSIS 3.0 modem I used the last time I had Comcast service. It let me skip the modem rental fee, but the website didn't allow me to not schedule an appointment to have a "professional" install it, nor skip the $90 fee that would entail. I picked up the phone and got it sorted out - apparently the previous tenant didn't schedule to disconnect their service, so the system insisted someone needed to go out and uninstall whatever was there. The service rep was able to sort it out for me, but I imagine a lot of people wouldn't bother picking up the phone and waiting on hold listening to badly-bandlimited Vivaldi for five minutes.
*Things* should be invariant. But labels for those things can change over varying place, if the label is inherently locational. What separates an asteroid from a meteorite, but whether it's on the ground or not?
The term "planet" is already somewhat orbit-based, because otherwise many of the moons, such as Titan, would be classified as planets, but I've never heard anyone argue it should be. That is to say, the intuitive understanding of the word "planet" seems to contain "directly orbits a star" as one of its criteria.
I would much rather resolve the ejected planet issue you found by classifying such things as "rogue planets" - things that would be planets except for that one criterion of a stellar orbit.
Yes. At the very least, any reasonable definition that excludes the "cleared the neighborhood" requirement would mean that Ceres would go back to planet status (currently dwarf planet, previously asteroid), as would Eris, and possibly the other dwarf planets.
Personally, I am of the opinion that "planet" is correctly defined, except for the "orbiting Sol" part, which should be revised to "orbiting a star" (with "star" defined as "non-singularity body capable of fusing hydrogen", which excludes deuterium-burning brown dwarfs but includes all other stars). "Planet" is primarily an orbit-based characteristic; it fits in a spectrum including "star", "moon", and "small fragmentary body" (dwarf planets, asteroids, comets, KBOs, etc.). I would gladly accept a second set of terms based on physical characteristics, such as "rocky body" (Earth, Luna, Ceres, Mercury), "gas giant" (Jupiter, Saturn), "dirtball" (small asteroids and comets, many moons), and so on.
No, just the prevailing wind patterns. Wind will blow wherever the hell it wants to, but for nuclear fallout planning purposes, it's normal to use prevailing winds.
We tried doing that with Syria, letting them just sort their own shit out. The resulting civil war led to a refugee crisis and the rapid growth of ISIS, and then let Russia expand its military reach into the Mediterranean.
A Saudi-Iran war would result in a refugee crisis bigger than any since WW2, an oil crisis bigger than any since ever, and if it went nuclear (Israel is a known-but-undeclared nuclear power, Iran and Saudi Arabia are just a serious political push and a year away from building their own nukes), a radioactive crisis when the winds carry it either eastward towards China, or southwestward into Africa.
Peace, if possible, is a vastly preferable alternative.
It wasn't a legally enacted treaty - never went to Congress for approval as all treaties must.
It's not a treaty. It's an agreement. Iran agreed to do a thing, the UNSC permanent members and the EU agreed to do a thing, all within the bounds of their respective executive powers. Congress's approval was not necessary, because nothing in the deal required legislative authority.
We were prevented from inspecting numerous locations considered 'military' by Iran's leaders - which is the most likely place to develop a nuclear program.
False - that is categorically and unquestionably incorrect.
The agreement provided for guaranteed inspection of *any* location the International Atomic Energy Agency inspectors deem potentially in violation. Iran has a limited ability to push back - they have a 24-day window to negotiate an alternative, but if we decide we *need* to see it, we will see it or the sanctions will kick back in. 24 days is not enough to hide a nuclear weapons facility from close inspection - particularly not when we have satellite surveillance and can easily see any large movement of equipment and materiel away from the site.
Additionally, a term of the agreement required Iran to accede to the "Additional Protocol", which has even more stringent requirements allowing short-notice inspections of any site by the IAEA - and that protocol will *not* expire with the rest of the agreement.
As I type this the news on tv is showing me Schumer, Menendez, and other democrats speaking AGAINST the Iran deal in 2015 - who now oddly embrace the deal they were against because Trump ended it.
Schumer and Menendez were the *only* two Democrat senators to oppose the deal. A symbolic resolution decrying the bill was passed through the House on party-line vote, and was never formally voted on in the Senate due to lack of sufficient votes. And I have not seen either of them publicly support the deal to this day. I strongly suspect your sources are being misleading on this, as they clearly are on other issues.
Slashdot Mirror
I did that, and in response they killed the product line.
I have a Nexus 5. I skipped the 5X because it didn't seem like a significant upgrade. Then they killed off the Nexus line for the twice-as-pricey Pixel, now on its second iteration. Now I actually need to upgrade, and I can't find any good 5", ~$300 phones with minimal OS changes and good long-term support. The Nokia 5.1 looks like just what I need, but it hasn't hit the US market yet.
Citation needed on the "would help us economically". AFAIHR, it would basically be as effective as handing a few hundred billion dollars to a couple construction conglomerates - once the construction is done, how does the wall itself improve the economy? It produces no value - no products or services that are needed or desired. And if we're going to stimulate the economy by throwing money at projects, how about ones that actually provide value - overhaul the highways and bridges (lower transportation costs = lower economic friction), or build some protection against natural disasters (a penny of prevention is worth a dollar of repair).
I'm also calling false dilemma on your "either we get a wall or we get unrestricted immigration". Nobody is arguing in favor of unrestricted immigration, and the status quo is in fact heavily restricted immigration. The best argument against the wall is that it already exists in the places where the cost/benefit makes sense to have a wall. Our net immigration with Mexico itself is negative - more people now emigrate to Mexico than immigrate from. The total flux is only positive because of immigrants from central/south America via Mexico. If your sole concern is preventing illegal immigration to the US, the best place to build the wall would be on the borders between Mexico and Guatemala, and Mexico and Belize. ("Mexico is going to build a wall, and we're gonna pay for it!", as it were)
The general argument being made by "us liberals" is:
1) Building a wall is a fuckton of money and a not insubstantial amount of environmental damage for negligible impact on immigration
2) We should not violate our own laws regarding due process, search and seizure in an attempt to enforce immigration laws
3) We really, really should not violate international humanitarian law while enforcing immigration (the "detention camps" are definitely breaking a few of these).
4) Illegal immigrants who have a valid claim for asylum should not be deported unless necessary - it is generally unethical to deport someone if it will result in their death at the hands of an angry dictator or warlord.
Additionally, I would present a further argument, which is not generally discussed among liberals so I'm not sure if it's popular or not:
5) Abuse of legal visas (H-1B etc.) causes economic problems on par with or greater than those caused by illegal entry, and a wall does jack shit to stop it.
Depending on the other power sources on the grid, it can make sense to pump.
Nuclear plants don't like to vary their output. They function best when outputting a constant amount of power 24/7. But power demand is roughly sinusoidal - a peak during the afternoon, and a trough in the early morning. So often they let the reservoir drain during the peak hours, and then use the excess power overnight to pump it back up. I've vacationed at an artificial lake that did this - the water level cycled by about two feet over the day, sort of like an artificial tide.
Solar plants vary their output daily. They produce a ton of power during the day, then none at night, unless they're one of those weird thermo-solar things. So you use the excess power during the day to pump water up, then let it drain overnight when the solar plants are dark (pun absolutely intended).
Wind and solar plants can overproduce. There have been times where the regional spot price of electricity went negative - the grid would pay you to use electricity, because they have so much they'll start going out of spec if the load doesn't climb to match. Even without going negative (that's usually the result of weird financial/political incentives), it makes sense to store excess power, then release it when the supply is low and demand is high, especially when the infrastructure to do so already exists.
A lot of them, actually.
A modern "core" has several "execution units". Unlike very early x86, where it was divided between ALU and FPU, these are divided more finely and evenly - one might do integer math, vector shifts, and branches, while another might do integer math, vector logic, and data stores. There's usually redundancy on common instruction types (eg. Haswell has three that can do address stores, but only one can do divides).
In a single thread, this is used for superscalar execution. If you have code something like "a = b / c; d = e * f;", both instructions can be run in parallel since neither depends on the other. This also hides the cost of x86's more complicated addressing modes - computing the address gets dispatched to an execution unit just like a normal multiply/add, and the result just gets sent to the store unit.
But sometimes a thread has lots of dependencies, or does mainly a single type of operation. Maybe it's crunching through a bunch of multiply-adds. Rather than let the remainder of the core sit idle, you can run another thread, or even another process, on it. If this second one mainly hits a different EU - say, it's doing a lot of shifting and bit-twiddling - you can get a 100% speedup.
You rarely get so much of a boost in practice. A worker-thread type of program, splitting a parallel task across cores, will generally be using the same execution units in each thread. And SMT doesn't help if you're bottlenecked on something besides execution - well-optimized code, as often as not, is limited by memory throughput rather than execution.
The other boost comes from covering memory latency. If one thread hits a load that isn't in L1 cache, it will stall while the load is served. If it's in L2 cache, that's not too long - a dozen cycles or so. If you're going out to main memory, you're looking at a few hundred, maybe a few thousand cycles of NOPs - so why not switch to another thread, that has all it's data in L1 cache already? Modern x86 processors have pretty low memory latency compared to other architectures, so two threads is generally the most you'd find useful for this, but other systems with harsher memory latency will go even wider - the latter-day SPARCs do eight threads per core, and some parts of a GPU will operate in the hundreds. This is why some non-superscalar architectures will still have multiple threads per core - it's only ever actually running one instruction, but it will rarely be running zero.
I tried desktop RSS feed readers many years ago, back in the Firefox 3.0 era, when the RSS update happened in the main UI thread and would draw your browser down to an absolute crawl. Nothing I found worked the way I actually use RSS feeds.
I mainly use Live Bookmarks for webcomics. I have around a hundred I follow. Every day, I roll down a big folder on my bookmarks bar that has a Live Bookmark "folder" for each of them, and read any unread entries, one at a time. This does take a while but most of that time is spent reading, not on the RSS process because I can just hover over an LB and see if I need to read anything. I've manually sorted them into groups by how often they're supposed to update - daily/5-a-week/3-a-week/weekly/etc. which speeds the process up even more. And the few webcomics that don't have a working RSS feed fit into the process nicely - they happen to have mostly-consistent schedules so I just remember "hey, it's wednesday, ___ should have updated". It's also not uncommon for a webcomic to go on a lengthy hiatus. I have a subfolder of "maybe dead?" LBs which I check less often.
I really do not want to miss even one update (most of the ones I read are story-based), which RSS readers designed for skimming a news feed generally don't focus on. In particular, ones that just shovel everything into one ginormous stream won't work for me. Live Bookmarks follows my browser history so it shows which entries I have and have not read, so even if I skip a few days I won't be unable to catch back up.
As extreme cases, it also needs to be able to deal with an RSS feed with several hundred entries in it (one comic includes its entire decade-plus archive into the feed), several feeds that include entries I don't want to read (news/blog entries), and one that goes in reverse order for whatever fucking reason (newest item at bottom).
I read from multiple devices, so Firefox syncing history between them mostly keeps that in line (this is maybe 50/50 whether it actually works or not, because Firefox doesn't sync redirect pages and like half the RSS feeds I read don't give canonical URLs).
Do you (or anyone else) have any desktop RSS feed reader that works for my use case as well as, or better than, Firefox Live Bookmarks?
The current year is 2771 AUC, and nobody can convince me otherwise!
New Shepard is a testbed for their orbital tech, which will be the New Glenn rocket. Specifically, where New Shepard uses a single BE-3 engine, New Glenn's second stage will use two BE-3 engines. The BE-3 engine is designed more for the orbital role than the suborbital one - hydrolox engines have very high Isp but low propellant density, making them much better suited for high altitude flight. The BE-4 engines that will power New Glenn's first stage use a methalox chemistry, and have been built and test-fired (they're also a leading contender for ULA's new Vulcan rocket, although due to typical government fuckery, Aerojet is somehow still in the running).
There's also the whole reusability thing. New Shepard is, in many ways, an equivalent of SpaceX's Grasshopper rocket - which was a suborbital-only test article, which made numerous test flights as they figured out the basics of vertical landing.
While there is indeed a very large gap between suborbital and orbital flights, it is not so large as you seem to think.
No decision either way has been made. The Aerospace Safety Advisory Panel has listed several hazards that need to be mitigated, whether by changing loading order or through other means (and, speaking informally, members of the panel suggested other methods were completely possible). A decision is "expected shortly", but that was two months ago and I couldn't find anything newer, so either "shortly" doesn't mean what it normally does, or the decision was not made widely known.
Blow does not care that his language sees wide adoption. The only user it *needs* to satisfy is himself - if nobody else uses it, that's their problem, not is. He's not trying to kill C++, he's trying to leave it. As a bizarrely wealthy*, completely independent developer, he's in a position to do that.
As a completely independent dev, I am watching Jai with interest. C++ is a bad language - it makes things that ought to be easy, very difficult, and things that ought to be difficult, easy. I am looking at several other C++ replacements, particularly Rust, but Jai looks the most appealing to me.
You are correct with regards to the larger industry - companies with large existing codebases cannot simply rewrite everything on a whim, and the C++ committee is limited in how much they can improve, because they cannot break backwards compatibility. But that is no reason why new projects must be in C++, nor why new languages cannot succeed.
* I have played both Braid and The Witness. They were okay. A bit hollowly pretentious. I don't see why they're put on such high pedestals, or how they earned him so much money. I don't begrudge him his success, but I am a bit mystified by it.
Is it really? Soyuz-2 has had quite a few failures (four complete, three partial, out of 77 launches, and not clustered towards the start as you normally see). Soyuz-FG (the one currently used for crewed flights) may have made 66 successful launches in a row, but they're clearly having quality issues with Soyuz-2.
I'm wondering when Soyuz will end. It's a horribly inefficient design by modern standards, even with the updates they've been doing. No insult to Korolyov, it was a great rocket for its time... but it's time is long past.
Soyuz's payload fits in between Angara 1 and Angara 5, which is probably why they aren't yet planning to discontinue Soyuz. Like Falcon and Delta IV, Angara is built around a small common design, which can be used as a side-mounted booster for heavier payloads, except in their case, they're strapping four boosters around the central core instead of one, to make the heavy Angara 5 which is replacing Proton.
I see an opening for a two-booster Angara 3. I think it would end up being somewhere between 150% and 200% the lifting capacity of Soyuz, which makes it less than ideal as a drop-in replacement, but should be serviceable as a lineup replacement.
Of course, the continued flight of any Russian rocket (for anything but Russian military/intelligence payloads) kind of depends on them getting some form of reusability. They designed a folding-wing, horizontal-landing version of the Angara URM, but apparently they don't have the funding to actually build it.
Often, management is relatively clueless about their own operations. They don't always know who the critical members of their team are.
A few years back, I was pushed out of a company because the chief of operations thought I was their weakest developer, and that the actual weakest dev was their rockstar. I was heavily specialized in database and backend coding, which management didn't see as important because they never got client complaints about it. Rather than seeing that as "this dev writes solid code", they saw it as "this dev is slower than Rockstar". Now, Rockstar would push code to production as soon as it compiled, only getting away with it because his main task was endless revisions on a social media network with maybe a dozen users that weren't funding the project, and he constantly relied on other devs to do anything outside his limited skillset (I wrote basically every database query for that project, despite never being assigned to it).
Had there been a "jury of my peers", I'd have been completely vindicated. But, they weren't the type to listen to their peons, let alone ask their opinion. So, out the door I went. And half the company followed in the next year.
Now, in my case, there was no higher level of management, and few people will set up a system to protect their company from their own mistakes. Amazon, however, is big enough to have several layers of management between the ground level and the board room. They can benefit from a mechanism to protect from bad frontline managers' staffing decisions. This is pretty unconventional, and I can see some potential defects in their specific implementation, but the principle seems sound.
I have often seen Moore's Law formulated as transistors per unit cost, which I think is a useful measurement. If we hit a wall on feature size, this could still allow continually improved performance (and to a lesser extent, performance per watt).
There are still programmers who optimize at that level - and then go one further, by pipelining in such a way that the core can execute both threads at as close to full speed as possible. Usually this ends when you're processing data as fast as the L1 cache can prefetch it - with SIMD instructions, you can hit 32 bytes/clock/thread (two 16-byte operations in one clock), while the L1 cache on the current-gen processors can read 64 bytes/clock/core.
This isn't done on every program, or even most programs, and nobody's optimizing their entire codebase to this level, but for stuff like compressors/decompressors, or codecs, yeah, there's still programmers who will optimize all the way down to the metal.
TFA includes a claim that the fix for this hole will actually improve performance. I don't see how that's possible - the problem is caused by a feature that improves performance, and the fix, as I understand it, is simply to not use that feature. Is there something I'm missing?
Gizmodo just discovered what a PRNG is
Rare seems to still be doing okay. They haven't been cranking out hits as reliably as they did in the 90s, but neither have the devs who left Rare to start separate companies (Free Radical/Crytek UK had flubs like Haze, and Playtonic's Yooka-Laylee was a mixed bag).
Agreed.
Swatting is like an auto-immune disease, of a nation instead of a person. The parts that are supposed to protect us (immune system/police system) have become so aggressive and powerful that they can be easily tricked into acting against us.
When a person has an auto-immune disease, we treat them by both suppressing the immune system (bringing it back down to normal, safe levels) and by eliminating any foreign bodies that are triggering the response. When a nation has it, I think it is sensible to do the same - demilitarize the police force, improve training, make it so that fake police calls don't regularly end in dead innocents, but also go after the bad actors who are trying to take advantage of an over-aggressive police response.
Modded -1: Not a car analogy
Signing up for Comcast at a new apartment. Selected "use my own modem", because I still have the DOCSIS 3.0 modem I used the last time I had Comcast service. It let me skip the modem rental fee, but the website didn't allow me to not schedule an appointment to have a "professional" install it, nor skip the $90 fee that would entail. I picked up the phone and got it sorted out - apparently the previous tenant didn't schedule to disconnect their service, so the system insisted someone needed to go out and uninstall whatever was there. The service rep was able to sort it out for me, but I imagine a lot of people wouldn't bother picking up the phone and waiting on hold listening to badly-bandlimited Vivaldi for five minutes.
*Things* should be invariant. But labels for those things can change over varying place, if the label is inherently locational. What separates an asteroid from a meteorite, but whether it's on the ground or not?
The term "planet" is already somewhat orbit-based, because otherwise many of the moons, such as Titan, would be classified as planets, but I've never heard anyone argue it should be. That is to say, the intuitive understanding of the word "planet" seems to contain "directly orbits a star" as one of its criteria.
I would much rather resolve the ejected planet issue you found by classifying such things as "rogue planets" - things that would be planets except for that one criterion of a stellar orbit.
Yes. At the very least, any reasonable definition that excludes the "cleared the neighborhood" requirement would mean that Ceres would go back to planet status (currently dwarf planet, previously asteroid), as would Eris, and possibly the other dwarf planets.
Personally, I am of the opinion that "planet" is correctly defined, except for the "orbiting Sol" part, which should be revised to "orbiting a star" (with "star" defined as "non-singularity body capable of fusing hydrogen", which excludes deuterium-burning brown dwarfs but includes all other stars). "Planet" is primarily an orbit-based characteristic; it fits in a spectrum including "star", "moon", and "small fragmentary body" (dwarf planets, asteroids, comets, KBOs, etc.). I would gladly accept a second set of terms based on physical characteristics, such as "rocky body" (Earth, Luna, Ceres, Mercury), "gas giant" (Jupiter, Saturn), "dirtball" (small asteroids and comets, many moons), and so on.
No, just the prevailing wind patterns. Wind will blow wherever the hell it wants to, but for nuclear fallout planning purposes, it's normal to use prevailing winds.
We tried doing that with Syria, letting them just sort their own shit out. The resulting civil war led to a refugee crisis and the rapid growth of ISIS, and then let Russia expand its military reach into the Mediterranean.
A Saudi-Iran war would result in a refugee crisis bigger than any since WW2, an oil crisis bigger than any since ever, and if it went nuclear (Israel is a known-but-undeclared nuclear power, Iran and Saudi Arabia are just a serious political push and a year away from building their own nukes), a radioactive crisis when the winds carry it either eastward towards China, or southwestward into Africa.
Peace, if possible, is a vastly preferable alternative.
It wasn't a legally enacted treaty - never went to Congress for approval as all treaties must.
It's not a treaty. It's an agreement. Iran agreed to do a thing, the UNSC permanent members and the EU agreed to do a thing, all within the bounds of their respective executive powers. Congress's approval was not necessary, because nothing in the deal required legislative authority.
We were prevented from inspecting numerous locations considered 'military' by Iran's leaders - which is the most likely place to develop a nuclear program.
False - that is categorically and unquestionably incorrect.
The agreement provided for guaranteed inspection of *any* location the International Atomic Energy Agency inspectors deem potentially in violation. Iran has a limited ability to push back - they have a 24-day window to negotiate an alternative, but if we decide we *need* to see it, we will see it or the sanctions will kick back in. 24 days is not enough to hide a nuclear weapons facility from close inspection - particularly not when we have satellite surveillance and can easily see any large movement of equipment and materiel away from the site.
Additionally, a term of the agreement required Iran to accede to the "Additional Protocol", which has even more stringent requirements allowing short-notice inspections of any site by the IAEA - and that protocol will *not* expire with the rest of the agreement.
As I type this the news on tv is showing me Schumer, Menendez, and other democrats speaking AGAINST the Iran deal in 2015 - who now oddly embrace the deal they were against because Trump ended it.
Schumer and Menendez were the *only* two Democrat senators to oppose the deal. A symbolic resolution decrying the bill was passed through the House on party-line vote, and was never formally voted on in the Senate due to lack of sufficient votes. And I have not seen either of them publicly support the deal to this day. I strongly suspect your sources are being misleading on this, as they clearly are on other issues.