But speculating that a big automotive competitor is possibly involved sounds nutty even if true. He should have quit when he was ahead and left that out.
Not sure why sabotage is any more unlikely than the corporate espionage that is alleged to have already happened in the self-driving car space.
This can also be seen in places like the StackOverflow user survey where you can see that women and racial minorities make up a larger percentage of the less experienced user base. All things being equal companies should be getting more skin-level diverse over time and not stuck in place like Google has gotten themselves.
Comparing the "less experienced user base" to people that a company like Google is likely to hire is misleading, because most people in that category do not have adequate skills or knowledge to get hired at any of the major tech companies. Also, you're comparing a location-neutral survey against location-specific hiring. Stack Overflow would have to break those numbers by geographical area to draw any useful comparisons, both because of the high cost of living in tech-heavy cities and because of limits on the number of people a company can bring in from other countries.
Worth noting: the minorities where the Stack Overflow survey showed the biggest gains were Asians and Middle Easterners. Asians are already overrepresented by about a factor of four at Google, assuming you're using the U.S. population as a basis. (If you're using the world population, Asians are still probably underrepresented, but there are geopolitical reasons that make such comparisons problematic). And Google doesn't even list "Middle Eastern" as an ethnicity in their report, so I can only assume that they are all lumped in under "white", making comparisons impossible.
That said, you're right that there is a big discrepancy between the percentage of blacks and Hispanics graduating with CS degrees and the percentage who are getting employed by major tech companies. I don't know what's going on there. It could be bias in hiring, or it could be that we're seeing the same effect among minorities now that we saw among whites in the 1990s, where everybody majored in CS even if they were bad at it, hoping to get a high-paying job. Realistically, it is probably some combination of both. The former problem is easy to test, though nontrivial to correct. The latter, of course, can be solved in basically the same way as we solve the problem of inadequate numbers of women grads — by finding ways to identify people with strong potential for programming early (like elementary school) and encouraging them to learn CS.
Considering the lengths to which they've recently gone to increase skin-level diversity, which includes literally suspending all application processes for lower level positions where the applicant is white or asian, I do have to say that I am somewhat surprised by how little they've been able to move the needle.
Kind of makes you wonder what could have nullified their efforts this heavily.
Well, I don't know anything about what Google may or may not be doing in terms of hiring, but I can tell you what would nullify pretty much any effort at achieving diversity through hiring practices: the fact that industry hiring is a zero-sum game.
The problem, at its core, is this: Every company out there is saying that it wants more diversity. Every company wants more women and minorities. Anything that Google or Apple or Facebook does is immediately cancelled out by every other company doing the same thing, and vice versa. When fewer than 20% of CS grads are women, it is physically impossible for the industry, on average, to employ an equal number of men and women. And the same statistical reality exists for minorities, for precisely the same reason. The best you can hope to do at the corporate level is to hire more women and minorities at the expense of some other company that ends up hiring fewer, hence the zero-sum-game thing.
The industry can't fix this — at least not directly. Even if they threw a pile of money at the problem, offering crazy salaries and crazy signing bonuses to women and minorities to such an extent that they all saw dollar signs at the mere mention of a tech career, it still wouldn't work. You'd get more grads, but most of the additional grads would still not be good grads, so the numbers in the industry likely would not move much, if at all.
And at some level, I think the C*Os in the industry must understand this. That's why every major tech company spends a lot of time and money trying to come up with new ways to try to improve diversity through hiring practices and trying to find ways to retain the women and minorities that they do have, in hopes that somehow they'll be able to steal more women and minorities from the competition than the competition steals from them. These are all good things to do in terms of improving working conditions for women and minorities, certainly, but short of changing the laws of physics, they can only have a very minimal effect on any given company's numbers unless every other company is asleep at the switch and doesn't follow suit. They aren't, and they do, which is why we keep having these articles asking why things aren't improving, even though the answer is obvious to anyone with even the most basic, high-school-level grasp of statistics.
There is only one way to improve diversity in computer science: get more women and minorities to major in computer science and stick with it through to graduation. How? Heck if I know.
One place to start might be creating programs in which elementary school teachers from around the country come to the Bay Area to work with industry engineers to design curricula for teaching computer science in schools, starting no later than first grade. Find ways to encourage girls and minority boys to get interested in computer science in primary school, when all the white boys (stereotyping a little here) are staying inside after school playing video games, and thus getting interested in tech. Until you do that, nothing, and I do mean nothing is going to move the diversity needle much at all.
While I agree that people need to be aware of how easily they can be manipulated and coerced, at least in the short-term, into doing things in the not-too-distant-future that are against their values now, bad science is bad science and should be exposed and retracted as soon as possible.
Except this really isn't bad science. Just some of the details of the experimental environment were omitted. There's no difference between what they're saying the professor did and what they did in the Milgram experiment, or any other similar experiment that showed a good percentage of people willingly engaging in horrible acts because of social pressure.
However, because of the omission, it doesn't quite show what is often claimed — that power inherently corrupts — but rather that people who are not used to power, who are given power, and who are then encouraged to abuse it, tend to do so. It's a subtle distinction, though, and largely empty when you really think about it, because in practice, there will always be someone or something encouraging people with power to abuse it. Expecting otherwise borders on pure fantasy.
More to the point, we've essentially seen this experiment reproduced in actual prisons without an experimenting professor encouraging the guards to be abusive. The results have still often been appalling.
So the burden of proof falls on the people making the accusation here to prove that the interference resulted in an invalid experiment from which nothing can be learned. I'm not convinced that this is the case. I'm also not convinced that it would be ethical to attempt to replicate the results, unfortunately, which makes it nontrivial to prove or disprove any such argument either way. Perhaps a similar experiment could be devised that involves an environment less extreme than prison. *shrugs*
IMO, the most practical personal vehicle would be a car first and a flying machine second. It would spend 90% of the time on the ground and 10% of time in the air, using the air as a way to fly over traffic jams and traffic lights. If you start from that assumption instead of the other way around, the power requirements become a lot more manageable.
Even for that, you're much better off hunting for a used four-core 2012 (even though at last check, they were still selling for more than the 2014 version).
The 2014 Mac Mini was a colossal failure. Between the elimination of the two-drive version and the elimination of the quad-core configuration, it went from being a great mini server to being an almost completely useless toy. I'd imagine their sales dropped commensurately, though they don't break out sales by product line enough to be certain. So I suspect it isn't getting updated because it has terrible sales, and it has terrible sales because the last upgrade was a huge downgrade.
However some people might prefer cheaper but more restricted pipes, like the "social media package + wikipedia" for $20/mo instead of $60 for totally unrestricted internet.
Have you ever tried to use social media without being able to follow off-site links? Nobody wants this except the social media sites.
"Full self driving", in this context, means an optional package that you can buy when you buy the car, but for which the software is not yet available. Part of the cost is additional hardware, so it is more expensive to add it later.
What "begin to enable full self-driving features" means is that they will begin delivering features that are only available to people who bought that optional package (which is not the same thing as actually delivering full self driving). This update could add traffic light detection, or even something as minor as speed limit sign detection (though the latter is probably planned for the lesser "enhanced autopilot" package). Or it might finally enable the smart summon feature (basically full self driving, but at 5 MPH or less).
I'd also like to note that if Opportunity wasn't designed to power down safely (or recover to a working state if someone were foolish enough to not have it power down before completely running out of power) recharge the batteries when there's enough sunlight, then have Opportunity restart, someone needs to loose their engineering degree.
The funny thing is going to be when that storm clears the panels on Spirit enough that it starts phoning home again.:-)
The headline is grossly misleading because "feature" is singular. If it had said "features", it would have been less misleading, because people would have immediately realized that there must be multiple parts to full self-driving capabilities, and they're just getting some of them. By using the singular form, the headline is saying that the car will be fully self-driving in August, which is almost certainly far from the truth.
The fact that you can figure out that the headline is pure B.S. by reading the article doesn't change the fact that the headline is clickbait.
Problem was, when that original MacBook with the single USB-C came out, there really wasn't that spectacular, to put it mildly. Now, it is quite different! There are several inexpensive (~$50) USB-C Docks that have a typical compliment of, for example, 3 or 4 USB 3.0 Ports, an SD/MiniSD (and sometimes also a CF) slot, Gigabit Ethernet, 4k HDMI out (and sometimes VGA), and occasionally even Audio I/O, plus a pass-through USB-C charger port.
But the better ones need two USB-C ports. Also, if you are giving up your only USB-C port for a dock that doesn't provide a functional USB-C port, then what's the point of having a device with USB-C ports? It's a lot more reasonable on the Pro, where you only give up half of your USB-C ports.
The biggest problem with the MacBook is that Apple completely misunderstood its target market. Students, on average, have limited desk space, so they have a laptop sitting on their desks, and their phone is plugged into the laptop to charge. With only one port, you can't do that. They designed the thing to be more like an iPad, acting like an accessory device, failing to recognize that laptops are more typically used as the central hub for your digital life.
I really don't know much about CRISPR, but I know at least one couple that was hoping for it to control the sex of their child. That couple already has a child with (pretty serious) autism and according to the dad at least, he was hoping that having a baby girl would cut the chances of having a second autistic child by a huge margin. Unfortuately, this new study doesn't sound like good news for them.
Well, then, I have good news. You can already do that with in vitro fertilization, which is commercially available and does not require editing genes at all.
USB-A is pretty darn robust on the whole, though I have bent the occasional housing by stepping on them. Chances are you just haven't used USB-C enough to have a similar number of failures.:-)
That said, I agree that micro-B was a disaster, and micro-B 3.0 doubly so, for all the reasons you mentioned. USB-C is what micro-USB and mini-USB should have been.
I'm not the one who wrote the article or asked the original question. I don't care much at all about this — certainly not enough to install Linux just to provide the statistics. And obviously, you don't either. The burden falls upon the folks who actually did the testing to produce the numbers in question.
My laptop and desktop are not ARM. One is Intel and the other is AMD. They both excel at power management as measured by battery life in the first case and measured power consumption in the second. Linux does just fine with Intel power management, thanks. Not in the least because they have a team of Linux developers working on it.
Doing "just fine" and doing as well as companies that are working directly with Intel under NDA are not necessarily the same thing. The original question was about numbers when running that version of Linux versus macOS on the same hardware. Show us the numbers. Otherwise, you're just speculating.
If your "sitting here on HS with a couple of windows using 13Gb" was correct, then none of the 8Gb machines would even boot.
Umm... there's this thing called paging that computers have been using since the 1950s. Might want to look into it. Oh, and for the past few years, Apple has been using a technique that was first pioneered in the mid-1990s called backing store compression. This is also useful for making systems boot even when there's not enough RAM.
Measuring RAM usage these days, on any OS, has been trickier for quite a while. Buffering, wired etc.
Yes, I realize that OS X aggressively uses RAM as a disk cache for performance. A little less than half of that 13 GB is file-backed, though that's somewhat misleading, as code pages are also file-backed, so we can't really tell what percentage of that is cache and what percentage consists of the app binaries themselves.
Either way, even if we assume the best case — that 100% of those file-backed pages are cache, and that code pages take no RAM at all — it would still leave 6.5 GB of physical RAM usage for a system that isn't doing much, which is just crazy.
But we definitely can't assume the best case, because I'm seeing a decent number of pageouts and a *lot* of compression. Of course, I have a flash-based disk, so the paging doesn't cause as much of a performance impact, but at least subjectively, things seem to be getting progressively slower and more bloated. Each upgrade of my HDD-based Mac Mini makes it less and less usable. At this point, it is unholy, with minor software updates taking an hour to install, and I haven't even made it to High Sierra on that one, I don't think.
One of the best seemingly minor decisions I made financially was putting a few thousand dollars into companies whose services will be needed no matter what happens to the economy, and whose growth is driven simultaneously by multiple levels of the economy — specifically, Mastercard and VISA.
I bought some MA at their IPO price of $4.435 (split adjusted, I guess) in May 2006, and it is sitting at $200, which is an annualized ROI of over 37%, with 65% growth in the past year. Over the 12 years that I've owned MA, the NASDAQ composite increased by only 247%, but MA increased by 4,057%.
In fact, the only stock I have that has done better than MA is TSLA (bought in 2010), and that's a much riskier investment. MA has consistently beaten AAPL on the average over that time period (remember that this time period includes the entirety of the existence of the iPhone), with the sole exception of a few months in 2012 where AAPL was slightly proportionately higher. But since then, MA has grown so much faster that on average, it has grown at twice the rate of AAPL. The difference is that when the economy tanks, MA hasn't typically dropped as much.
I also bought some V in 2010. It has shown an annualized ROI of about 29% since then, with 43% growth in the last year. So not quite as good, but still one of my better stocks percentage-wise.
As always, YMMV.
I am, of course, kicking myself for not putting more money into MA, but oh well. At this point, it has grown from being one of the smallest single stocks in my portfolio to being worth nearly as much as all of the others put together, if you exclude companies where I work or have previously worked, so I'm not about to increase my exposure now.
Yeah, I'm sitting here in High Sierra with only a few Safari windows open, and top is telling me I'm using 13 GB of physical RAM. And of course, the fact that you can't even get Mac laptops with more than 16 GB of physical RAM makes this all the more problematic.
And I, too, agree that 10.6.8 was the pinnacle of OS X engineering. They spent two years basically doing nothing but performance optimizations and bug fixes. The result was wildly better than what they had before. Unfortunately, in the 9 years since 10.6 came out, subsequent releases have slowly eroded those robustness and performance gains. Apple really needs to do another two-year bake cycle, doing nothing but triaging and fixing bugs. IMO, it's time.
When did battery life become an issue in performance testing?
When was it not?
For maximum performance, assuming adequate cooling, you could keep all the CPUs/cores in their highest power state and always have the maximum amount of computing power available. Your battery life will suck.
On the flip side, for maximum battery life, you could never allow your CPUs/cores to go much past the lowest non-idle power state. You'll have terrible performance, but your battery will last longer.
So to the extent that the host OS can influence the aggressiveness of CPU state changes, improving performance can significantly impact battery life and vice versa.
What the GP meant is that Apple has always been better at Intel power management. Chromebooks and Android are usually not x86, and when they are, they're typically much slower, lower-power x86 hardware than what ships in a Mac laptop. So of course they have long battery life. That goes without saying.
To use a car analogy, comparing a Chromebook to a Mac laptop is like comparing an electric golf cart to a Tesla and saying that the golf cart is as good as the Tesla at battery management because it can run for as many hours on a charge as the Tesla, ignoring that the golf cart went twenty-five miles in those five hours, and the Tesla went 300.
Slashdot Mirror
Not sure why sabotage is any more unlikely than the corporate espionage that is alleged to have already happened in the self-driving car space.
Does she charge them money to do so? Beyond other parents chipping in for gas, I mean?
Comparing the "less experienced user base" to people that a company like Google is likely to hire is misleading, because most people in that category do not have adequate skills or knowledge to get hired at any of the major tech companies. Also, you're comparing a location-neutral survey against location-specific hiring. Stack Overflow would have to break those numbers by geographical area to draw any useful comparisons, both because of the high cost of living in tech-heavy cities and because of limits on the number of people a company can bring in from other countries.
Worth noting: the minorities where the Stack Overflow survey showed the biggest gains were Asians and Middle Easterners. Asians are already overrepresented by about a factor of four at Google, assuming you're using the U.S. population as a basis. (If you're using the world population, Asians are still probably underrepresented, but there are geopolitical reasons that make such comparisons problematic). And Google doesn't even list "Middle Eastern" as an ethnicity in their report, so I can only assume that they are all lumped in under "white", making comparisons impossible.
That said, you're right that there is a big discrepancy between the percentage of blacks and Hispanics graduating with CS degrees and the percentage who are getting employed by major tech companies. I don't know what's going on there. It could be bias in hiring, or it could be that we're seeing the same effect among minorities now that we saw among whites in the 1990s, where everybody majored in CS even if they were bad at it, hoping to get a high-paying job. Realistically, it is probably some combination of both. The former problem is easy to test, though nontrivial to correct. The latter, of course, can be solved in basically the same way as we solve the problem of inadequate numbers of women grads — by finding ways to identify people with strong potential for programming early (like elementary school) and encouraging them to learn CS.
Well, I don't know anything about what Google may or may not be doing in terms of hiring, but I can tell you what would nullify pretty much any effort at achieving diversity through hiring practices: the fact that industry hiring is a zero-sum game.
The problem, at its core, is this: Every company out there is saying that it wants more diversity. Every company wants more women and minorities. Anything that Google or Apple or Facebook does is immediately cancelled out by every other company doing the same thing, and vice versa. When fewer than 20% of CS grads are women, it is physically impossible for the industry, on average, to employ an equal number of men and women. And the same statistical reality exists for minorities, for precisely the same reason. The best you can hope to do at the corporate level is to hire more women and minorities at the expense of some other company that ends up hiring fewer, hence the zero-sum-game thing.
The industry can't fix this — at least not directly. Even if they threw a pile of money at the problem, offering crazy salaries and crazy signing bonuses to women and minorities to such an extent that they all saw dollar signs at the mere mention of a tech career, it still wouldn't work. You'd get more grads, but most of the additional grads would still not be good grads, so the numbers in the industry likely would not move much, if at all.
And at some level, I think the C*Os in the industry must understand this. That's why every major tech company spends a lot of time and money trying to come up with new ways to try to improve diversity through hiring practices and trying to find ways to retain the women and minorities that they do have, in hopes that somehow they'll be able to steal more women and minorities from the competition than the competition steals from them. These are all good things to do in terms of improving working conditions for women and minorities, certainly, but short of changing the laws of physics, they can only have a very minimal effect on any given company's numbers unless every other company is asleep at the switch and doesn't follow suit. They aren't, and they do, which is why we keep having these articles asking why things aren't improving, even though the answer is obvious to anyone with even the most basic, high-school-level grasp of statistics.
There is only one way to improve diversity in computer science: get more women and minorities to major in computer science and stick with it through to graduation. How? Heck if I know.
One place to start might be creating programs in which elementary school teachers from around the country come to the Bay Area to work with industry engineers to design curricula for teaching computer science in schools, starting no later than first grade. Find ways to encourage girls and minority boys to get interested in computer science in primary school, when all the white boys (stereotyping a little here) are staying inside after school playing video games, and thus getting interested in tech. Until you do that, nothing, and I do mean nothing is going to move the diversity needle much at all.
Except this really isn't bad science. Just some of the details of the experimental environment were omitted. There's no difference between what they're saying the professor did and what they did in the Milgram experiment, or any other similar experiment that showed a good percentage of people willingly engaging in horrible acts because of social pressure.
However, because of the omission, it doesn't quite show what is often claimed — that power inherently corrupts — but rather that people who are not used to power, who are given power, and who are then encouraged to abuse it, tend to do so. It's a subtle distinction, though, and largely empty when you really think about it, because in practice, there will always be someone or something encouraging people with power to abuse it. Expecting otherwise borders on pure fantasy.
More to the point, we've essentially seen this experiment reproduced in actual prisons without an experimenting professor encouraging the guards to be abusive. The results have still often been appalling.
So the burden of proof falls on the people making the accusation here to prove that the interference resulted in an invalid experiment from which nothing can be learned. I'm not convinced that this is the case. I'm also not convinced that it would be ethical to attempt to replicate the results, unfortunately, which makes it nontrivial to prove or disprove any such argument either way. Perhaps a similar experiment could be devised that involves an environment less extreme than prison. *shrugs*
IMO, the most practical personal vehicle would be a car first and a flying machine second. It would spend 90% of the time on the ground and 10% of time in the air, using the air as a way to fly over traffic jams and traffic lights. If you start from that assumption instead of the other way around, the power requirements become a lot more manageable.
Even for that, you're much better off hunting for a used four-core 2012 (even though at last check, they were still selling for more than the 2014 version).
The 2014 Mac Mini was a colossal failure. Between the elimination of the two-drive version and the elimination of the quad-core configuration, it went from being a great mini server to being an almost completely useless toy. I'd imagine their sales dropped commensurately, though they don't break out sales by product line enough to be certain. So I suspect it isn't getting updated because it has terrible sales, and it has terrible sales because the last upgrade was a huge downgrade.
However some people might prefer cheaper but more restricted pipes, like the "social media package + wikipedia" for $20/mo instead of $60 for totally unrestricted internet.
Have you ever tried to use social media without being able to follow off-site links? Nobody wants this except the social media sites.
Alternatively, this seems like a great use for a thin client, i.e. use your laptop to VNC into a beefier computer.
Less than 1% of San Francisco residents are homeless. That's still a lot, but nowhere near 20%.
"Full self driving", in this context, means an optional package that you can buy when you buy the car, but for which the software is not yet available. Part of the cost is additional hardware, so it is more expensive to add it later.
What "begin to enable full self-driving features" means is that they will begin delivering features that are only available to people who bought that optional package (which is not the same thing as actually delivering full self driving). This update could add traffic light detection, or even something as minor as speed limit sign detection (though the latter is probably planned for the lesser "enhanced autopilot" package). Or it might finally enable the smart summon feature (basically full self driving, but at 5 MPH or less).
The funny thing is going to be when that storm clears the panels on Spirit enough that it starts phoning home again. :-)
The headline is grossly misleading because "feature" is singular. If it had said "features", it would have been less misleading, because people would have immediately realized that there must be multiple parts to full self-driving capabilities, and they're just getting some of them. By using the singular form, the headline is saying that the car will be fully self-driving in August, which is almost certainly far from the truth.
The fact that you can figure out that the headline is pure B.S. by reading the article doesn't change the fact that the headline is clickbait.
My prediction: It will (sometimes) detect traffic lights and chime at you to remind you to push the brake pedal. :-D
But the better ones need two USB-C ports. Also, if you are giving up your only USB-C port for a dock that doesn't provide a functional USB-C port, then what's the point of having a device with USB-C ports? It's a lot more reasonable on the Pro, where you only give up half of your USB-C ports.
The biggest problem with the MacBook is that Apple completely misunderstood its target market. Students, on average, have limited desk space, so they have a laptop sitting on their desks, and their phone is plugged into the laptop to charge. With only one port, you can't do that. They designed the thing to be more like an iPad, acting like an accessory device, failing to recognize that laptops are more typically used as the central hub for your digital life.
Well, then, I have good news. You can already do that with in vitro fertilization, which is commercially available and does not require editing genes at all.
USB-A is pretty darn robust on the whole, though I have bent the occasional housing by stepping on them. Chances are you just haven't used USB-C enough to have a similar number of failures. :-)
That said, I agree that micro-B was a disaster, and micro-B 3.0 doubly so, for all the reasons you mentioned. USB-C is what micro-USB and mini-USB should have been.
I'm not the one who wrote the article or asked the original question. I don't care much at all about this — certainly not enough to install Linux just to provide the statistics. And obviously, you don't either. The burden falls upon the folks who actually did the testing to produce the numbers in question.
Doing "just fine" and doing as well as companies that are working directly with Intel under NDA are not necessarily the same thing. The original question was about numbers when running that version of Linux versus macOS on the same hardware. Show us the numbers. Otherwise, you're just speculating.
Umm... there's this thing called paging that computers have been using since the 1950s. Might want to look into it. Oh, and for the past few years, Apple has been using a technique that was first pioneered in the mid-1990s called backing store compression. This is also useful for making systems boot even when there's not enough RAM.
Yes, I realize that OS X aggressively uses RAM as a disk cache for performance. A little less than half of that 13 GB is file-backed, though that's somewhat misleading, as code pages are also file-backed, so we can't really tell what percentage of that is cache and what percentage consists of the app binaries themselves.
Either way, even if we assume the best case — that 100% of those file-backed pages are cache, and that code pages take no RAM at all — it would still leave 6.5 GB of physical RAM usage for a system that isn't doing much, which is just crazy.
But we definitely can't assume the best case, because I'm seeing a decent number of pageouts and a *lot* of compression. Of course, I have a flash-based disk, so the paging doesn't cause as much of a performance impact, but at least subjectively, things seem to be getting progressively slower and more bloated. Each upgrade of my HDD-based Mac Mini makes it less and less usable. At this point, it is unholy, with minor software updates taking an hour to install, and I haven't even made it to High Sierra on that one, I don't think.
One of the best seemingly minor decisions I made financially was putting a few thousand dollars into companies whose services will be needed no matter what happens to the economy, and whose growth is driven simultaneously by multiple levels of the economy — specifically, Mastercard and VISA.
I bought some MA at their IPO price of $4.435 (split adjusted, I guess) in May 2006, and it is sitting at $200, which is an annualized ROI of over 37%, with 65% growth in the past year. Over the 12 years that I've owned MA, the NASDAQ composite increased by only 247%, but MA increased by 4,057%.
In fact, the only stock I have that has done better than MA is TSLA (bought in 2010), and that's a much riskier investment. MA has consistently beaten AAPL on the average over that time period (remember that this time period includes the entirety of the existence of the iPhone), with the sole exception of a few months in 2012 where AAPL was slightly proportionately higher. But since then, MA has grown so much faster that on average, it has grown at twice the rate of AAPL. The difference is that when the economy tanks, MA hasn't typically dropped as much.
I also bought some V in 2010. It has shown an annualized ROI of about 29% since then, with 43% growth in the last year. So not quite as good, but still one of my better stocks percentage-wise.
As always, YMMV.
I am, of course, kicking myself for not putting more money into MA, but oh well. At this point, it has grown from being one of the smallest single stocks in my portfolio to being worth nearly as much as all of the others put together, if you exclude companies where I work or have previously worked, so I'm not about to increase my exposure now.
Yeah, I'm sitting here in High Sierra with only a few Safari windows open, and top is telling me I'm using 13 GB of physical RAM. And of course, the fact that you can't even get Mac laptops with more than 16 GB of physical RAM makes this all the more problematic.
And I, too, agree that 10.6.8 was the pinnacle of OS X engineering. They spent two years basically doing nothing but performance optimizations and bug fixes. The result was wildly better than what they had before. Unfortunately, in the 9 years since 10.6 came out, subsequent releases have slowly eroded those robustness and performance gains. Apple really needs to do another two-year bake cycle, doing nothing but triaging and fixing bugs. IMO, it's time.
When was it not?
For maximum performance, assuming adequate cooling, you could keep all the CPUs/cores in their highest power state and always have the maximum amount of computing power available. Your battery life will suck.
On the flip side, for maximum battery life, you could never allow your CPUs/cores to go much past the lowest non-idle power state. You'll have terrible performance, but your battery will last longer.
So to the extent that the host OS can influence the aggressiveness of CPU state changes, improving performance can significantly impact battery life and vice versa.
What the GP meant is that Apple has always been better at Intel power management. Chromebooks and Android are usually not x86, and when they are, they're typically much slower, lower-power x86 hardware than what ships in a Mac laptop. So of course they have long battery life. That goes without saying.
To use a car analogy, comparing a Chromebook to a Mac laptop is like comparing an electric golf cart to a Tesla and saying that the golf cart is as good as the Tesla at battery management because it can run for as many hours on a charge as the Tesla, ignoring that the golf cart went twenty-five miles in those five hours, and the Tesla went 300.