And phone companies are often subsidized by government,
Not just subsidized - all of them receive their corporate charter from government. In telecommunications, they rely on government-created and enforced rights on parts of the electromagnetic spectrum. Verizon is just as much beholden to the US government as the East India Company was to the UK - I don't think you can really separate them.
Ahh, a retreat to identity politics. This is what I meant when I said Democrats still haven't learned. I don' t suspect I'll get anywhere, but I'll try anyway... if you find yourself in a minority group and your goal is to have someone who represents you in office, you need to create a coalition with an inclusive message. This should be a very natural way for liberals to operate, and historically Democrats did very well with it. Or, you can chase short-term advantage with a message of fear and division - something Republicans have historically done well with... though at the expense of long-term membership decline as demographics and sentiment shift.
Democrats have recently latched on to this poisonous Republican strategy, and it is killing them. I can think of no worse tactical mistake then getting everyone to circle their wagons and vote with their "tribe". White people currently vote all over the map because they aren't a single identity. If you force an identity on them, they will be the largest single ethnic group long into the foreseeable future. This is a bad strategy, and it's bad for society. The end-goal is (or at least, should be IMHO) a society where everyone enjoys equal treatment, both by government and by society. Stratifying us into tribes will not accomplish this, and will only cement "white privilege" (another terrible term for converting people to your side).
Thanks for pointing that out - I followed the link to the abstract and then downloaded the paper, and you are correct. The Register article is misleading... lesson learned.
And when you hire "coders", you're going to direct them towards an Excel solution? Why on earth would you do that when a "coder" could use almost anything?
Where does it say that? I just re-read it and I'm pretty sure you made that up. There is a mention of copy and paste also contributing at the bottom, but that's it.
If you want to change the rules, then change the rules and the programmers will make the appropriate changes. Don't blame a robot for taking the laws literally. For this particular robot, they'll change the rules for horn blowing. The trucker is liable, just like if he hit any other immobile object. A robot truck probably would not have backed into the bus.
Once more of these idiot vehicles are released, driving may become more dangerous for everyone.
Oh, please. Pretty soon we'll have data and this kind of absurd statement will die down. Autonomous vehicles will become just another road feature - and a predictable one at that - that people navigate around.
If a person drives 10 MPH on a freeway and people get in accidents while avoiding the slow vehicle, who has really caused it?
If that's your concern, change the rules to prohibit going 10MPH on the freeway. They are just computers - they'll follow the rules.
And yes, to answer your question - if you hit someone going 10 MPH on the freeway, you were either not paying attention or you were driving recklessly. Yes, they created a hazard, but hazards are part of driving - it could just as easily have been an emergency vehicle, a tire tread, or a mattress. The autonomous cars will not hit the hazard.
And the only way to push a change back to a repository you don't control! You fork, push your change to your fork, then create a pull request. This is by design - I have no idea why this is in any way a surprise.
Please... humans crash into each other several times a minute. The automatic bus in question was back on the road the next day and there have been no further incidents. Perfection is not the bar here, human drivers are the bar to meet or exceed. It's not as high a bar as you are implying.
In NYC, they pay the "conductor" (the guy who closes the doors to separate you from your family) more than a rookie cop. Turn on the automatic doors and put a rookie cop on each train. Problem solved.
That's misleading. A human-driven truck ran into a stopped automated bus at creeping speed. No injuries and there was a human "attendant" on board, which obviously did not help since the bus already detected the hazard and stopped.
That's my wife's workflow. I think it's silly to scrub every dish and then load the dishwasher, so I put them in fairly filthy and only scrub the occasional one that comes out still soiled. She complains about my method because she doesn't like scrubbing "clean" dishes. Oh, well - married life... at least I put the seat down and fold the laundry fresh out of the dryer.
Incidentally, a dishwasher is a dishwashing robot. I guess people don't want to put the clean dishes away? That seems like a more trivial task for a robot to be trained.
Data sources don't fail silently (unless they open the sheet "wrong", like from a network drive and the security settings aren't right). Custom functions fail silently. Once a function fails, all the calculations that rely on that result also fail and your spreadsheet is in an undefined state.
Oh, and once you buy them, there's a quick-and-dirty trick you can use to check for flicker: put your smartphone camera into continuous mode (on Apple I think you just hold down the shutter button) and take a screenful of photos. When you review the thumbnails all on one screen, it will be very clear if the bulb is flickering on and off or if it is steady.
You can find online reviews of many bulbs, but the strategy I've been (mostly) successful with is to look at the claimed lifetime on the name-brand bulbs. They generally fall into two buckets, 10 and 20 year. The 10 year bulbs are clearly over-driven and lots of corners cut, the 20-year are generally properly designed and more conservative. But honestly, the flicker doesn't bother me so I only notice it in places with running water. My $60 wet-rated recessed shower LED flickers, for instance - so cost clearly isn't a guide. On the other hand, the filament bulbs that I use for decorative purposes (mostly outside, but also in one bathroom) flicker for sure (and I one of them apart to find no electrolytic capacitor at all!), but I only notice it in the bathroom because it freeze-frames your pee and the kids picked up on that.
Again, not all LEDs do this - just LEDs with crappy driver circuitry. LEDs are DC, so they don't flicker at all unless you do so deliberately. I have some of these flickery LEDs in my bathroom, because the "filament" bulbs look nice since the driver fits into the little candelabra base. But the penalty I pay is a roughly 60Hz flicker that is quite pronounced, as all they do is rectify the AC and then have some kind of a LED driver - no room for a capacitor at all. If your bulb has room for a full bridge rectifier, decent capacitor, and a constant current driver you shouldn't have any measurable "flicker" no matter how nice an instrument you have (obviously you'd be able to see some slight periodic variation, just not "flicker").
using those as sources for Excel sheets with live data
That's true, but if you are just using Access to feed Excel, there are easier, cleaner, faster server-side ways to do that. I just saw little value in going with Access vs. other options, once I got to know it. I have a bad taste in my mouth from using multiple Office products strung together in a workflow. Too much of the MS stuff fails silently to aid newbies, and eventually you end up accidentally using old or half-refreshed data.
This comment perfectly illustrates why it is very difficult to train computers to parse natural language. If some humans can't recognize that the context has shifted away from Excel and towards spreadsheets in general, how can you train a computer to do it?
I thought Access seemed like a nice tool. Then I took a course and really learned it and realized that most of the things it could do were better done elsewhere. Oh, well - I try to target those brain cells when drinking now.
It's definitely a workflow problem, but I suspect one of the problems is that Excel is in the middle of the workflow. It's fine for light-duty data exploration, prototyping, one-offs... things like that. But if you find that you are using Excel as glue, you are probably Doing It Wrong(tm):)
Slashdot Mirror
And phone companies are often subsidized by government,
Not just subsidized - all of them receive their corporate charter from government. In telecommunications, they rely on government-created and enforced rights on parts of the electromagnetic spectrum. Verizon is just as much beholden to the US government as the East India Company was to the UK - I don't think you can really separate them.
Ahh, a retreat to identity politics. This is what I meant when I said Democrats still haven't learned. I don' t suspect I'll get anywhere, but I'll try anyway... if you find yourself in a minority group and your goal is to have someone who represents you in office, you need to create a coalition with an inclusive message. This should be a very natural way for liberals to operate, and historically Democrats did very well with it. Or, you can chase short-term advantage with a message of fear and division - something Republicans have historically done well with... though at the expense of long-term membership decline as demographics and sentiment shift.
Democrats have recently latched on to this poisonous Republican strategy, and it is killing them. I can think of no worse tactical mistake then getting everyone to circle their wagons and vote with their "tribe". White people currently vote all over the map because they aren't a single identity. If you force an identity on them, they will be the largest single ethnic group long into the foreseeable future. This is a bad strategy, and it's bad for society. The end-goal is (or at least, should be IMHO) a society where everyone enjoys equal treatment, both by government and by society. Stratifying us into tribes will not accomplish this, and will only cement "white privilege" (another terrible term for converting people to your side).
Just not the people that elected the Republicans. It's OK, the Republicans are just as bad when the Democrats win.
It's amazing what winning all three branches of government can get you. Hopefully the Democrats learn from this, but so far they haven't.
I'd like that stolen battery life back please
OK, just pay more for the extra development time, or use a version with fewer features. Every decision has trade-offs.
The Czech Republic, EU and NATO member, has to pick a side? This will be a very difficult decision.
Thanks for pointing that out - I followed the link to the abstract and then downloaded the paper, and you are correct. The Register article is misleading... lesson learned.
And when you hire "coders", you're going to direct them towards an Excel solution? Why on earth would you do that when a "coder" could use almost anything?
Where does it say that? I just re-read it and I'm pretty sure you made that up. There is a mention of copy and paste also contributing at the bottom, but that's it.
If you want to change the rules, then change the rules and the programmers will make the appropriate changes. Don't blame a robot for taking the laws literally. For this particular robot, they'll change the rules for horn blowing. The trucker is liable, just like if he hit any other immobile object. A robot truck probably would not have backed into the bus.
Once more of these idiot vehicles are released, driving may become more dangerous for everyone.
Oh, please. Pretty soon we'll have data and this kind of absurd statement will die down. Autonomous vehicles will become just another road feature - and a predictable one at that - that people navigate around.
If a person drives 10 MPH on a freeway and people get in accidents while avoiding the slow vehicle, who has really caused it?
If that's your concern, change the rules to prohibit going 10MPH on the freeway. They are just computers - they'll follow the rules.
And yes, to answer your question - if you hit someone going 10 MPH on the freeway, you were either not paying attention or you were driving recklessly. Yes, they created a hazard, but hazards are part of driving - it could just as easily have been an emergency vehicle, a tire tread, or a mattress. The autonomous cars will not hit the hazard.
And the only way to push a change back to a repository you don't control! You fork, push your change to your fork, then create a pull request. This is by design - I have no idea why this is in any way a surprise.
Please... humans crash into each other several times a minute. The automatic bus in question was back on the road the next day and there have been no further incidents. Perfection is not the bar here, human drivers are the bar to meet or exceed. It's not as high a bar as you are implying.
In NYC, they pay the "conductor" (the guy who closes the doors to separate you from your family) more than a rookie cop. Turn on the automatic doors and put a rookie cop on each train. Problem solved.
That's misleading. A human-driven truck ran into a stopped automated bus at creeping speed. No injuries and there was a human "attendant" on board, which obviously did not help since the bus already detected the hazard and stopped.
That's my wife's workflow. I think it's silly to scrub every dish and then load the dishwasher, so I put them in fairly filthy and only scrub the occasional one that comes out still soiled. She complains about my method because she doesn't like scrubbing "clean" dishes. Oh, well - married life... at least I put the seat down and fold the laundry fresh out of the dryer.
Incidentally, a dishwasher is a dishwashing robot. I guess people don't want to put the clean dishes away? That seems like a more trivial task for a robot to be trained.
Data sources don't fail silently (unless they open the sheet "wrong", like from a network drive and the security settings aren't right). Custom functions fail silently. Once a function fails, all the calculations that rely on that result also fail and your spreadsheet is in an undefined state.
Oh, and once you buy them, there's a quick-and-dirty trick you can use to check for flicker: put your smartphone camera into continuous mode (on Apple I think you just hold down the shutter button) and take a screenful of photos. When you review the thumbnails all on one screen, it will be very clear if the bulb is flickering on and off or if it is steady.
You can find online reviews of many bulbs, but the strategy I've been (mostly) successful with is to look at the claimed lifetime on the name-brand bulbs. They generally fall into two buckets, 10 and 20 year. The 10 year bulbs are clearly over-driven and lots of corners cut, the 20-year are generally properly designed and more conservative. But honestly, the flicker doesn't bother me so I only notice it in places with running water. My $60 wet-rated recessed shower LED flickers, for instance - so cost clearly isn't a guide. On the other hand, the filament bulbs that I use for decorative purposes (mostly outside, but also in one bathroom) flicker for sure (and I one of them apart to find no electrolytic capacitor at all!), but I only notice it in the bathroom because it freeze-frames your pee and the kids picked up on that.
Again, not all LEDs do this - just LEDs with crappy driver circuitry. LEDs are DC, so they don't flicker at all unless you do so deliberately. I have some of these flickery LEDs in my bathroom, because the "filament" bulbs look nice since the driver fits into the little candelabra base. But the penalty I pay is a roughly 60Hz flicker that is quite pronounced, as all they do is rectify the AC and then have some kind of a LED driver - no room for a capacitor at all. If your bulb has room for a full bridge rectifier, decent capacitor, and a constant current driver you shouldn't have any measurable "flicker" no matter how nice an instrument you have (obviously you'd be able to see some slight periodic variation, just not "flicker").
Dimming is another whole kettle of fish.
CFLs also flicker - that's how they work. LEDs at least can be made steady if you use a good one with a decent capacitor.
The sodium lights are all steadily being replaced by LED.
using those as sources for Excel sheets with live data
That's true, but if you are just using Access to feed Excel, there are easier, cleaner, faster server-side ways to do that. I just saw little value in going with Access vs. other options, once I got to know it. I have a bad taste in my mouth from using multiple Office products strung together in a workflow. Too much of the MS stuff fails silently to aid newbies, and eventually you end up accidentally using old or half-refreshed data.
This comment perfectly illustrates why it is very difficult to train computers to parse natural language. If some humans can't recognize that the context has shifted away from Excel and towards spreadsheets in general, how can you train a computer to do it?
I thought Access seemed like a nice tool. Then I took a course and really learned it and realized that most of the things it could do were better done elsewhere. Oh, well - I try to target those brain cells when drinking now.
It's definitely a workflow problem, but I suspect one of the problems is that Excel is in the middle of the workflow. It's fine for light-duty data exploration, prototyping, one-offs... things like that. But if you find that you are using Excel as glue, you are probably Doing It Wrong(tm) :)