Unless the problem isn't that the batteries fail under high stress -- it could be that a very small fraction of the batteries fail under normal use. I suspect that this is much nastier to test for...
Oh and Apple, you can keep CUPS. It is a lousy technology that hardly compensates the community for everything else you took to build X.
Speak for yourself -- I routinely have an easier time setting up printers on my Linux box using CUPS than my Windows friends do. And, as much as I think editing raw text files is great from flexibility/power standpoint, the ol' localhost:631 interface is very nice in my experience.
If it's all on *NIX, I know this would be pretty easy to do (probably not difficult with Windows, I just don't know how). Just have the VM, upon booting up, download an image of itself and spool up a new instance. Would perhaps need to take care that the memory and storage allocated for each additional VM is somewhat smaller than the host VM, but this is easily scriptable.
It was a pretty basic setup. My University has some deal with MS so each student has access to one Windows 10 (and Office) key, with a download link via Microsoft. I just downloaded and set up a vanilla VirtualBox Windows 10 guest. (I believe I just installed VirtualBox from the Debian repositories, nothing fancy.)
It's not perfect -- audio can be choppy at times, and the 3D acceleration doesn't work (at least on my setup), and the seamless mode doesn't really work (so I just run it fullscreen on one of my workspaces). But performance is completely acceptable for Word, and I have it using only one core and 2GB RAM.
I have my Windows 10 install in a VM (Linux host), and I'm very pleased with the setup (a multi-monitor setup is helpful but not necessary). Performance is adequate for the basics, although it's a little silly that I need to run a VM just to be able to properly edit a document. Yes, I know there are open alternatives, but for documents with images, formulae, tracked changes, etc., I find it easier to just use Word (obviously I prefer a TeX solution [I highly recommend Overleaf], but it's not always up to me...).
True, but I think you're being a little too literal.
Yes, everyone experiences everything differently. But that doesn't mean there aren't common features/tasks/workflows -- dare I say, experiences -- that affect many people.
We've all been in terrible airports, and we've all been in airports that were actually pretty smooth; the "user experience" didn't magically become lousy in one and great in another, it was deliberately (not) thought out. That's not to say that we both agree 100% on which are the good airports and which are the lousy ones, but I'm guessing there's some overlap. It's similar with software -- we've all used programs that had about zero discoverability and were a pain to use, and we've (hopefully) all experienced software which was actually pretty well thought out from a user experience perspective.
How a user experiences the result is something that the user does indeed contribute to the interaction -- but they're not the sole contributor.
Not to mention the fact that, AFAIK, there's no real ban on cellphones vis-a-vis data usage/spectrum usage in theaters -- there's only a ban on making noise/light. There shouldn't be a problem downloading Wikipedia in the background during a movie, so long as your phone is on silent.
...and seriously, $200? And I thought the in-flight wifi was expensive...
Sure, but it seems to be like ISPs are trying to have their cake and eat it too with data caps.
To use the power analogy, bandwidth ~ power capacity, data ~ energy use (or similarly with water, gas,...). Now, it's true that a basic home electrical hookup has a "bandwidth" limit -- but this should easily be enough for everyone in the family to have their own computer or TV running, in addition to a fridge, washing machine, dryer, etc. So ideally, the "bandwidth" isn't a problem (brown-outs notwithstanding), and you pay for the energy used (data).
Contrast this to ISPs, where the norm has been to charge for bandwith. Now, bandwidth has historically not been something that we "have enough of" -- a starter package probably doesn't allow everyone in a family to do "normal internet things" at a respectable rate (download something in the background, watch HD or UHD streams, etc.). So, not only are we supposed to pay extra for acceptable bandwidth, but we are also supposed to pay for each bit?
It strikes me as double-dipping: either charge for the bits but give us "unlimited" bandwidth (fast enough for everyone in the house to simultaneously use it at adequate speeds), or give us the bits for free and charge us for the bandwidth. I certainly prefer the latter, but perhaps only because that's what I'm used to.
Would it be legal for Samsung to issue an OTA update to essentially brick the device (ideally affecting the charging controller, too)?
Would this be legal? Not that I'm advocating that sort of behavior, just wondering...as-is, it seems we barely own anything and are just borrowing it from the company...
Although UCSF may not have a robust athletics department, I believe (?) that this decision comes from the University of California system -- not just the UCSF campus.
This layoff may have huge implications. That's because the university's IT services agreement with HCL can be leveraged by any institution in the 10-campus University of California system, which serves some 240,000 students and employs some 190,000 faculty and staff.
I think it's a reflection on the current fixed-line internet options available to some people. I'd bet that if these people had the option of reasonably priced ~100Mbps symmetric connections, they wouldn't consider using cell as their primary ISP.
It sounds like you're talking about shot noise. But there's also read noise, etc., which may not be (afaik...) spatially correlated.
Now, if Canon properly implements some sort of binning, they could get the best of both worlds: high resolution when read noise isn't a problem (i.e., bright scene) and good SNR in dim scenes, albeit at lower resolution.
Except that (as pointed out elsewhere) this means that car passengers, public transit riders, etc., are all unfairly locked out.
Out of an abundance of caution, yes, that could work -- I guess the question is, should Nintendo be responsible for everything, or should the user? If the former, then is it also the company's responsibility to, say, mandate breaks every X minutes (for ergonomic and/or eyesight reasons)? It presumably also knows (or can know) when your alarms are set, so should it shut off in time to allow a full 8 hours of sleep (as driving under sleep deprivation can be just as dangerous as driving drunk)?
Something tells me that if you already have a car with built-in nav and OTA update ability, it already has the means to track you.
Besides, if you travel with your cell phone then you can easily be tracked by the cell network (albeit with perhaps less accuracy than GPS). And I'm sure there are various apps -- in particular map apps -- which track you (I believe Google's traffic data comes in part from this?).
If you want to be un-tracked, I suspect the only decent way of doing it would be to turn off your phone and drive a car with zero networking abilities. And even this can be defeated with a camera, OCR, and a license-plate registration database...
I had a lot of trouble with garbled SD cards until I switched to a different power adapter. The newish, 2A Samsung supply I had ended up causing stability issues, whereas my old flip-phone micro-USB power supply works just fine. (Although before I figured that out, my initial workaround was to use an NFS root filesystem, which does have its advantages I guess.)
But the technology and standards could be useful for things other than just raw 8k video. For instance, one could imagine some "user-pannable" (or VR) applications where you're streaming sporting events with lots going on. Perhaps you're only displaying 1920x1080 on your TV, but you could pan your view smoothly over a grid 4 times the width and height of your TV. And for the 22.2 channel audio, yeah, that does seem a little ridiculous (though it would make sense to me to just encode it in a spatially-resolved way, which is iirc what Dolby Atmos does).
That said, yeah, I agree that it's overkill for just watching TV.
Right. Comcast (allegedly) rips off the consumers and the state fills its pockets with another 100 million dollars of taxpayers funds paid via Comcast. That's just fucking fantastic.
From TFA:
According to KOMO News, the lawsuit is seeking more than $73 million in restitution to pay back Service Protection Plan subscriber payments; full restitution for all service calls that applied an improper resolution code, estimated to be at least $1 million...
I could be reading it incorrectly, but it seems that (if the state wins) >$73M+$1M goes directly to the customers.
IANAL, but I would guess (???) that if the money *didn't* go back to the subscribers/taxpayers, then they could turn around and file a class action (though again, my understanding is that the subscribers will be getting a fair chunk of the $100M if they win).
Sure, but if you're trying to do, say, some GPU computing with a limited budget, it might (?) be wise to shop around and perhaps settle for multiple budget cards as opposed to one mid/high-end card. Taking the numbers at face value, the RX470 claims to offer 4.9 TFLOPs for $150, whereas the GTX1080 is something like 9 TFLOPS for ~$600 (I'm assuming these are comparable floating point tests, but perhaps they're not...).
Yes, I'd rather an Nvidia card that offers good performance and "just works" under Linux -- but everyone has their own requirements.
Slashdot Mirror
Unless the problem isn't that the batteries fail under high stress -- it could be that a very small fraction of the batteries fail under normal use. I suspect that this is much nastier to test for...
Oh and Apple, you can keep CUPS. It is a lousy technology that hardly compensates the community for everything else you took to build X.
Speak for yourself -- I routinely have an easier time setting up printers on my Linux box using CUPS than my Windows friends do. And, as much as I think editing raw text files is great from flexibility/power standpoint, the ol' localhost:631 interface is very nice in my experience.
...starring Jamie Dornan ('Fifty Shades of Grey')...
So to answer your question...the buggy whip, uh, finds a way.
If it's all on *NIX, I know this would be pretty easy to do (probably not difficult with Windows, I just don't know how). Just have the VM, upon booting up, download an image of itself and spool up a new instance. Would perhaps need to take care that the memory and storage allocated for each additional VM is somewhat smaller than the host VM, but this is easily scriptable.
It was a pretty basic setup. My University has some deal with MS so each student has access to one Windows 10 (and Office) key, with a download link via Microsoft. I just downloaded and set up a vanilla VirtualBox Windows 10 guest. (I believe I just installed VirtualBox from the Debian repositories, nothing fancy.)
It's not perfect -- audio can be choppy at times, and the 3D acceleration doesn't work (at least on my setup), and the seamless mode doesn't really work (so I just run it fullscreen on one of my workspaces). But performance is completely acceptable for Word, and I have it using only one core and 2GB RAM.
I have my Windows 10 install in a VM (Linux host), and I'm very pleased with the setup (a multi-monitor setup is helpful but not necessary). Performance is adequate for the basics, although it's a little silly that I need to run a VM just to be able to properly edit a document. Yes, I know there are open alternatives, but for documents with images, formulae, tracked changes, etc., I find it easier to just use Word (obviously I prefer a TeX solution [I highly recommend Overleaf], but it's not always up to me...).
True, but I think you're being a little too literal.
Yes, everyone experiences everything differently. But that doesn't mean there aren't common features/tasks/workflows -- dare I say, experiences -- that affect many people.
We've all been in terrible airports, and we've all been in airports that were actually pretty smooth; the "user experience" didn't magically become lousy in one and great in another, it was deliberately (not) thought out. That's not to say that we both agree 100% on which are the good airports and which are the lousy ones, but I'm guessing there's some overlap. It's similar with software -- we've all used programs that had about zero discoverability and were a pain to use, and we've (hopefully) all experienced software which was actually pretty well thought out from a user experience perspective.
How a user experiences the result is something that the user does indeed contribute to the interaction -- but they're not the sole contributor.
Not to mention the fact that, AFAIK, there's no real ban on cellphones vis-a-vis data usage/spectrum usage in theaters -- there's only a ban on making noise/light. There shouldn't be a problem downloading Wikipedia in the background during a movie, so long as your phone is on silent.
...and seriously, $200? And I thought the in-flight wifi was expensive...
Sure, but it seems to be like ISPs are trying to have their cake and eat it too with data caps.
...). Now, it's true that a basic home electrical hookup has a "bandwidth" limit -- but this should easily be enough for everyone in the family to have their own computer or TV running, in addition to a fridge, washing machine, dryer, etc. So ideally, the "bandwidth" isn't a problem (brown-outs notwithstanding), and you pay for the energy used (data).
To use the power analogy, bandwidth ~ power capacity, data ~ energy use (or similarly with water, gas,
Contrast this to ISPs, where the norm has been to charge for bandwith. Now, bandwidth has historically not been something that we "have enough of" -- a starter package probably doesn't allow everyone in a family to do "normal internet things" at a respectable rate (download something in the background, watch HD or UHD streams, etc.). So, not only are we supposed to pay extra for acceptable bandwidth, but we are also supposed to pay for each bit?
It strikes me as double-dipping: either charge for the bits but give us "unlimited" bandwidth (fast enough for everyone in the house to simultaneously use it at adequate speeds), or give us the bits for free and charge us for the bandwidth. I certainly prefer the latter, but perhaps only because that's what I'm used to.
Would it be legal for Samsung to issue an OTA update to essentially brick the device (ideally affecting the charging controller, too)?
Would this be legal? Not that I'm advocating that sort of behavior, just wondering...as-is, it seems we barely own anything and are just borrowing it from the company...
According to this article:
This layoff may have huge implications. That's because the university's IT services agreement with HCL can be leveraged by any institution in the 10-campus University of California system, which serves some 240,000 students and employs some 190,000 faculty and staff.
I think it's a reflection on the current fixed-line internet options available to some people. I'd bet that if these people had the option of reasonably priced ~100Mbps symmetric connections, they wouldn't consider using cell as their primary ISP.
It sounds like you're talking about shot noise. But there's also read noise, etc., which may not be (afaik...) spatially correlated.
Now, if Canon properly implements some sort of binning, they could get the best of both worlds: high resolution when read noise isn't a problem (i.e., bright scene) and good SNR in dim scenes, albeit at lower resolution.
Except that (as pointed out elsewhere) this means that car passengers, public transit riders, etc., are all unfairly locked out.
Out of an abundance of caution, yes, that could work -- I guess the question is, should Nintendo be responsible for everything, or should the user? If the former, then is it also the company's responsibility to, say, mandate breaks every X minutes (for ergonomic and/or eyesight reasons)? It presumably also knows (or can know) when your alarms are set, so should it shut off in time to allow a full 8 hours of sleep (as driving under sleep deprivation can be just as dangerous as driving drunk)?
(Here's the patch from a more familiar source, kernel.org.)
The link was from here, which also suggests a fix for unpatched systems:
echo 'net.ipv4.tcp_challenge_ack_limit = 999999999' >>/etc/sysctl.conf;sysctl -p
(Courtesy of this site.)
Something tells me that if you already have a car with built-in nav and OTA update ability, it already has the means to track you.
Besides, if you travel with your cell phone then you can easily be tracked by the cell network (albeit with perhaps less accuracy than GPS). And I'm sure there are various apps -- in particular map apps -- which track you (I believe Google's traffic data comes in part from this?).
If you want to be un-tracked, I suspect the only decent way of doing it would be to turn off your phone and drive a car with zero networking abilities. And even this can be defeated with a camera, OCR, and a license-plate registration database...
On the other hand, you could make a slick implementation of this.
I had a lot of trouble with garbled SD cards until I switched to a different power adapter. The newish, 2A Samsung supply I had ended up causing stability issues, whereas my old flip-phone micro-USB power supply works just fine. (Although before I figured that out, my initial workaround was to use an NFS root filesystem, which does have its advantages I guess.)
Ah, you're right.
...they use helium in the drives, so all your music sounds like Alvin and the Chipmunks.
But the technology and standards could be useful for things other than just raw 8k video. For instance, one could imagine some "user-pannable" (or VR) applications where you're streaming sporting events with lots going on. Perhaps you're only displaying 1920x1080 on your TV, but you could pan your view smoothly over a grid 4 times the width and height of your TV. And for the 22.2 channel audio, yeah, that does seem a little ridiculous (though it would make sense to me to just encode it in a spatially-resolved way, which is iirc what Dolby Atmos does).
That said, yeah, I agree that it's overkill for just watching TV.
Right. Comcast (allegedly) rips off the consumers and the state fills its pockets with another 100 million dollars of taxpayers funds paid via Comcast. That's just fucking fantastic.
From TFA:
According to KOMO News, the lawsuit is seeking more than $73 million in restitution to pay back Service Protection Plan subscriber payments; full restitution for all service calls that applied an improper resolution code, estimated to be at least $1 million...
I could be reading it incorrectly, but it seems that (if the state wins) >$73M+$1M goes directly to the customers.
IANAL, but I would guess (???) that if the money *didn't* go back to the subscribers/taxpayers, then they could turn around and file a class action (though again, my understanding is that the subscribers will be getting a fair chunk of the $100M if they win).
Nice to see some real money for these violations. $100M certainly wont bankrupt Comcast, but hopefully it will make them sit up and take notice.
Sure, but if you're trying to do, say, some GPU computing with a limited budget, it might (?) be wise to shop around and perhaps settle for multiple budget cards as opposed to one mid/high-end card. Taking the numbers at face value, the RX470 claims to offer 4.9 TFLOPs for $150, whereas the GTX1080 is something like 9 TFLOPS for ~$600 (I'm assuming these are comparable floating point tests, but perhaps they're not...).
Yes, I'd rather an Nvidia card that offers good performance and "just works" under Linux -- but everyone has their own requirements.
In a nonlinear world, I might be the OP of this thread!