Huh? 20% of what? It's a clock, and time is infinite(-ish). The clock face covers 720 minutes, and it did not advance by 20% of that (144 minutes). If this clock only ever advanced, than *maybe* you could say it was 20% of the remaining time, and always use remaining time as the scale of what's left, but this clock has moved backwards multiple times before. That means the scale goes all the way back to however far back it can go, which is entirely undefined. The clock is just a really bad representation.
doom said, "Hm perhaps we could use similar techniques to avoid making those hundreds of network connections in the first place..."... I don't really know what he's getting at here. I took it as a joke - don't get all the advertisments and crap to begin with, but maybe he meant HTTP2 style with one TCP connections and multiplexed requests. jellomizer stated that HTTP is, "Send the data and disconnect" tepples pointed out that, "HTTP has supported keep-alive and pipelining since 1.1" I noted that, "Neither of those solve the issue of checking the status of 100's of elements. They help... but can't be done completely parallel..." You said, "Doing them in parallel is bad. Doing them async concurrently is good." yada yada HTTP2.
Under HTTP2, using a classic style reload of a page, there would still be 100's of requests sent - they'd just be multiplexed over the same TCP connection. Today, with keepalives and pipelining, the browser would open several connections, using them in parallel but sending requests in each in series. While HTTP2 may be slightly faster (though it will lose out on the benefit of talking to multiple load balanced servers), it's not going to solve the 100's of requests problem either. That problem is simply that there are 100's of requests.
Neither of those solve the issue of checking the status of 100's of elements. They help, but each check still needs to be done, and they can't be done completely parallel (sane connection count limits on servers limit that, and so browsers only do N parallel requests, each keepalive'd).
They have simply decided not to check any of those other elements. Ignoring 100's of elements for a 28% speedup seems like bad math to me. They're only reloading 1 in 100's of elements, so that should be a 100x speedup (10000% speedup?)... that might be worth it, but not for just 28% faster.
I feel like I read a book and somebody ripped out the last several chapters.
Don't feel too bad. You won't find those in the article either. Here's the only sentence you missed:
The team simplified Chrome’s reload behavior and it now only validates the main resource.
IE. it used to reload 100's of page elements (eg. checking the etag of each), and now it only checks the main page.
I occasionally find I need to do a SHIFT+Reload (or CTRL+SHIFT+R) to force reload of all elements. With this new "only reload main resource" feature, I sure hope they add a something similar to reproduce the old behavior because a full reload is MUCH more overhead than the current reloads, or provide an option to use the old behavior (hahaha, yeah right! modern app *adding* a setting/option!).
It should probably be closer to a minute from midnight right now, but what's their scale?
Scale is my biggest issue with this "clock". A typical clock face represents 12 hours of time. Thus far, they have only adjusted it by minutes, starting at 7 minutes until midnight, maxing at 17, min at 2.
IMO, that means they thought the introduction of nuclear bombs changed our doomsday scale from (I assume) 720 minutes until midnight (ie. noon) to 7 minutes until midnight. Since then, we've been living in the bottom 2.3% of the scale.
Assuming I buy into that (that nukes permanently moved us that much closer to doomsday), then this latest move is a change of only 0.069% (100 * (0.5min / 720)). The largest change was in 1991 with +7minutes added, and that was only a 0.097% change on the scale.
IMO, the should have done away with the doomsday clock long ago and moved to a something almost as silly but more accurate, such as the Homeland Security Advisory System (ie. Terror Alert Level going from green - blue - yellow - orange - red). Under that system, I'd probably put us down at blue, but I suspect the same group doing the doomsday clock would abuse this one too and we would have been in red since it started. That'd be fine by me though, cause there'd be no announcements of it changing by 0.069%:-)
The same way "Defense cuts" mean the Defense Department budget would be smaller
No. What was asked was, "Would you cut departments?" Answering in the affirmative regarding the defense department would mean, "Yes, I would cut the defense department", which is very different than "defense cuts".
I do think it's obvious that he rarely speak or writes clearly. In this case, I think what he meant was that he is currently angry with something or someone EPA related, and he intended to do something to them that may be anywhere between fairly significant cuts and complete elimination of the department, but his opinion on the matter had already shifted within 2 back to back phrases, so it's anybodies guess what he'll do today.
Couple that with Trump's stated goal of eliminating the EPA entirely
Can you provide a quote for that?
Q: Would you cut departments? TRUMP: Environmental Protection, what they do is a disgrace. Every week they come out with new regulations. Q: Who's going to protect the environment? TRUMP: We'll be fine with the environment. We can leave a little bit, but you can't destroy businesses. Source: Fox News Sunday 2015 Coverage of 2016 presidential hopefuls , Oct 18, 2015
This is nice, except for docs that have parts redacted. They could sign chunks of the doc, but what chunk size? And if too small, it could be possible to determine the contents of the chunk based on other sigs (maybe). Regardless, that's not a bad idea, even if it's just used for pulling out the originals and verifying before redacting, and it'd still be good for docs that don't have anything redacted.
Sure, not all people leading these positions are experts at those fields. I'd argue they should be, but if they're competent enough at leading people that are experts, that'd probably do as well.
I'd also concede that Giuliani almost certainly didn't set up this server himself, so he's not directly to blame for that.
However, when those two are combined, it's an utter failure. He is not qualified to do the actual work, and when he has had others do the work (for an "infosec consultancy firm", no less), they utterly failed - thus his leadership of them is also an utter failure. To fill the cyber security advisor role, one should be able to either do the work directly, or be smart enough to interface with those that can do the work. As Trump would say, so sad!
Where have you been? The courts have determined an IP doesn't prove an identity, therefore it's just more anonymous information.
Look up what PII is (personally identifiable information). Your full name is PII, as is your address, but neither prove an identify. The IP does not have to prove it was you to be PII. I believe it would fall under non-sensitive, as opposed to sensitive PII such as biometric info, medical info, SSN, drivers license number, etc, but it's still PII (at least IMO, since each country/court/etc could define it however they see fit).
We already have semi-autonomous killing machines. It's not a big stretch to get to fully autonomous killing machines (tech wise, if we can have a self driving car, then we're there - though it'd be scary as hell and a bad idea to use it just yet).
If we get to autonomous ones, we can fight them back if they go haywire so long as they don't have a desire for continued existence and/or the ability to self replicate. I may be reading into it some, but I think that was GP's point. Not to get to 100% safe, because we already have killing machines, so we're already less than 100% safe, but just to get to avoid a situation where they would wipe us out.
The EU laws are similar, but different. They apply to companies with a presence in the EU, and protect the personal information of their citizens (data may not leave the country without dealing with more red tape). The RU law requires that the data be written to a server in Russia first, but the data may then leave the country - it protects their ability to access that data, rather than protecting the data. The impact is similar, but the differences are quite important. https://slashdot.org/comments....
LinkedIn could make their app still closed source and with a simple ifRussian config, tell it to pull information from a different set of servers.
(bold by me) This is the point. How are you stating it so clearly and yet still missing it? With proprietary software, the proprietor is the only one that can make that decision, and you are subject to their choices so long as you wish to use said software/service. If it were free/open/libre, anyone could decide it was worth the effort to setup and configure a Russian version that complies with their laws.
LinkedIn has decided not to comply with the Russian law at this time, so they are being blocked and everyone loses** (Russians can't use the service anymore, and no one else will see new/updated Russian data anymore).
I have 2 24" monitors, so it's not like any reasonable panel would supersede *those*
Really? If you think a 30" 2160x1600 screen is (or could be) superior to two 17" panels, then why wouldn't a 40 - 50" 4k screen be better than your 2x 24"? More square inches, (probably) way more pixels, and cheaper than the aforementioned Dell U3014
Exactly what would your non-internet-connected echo-like device that does all processing locally be capable of doing?
FWIW, I'd ideally prefer the arrangement to be a bit different, and I think my ideal setup would probably suit most of those complaining about some cloud connected always on mic. I'd prefer a local voice to text thing, a separate array mic piece, and a separate thing to process the commands and send results and keep state, and I'd want it all open source and no requirement for any external accounts. The end result would almost certainly have far more room for vulnerabilities and exploits. I'd still want it to have an internet connection so it would work with all the wonderful things on the internet. I'd still want it to be able to tie into IFTTT (if you don't know what that is, look it up - it's pretty awesome). I'd definitely want to be able to customize the action word (ie. use something other than "Alexa", "Amazon", "Echo", "Siri", "OK Google", "Cortana", etc)... which means that part couldn't use a low power hardcoded hardware bit to pick up the keyword, so it would have to listen and process all text instead. With all that stuff done locally, there's also a lot more that can break, and it'll need a lot more resources (CPU, memory, ram, and storage). I'd also want it to train based on what's been said before, which is also built into the Alexa service (you can review every clip it sent, and what it thought you said, and correct them as needed).
I don't think everyone in the world should have one, but if you're avoiding them because, "there's no way I'm putting an internet-connected microphone in my living room", that just seems silly. You'd be doing it anyway if you had anything with similar features, and your probably carrying around an internet connected mic right now, and your laptop, PC, xbox, ps4, and whatever kit you have connected to your TV probably all have one as well. If you have none of those and you're that paranoid (maybe even justifiably so), that's fine, but obviously this isn't for you then.
If you really would, "love to have something like this in my house", then using a low powered net connected device like the Echo is the smart way to go... there's no way I could justify a more-or-less fully fledged PC at every point I wanted to do that stuff. The mic alone pretty much justifies the cost, let alone the "wife factor", so to speak (and sorry to sound sexist). But hey.. if you're worried about recording devices, ignore this whole segment of tech forever. If it's voice activated, something has to be listening.
If you don't want one, that's fine, but the excessive paranoia around the Echo is just stupid. The Echo has been torn apart many times, and it barely has anything in it. It can't record excessive amounts of anything to queue up. It doesn't have the guts to do full voice to text locally. It has a small chip that is pre-programmed for specific key words (2 of them... my older Echo can use either "Alexa" or "Amazon", while my friends is either "Alexa" or "Echo").
It's pretty plain and clear what it does. It has a fifo that records a small buffer of audio, and a chip that detects the keywords. When a keyword is detected, it starts streaming the audio out (starting with that fifo, so it can catch the first utterances of your request). It gets back a textual response and speaks it, optionally asking for a follow up question.
That may be more than enough for some people to rule it out as an option in their home. That's fine and even somewhat understandable. There's no need to exaggerate the situation further.
If someone wants to record all you say, you're probably carrying around a far more advanced and more powerful computer with TONS more sensors with multiple audio and video sources, accelerometer, GPS, beromiter, finger print reader, etc... all in your pocket, or next to you on the desk/table/etc. Where's the hard off switch for those sensors? The Echo is the least of your worries.
So if in the code I say "if ( seconds assumes that the value...
Kinda amusing that your post is an example of unexpected (though well known) data causing an incorrect outcome. IE. slashdot ate part of your comment (I'm hoping that assumption is correct. Otherwise, your brain ate part of it). Sadly, we have to manually escape < (ie: <) and friends here (and I have no idea what all must be escaped).
Read the article then. It shows it pretty plainly: https://blog.cloudflare.com/ho... I was going to try to guess what they were doing, but they have some actual code snippets.
AFAICT, a unit test wouldn't have caught this either (unless they planned for this sort of error, in which case the code wouldn't have been broken either). From TFA:
RRDNS doesn’t just keep a single measurement for each resolver, it takes many measurements and smoothes them. So, the single measurement wouldn’t cause RRDNS to think the resolver was working in negative time, but after a few measurements the smoothed value would eventually become negative.
So, a unit test with one negative example (which may have been difficult to mimic anyway, due to the direct usage of Time.Now()) probably wouldn't have triggered the issue on its own.
IMHO, blaming a misconception of time always going forward is just convenient here. The fix was changing this bit: if rttMax == 0 {
rttMax = DefaultTimeout }
They just changed "==" to "<=". There was no reason not to have it as "<=" to begin with, even if one ignores where rttMax comes from. Any time I check if something is == to something else, and I don't have else conditions covering the other cases, I ask myself what should happen in those other else cases and ensure I'm covered. That may still have caused it to break, but it could have done: if rttMax == 0 {
rttMax = DefaultTimeout } else if rttMax < 0 {
panic("What the fuck happened to rttMax to make it negative!?!") } ...though it probably would have been better to just log that somewhere and set it to the DefaultTimeout.
Anyway, I think it's a great example of a one character bug that only triggers on very obscure events under significant load.
That's a terrible analogy. When you go to a restaurant you only "own" the food on your plate and your body gets rid of it a few hours later.
I think you're on to a better analogy. For TV, you "own" the show content until you're done processing it (ie. you rent it, or have a temporary license to it). So, where's the TV/software/restaurant fall in there? * food = content/shows/etc * salt/pepper = ff/pause/rewind/mute * restaurant = network (abc/nbc/etc) * chef/waiters/staff = content producers (director/actor/etc) * tv = plate/fork/knife/etc * tv OS = Miss Manners rules for using your utensils and how to properly consume your food
If you eat at a restaurant, it's then like going out to the movie theater. If you order delivery (and thus supply your own dishes and your own house rules for the handling thereof), it's like watching the content on your TV. As such, they have no business enforcing Judith Martin's rules while in the privacy of your own home.
If you went to their site, they already know your IP address. They let you type in other IP addresses in addition to your own, which every server knows whenever a client connects (barring tor or proxy usage).
This is a double blessing then. I could never wrap my brain around the name cyanogenmod, could never remember it, had few friends that ever even heard of it, etc. People would just say (incorrectly), "root your phone", rather than something like, "install cyanogenmod or another alternative OS". LineageOS sounds like a great name.
The one thing that bothers me with this drive is that it requires a somewhat large amount of energy to work,... plus they add a lot of weight to the probes, which would nullify the EM drive's point to some extent.
I don't see the connection to those points from TFS. This drive seems plenty interesting, but from a power to weight ratio, it sucks. Loads of energy in, and micronewtons of thrust out? It's still very interesting, but I don't see how this looks viable for small satellites. Hopefully I'm overlooking something obvious.
Why in the world are they saying this is, "a software category it essentially gave birth to"? There were loads of others with far more features for years before git, let alone github. It's like saying facebook gave birth to social networks.
Yes, you can. Put in their info, and have them bring their ID.
I can't let someone use my season pass ticket if I can't go or don't want to go?
Completely optional. Season passes could easily be exempted (they don't sell an unlimited number of those). If the pass is something re-usable, then you wouldn't be selling it to just anyone anyway, so that's not a problem.
You're saying that an event ticket isn't really my property but only a license to a particular seat?
Are you saying you think you own that seat? Of course not. The ticket grants YOU access (or could, under that setup).
If they want to allow and facilitate resale, as they do now, then they should simple auction them off. If they want to prevent resale, they can easily do so without this stupid bot law. I don't really care which way it's done; I don't think this has any place as a law.
Slashdot Mirror
Huh? 20% of what?
It's a clock, and time is infinite(-ish). The clock face covers 720 minutes, and it did not advance by 20% of that (144 minutes).
If this clock only ever advanced, than *maybe* you could say it was 20% of the remaining time, and always use remaining time as the scale of what's left, but this clock has moved backwards multiple times before. That means the scale goes all the way back to however far back it can go, which is entirely undefined. The clock is just a really bad representation.
This thread keeps flying off on tangents!
doom said, "Hm perhaps we could use similar techniques to avoid making those hundreds of network connections in the first place..." ... I don't really know what he's getting at here. I took it as a joke - don't get all the advertisments and crap to begin with, but maybe he meant HTTP2 style with one TCP connections and multiplexed requests. ... but can't be done completely parallel ..."
jellomizer stated that HTTP is, "Send the data and disconnect"
tepples pointed out that, "HTTP has supported keep-alive and pipelining since 1.1"
I noted that, "Neither of those solve the issue of checking the status of 100's of elements. They help
You said, "Doing them in parallel is bad. Doing them async concurrently is good." yada yada HTTP2.
Under HTTP2, using a classic style reload of a page, there would still be 100's of requests sent - they'd just be multiplexed over the same TCP connection.
Today, with keepalives and pipelining, the browser would open several connections, using them in parallel but sending requests in each in series. While HTTP2 may be slightly faster (though it will lose out on the benefit of talking to multiple load balanced servers), it's not going to solve the 100's of requests problem either. That problem is simply that there are 100's of requests.
Neither of those solve the issue of checking the status of 100's of elements. They help, but each check still needs to be done, and they can't be done completely parallel (sane connection count limits on servers limit that, and so browsers only do N parallel requests, each keepalive'd).
They have simply decided not to check any of those other elements. Ignoring 100's of elements for a 28% speedup seems like bad math to me. They're only reloading 1 in 100's of elements, so that should be a 100x speedup (10000% speedup?)... that might be worth it, but not for just 28% faster.
I feel like I read a book and somebody ripped out the last several chapters.
Don't feel too bad. You won't find those in the article either. Here's the only sentence you missed:
The team simplified Chrome’s reload behavior and it now only validates the main resource.
IE. it used to reload 100's of page elements (eg. checking the etag of each), and now it only checks the main page.
I occasionally find I need to do a SHIFT+Reload (or CTRL+SHIFT+R) to force reload of all elements. With this new "only reload main resource" feature, I sure hope they add a something similar to reproduce the old behavior because a full reload is MUCH more overhead than the current reloads, or provide an option to use the old behavior (hahaha, yeah right! modern app *adding* a setting/option!).
It should probably be closer to a minute from midnight right now, but what's their scale?
Scale is my biggest issue with this "clock". A typical clock face represents 12 hours of time. Thus far, they have only adjusted it by minutes, starting at 7 minutes until midnight, maxing at 17, min at 2.
IMO, that means they thought the introduction of nuclear bombs changed our doomsday scale from (I assume) 720 minutes until midnight (ie. noon) to 7 minutes until midnight. Since then, we've been living in the bottom 2.3% of the scale.
Assuming I buy into that (that nukes permanently moved us that much closer to doomsday), then this latest move is a change of only 0.069% (100 * (0.5min / 720)). The largest change was in 1991 with +7minutes added, and that was only a 0.097% change on the scale.
IMO, the should have done away with the doomsday clock long ago and moved to a something almost as silly but more accurate, such as the Homeland Security Advisory System (ie. Terror Alert Level going from green - blue - yellow - orange - red). Under that system, I'd probably put us down at blue, but I suspect the same group doing the doomsday clock would abuse this one too and we would have been in red since it started. That'd be fine by me though, cause there'd be no announcements of it changing by 0.069% :-)
The same way "Defense cuts" mean the Defense Department budget would be smaller
No.
What was asked was, "Would you cut departments?"
Answering in the affirmative regarding the defense department would mean, "Yes, I would cut the defense department", which is very different than "defense cuts".
I do think it's obvious that he rarely speak or writes clearly. In this case, I think what he meant was that he is currently angry with something or someone EPA related, and he intended to do something to them that may be anywhere between fairly significant cuts and complete elimination of the department, but his opinion on the matter had already shifted within 2 back to back phrases, so it's anybodies guess what he'll do today.
AFAICT, the Wi-Fi assistant is only available on android 5.1 nexus devices (ex. galaxy S6 has no support for it).
Couple that with Trump's stated goal of eliminating the EPA entirely
Can you provide a quote for that?
Q: Would you cut departments?
TRUMP: Environmental Protection, what they do is a disgrace. Every week they come out with new regulations.
Q: Who's going to protect the environment?
TRUMP: We'll be fine with the environment. We can leave a little bit, but you can't destroy businesses.
Source: Fox News Sunday 2015 Coverage of 2016 presidential hopefuls , Oct 18, 2015
This is nice, except for docs that have parts redacted. They could sign chunks of the doc, but what chunk size? And if too small, it could be possible to determine the contents of the chunk based on other sigs (maybe).
Regardless, that's not a bad idea, even if it's just used for pulling out the originals and verifying before redacting, and it'd still be good for docs that don't have anything redacted.
Agreed, and I'd take it several steps further...
Sure, not all people leading these positions are experts at those fields. I'd argue they should be, but if they're competent enough at leading people that are experts, that'd probably do as well.
I'd also concede that Giuliani almost certainly didn't set up this server himself, so he's not directly to blame for that.
However, when those two are combined, it's an utter failure. He is not qualified to do the actual work, and when he has had others do the work (for an "infosec consultancy firm", no less), they utterly failed - thus his leadership of them is also an utter failure. To fill the cyber security advisor role, one should be able to either do the work directly, or be smart enough to interface with those that can do the work. As Trump would say, so sad!
Where have you been? The courts have determined an IP doesn't prove an identity, therefore it's just more anonymous information.
Look up what PII is (personally identifiable information). Your full name is PII, as is your address, but neither prove an identify. The IP does not have to prove it was you to be PII. I believe it would fall under non-sensitive, as opposed to sensitive PII such as biometric info, medical info, SSN, drivers license number, etc, but it's still PII (at least IMO, since each country/court/etc could define it however they see fit).
We already have semi-autonomous killing machines. It's not a big stretch to get to fully autonomous killing machines (tech wise, if we can have a self driving car, then we're there - though it'd be scary as hell and a bad idea to use it just yet).
If we get to autonomous ones, we can fight them back if they go haywire so long as they don't have a desire for continued existence and/or the ability to self replicate. I may be reading into it some, but I think that was GP's point. Not to get to 100% safe, because we already have killing machines, so we're already less than 100% safe, but just to get to avoid a situation where they would wipe us out.
The EU laws are similar, but different. They apply to companies with a presence in the EU, and protect the personal information of their citizens (data may not leave the country without dealing with more red tape). The RU law requires that the data be written to a server in Russia first, but the data may then leave the country - it protects their ability to access that data, rather than protecting the data. The impact is similar, but the differences are quite important.
https://slashdot.org/comments....
LinkedIn could make their app still closed source and with a simple ifRussian config, tell it to pull information from a different set of servers.
(bold by me)
This is the point. How are you stating it so clearly and yet still missing it?
With proprietary software, the proprietor is the only one that can make that decision, and you are subject to their choices so long as you wish to use said software/service. If it were free/open/libre, anyone could decide it was worth the effort to setup and configure a Russian version that complies with their laws.
LinkedIn has decided not to comply with the Russian law at this time, so they are being blocked and everyone loses** (Russians can't use the service anymore, and no one else will see new/updated Russian data anymore).
** or wins, if you happen to dislike LinkedIn
I have 2 24" monitors, so it's not like any reasonable panel would supersede *those*
Really? If you think a 30" 2160x1600 screen is (or could be) superior to two 17" panels, then why wouldn't a 40 - 50" 4k screen be better than your 2x 24"? More square inches, (probably) way more pixels, and cheaper than the aforementioned Dell U3014
Exactly what would your non-internet-connected echo-like device that does all processing locally be capable of doing?
FWIW, I'd ideally prefer the arrangement to be a bit different, and I think my ideal setup would probably suit most of those complaining about some cloud connected always on mic. I'd prefer a local voice to text thing, a separate array mic piece, and a separate thing to process the commands and send results and keep state, and I'd want it all open source and no requirement for any external accounts. The end result would almost certainly have far more room for vulnerabilities and exploits. I'd still want it to have an internet connection so it would work with all the wonderful things on the internet. I'd still want it to be able to tie into IFTTT (if you don't know what that is, look it up - it's pretty awesome). I'd definitely want to be able to customize the action word (ie. use something other than "Alexa", "Amazon", "Echo", "Siri", "OK Google", "Cortana", etc)... which means that part couldn't use a low power hardcoded hardware bit to pick up the keyword, so it would have to listen and process all text instead. With all that stuff done locally, there's also a lot more that can break, and it'll need a lot more resources (CPU, memory, ram, and storage). I'd also want it to train based on what's been said before, which is also built into the Alexa service (you can review every clip it sent, and what it thought you said, and correct them as needed).
I don't think everyone in the world should have one, but if you're avoiding them because, "there's no way I'm putting an internet-connected microphone in my living room", that just seems silly. You'd be doing it anyway if you had anything with similar features, and your probably carrying around an internet connected mic right now, and your laptop, PC, xbox, ps4, and whatever kit you have connected to your TV probably all have one as well. If you have none of those and you're that paranoid (maybe even justifiably so), that's fine, but obviously this isn't for you then.
If you really would, "love to have something like this in my house", then using a low powered net connected device like the Echo is the smart way to go... there's no way I could justify a more-or-less fully fledged PC at every point I wanted to do that stuff. The mic alone pretty much justifies the cost, let alone the "wife factor", so to speak (and sorry to sound sexist). But hey.. if you're worried about recording devices, ignore this whole segment of tech forever. If it's voice activated, something has to be listening.
If you don't want one, that's fine, but the excessive paranoia around the Echo is just stupid.
The Echo has been torn apart many times, and it barely has anything in it. It can't record excessive amounts of anything to queue up. It doesn't have the guts to do full voice to text locally. It has a small chip that is pre-programmed for specific key words (2 of them... my older Echo can use either "Alexa" or "Amazon", while my friends is either "Alexa" or "Echo").
It's pretty plain and clear what it does. It has a fifo that records a small buffer of audio, and a chip that detects the keywords. When a keyword is detected, it starts streaming the audio out (starting with that fifo, so it can catch the first utterances of your request). It gets back a textual response and speaks it, optionally asking for a follow up question.
That may be more than enough for some people to rule it out as an option in their home. That's fine and even somewhat understandable. There's no need to exaggerate the situation further.
If someone wants to record all you say, you're probably carrying around a far more advanced and more powerful computer with TONS more sensors with multiple audio and video sources, accelerometer, GPS, beromiter, finger print reader, etc... all in your pocket, or next to you on the desk/table/etc. Where's the hard off switch for those sensors? The Echo is the least of your worries.
So if in the code I say "if ( seconds assumes that the value ...
Kinda amusing that your post is an example of unexpected (though well known) data causing an incorrect outcome. IE. slashdot ate part of your comment (I'm hoping that assumption is correct. Otherwise, your brain ate part of it). Sadly, we have to manually escape < (ie: <) and friends here (and I have no idea what all must be escaped).
Read the article then. It shows it pretty plainly: https://blog.cloudflare.com/ho...
I was going to try to guess what they were doing, but they have some actual code snippets.
AFAICT, a unit test wouldn't have caught this either (unless they planned for this sort of error, in which case the code wouldn't have been broken either). From TFA:
RRDNS doesn’t just keep a single measurement for each resolver, it takes many measurements and smoothes them. So, the single measurement wouldn’t cause RRDNS to think the resolver was working in negative time, but after a few measurements the smoothed value would eventually become negative.
So, a unit test with one negative example (which may have been difficult to mimic anyway, due to the direct usage of Time.Now()) probably wouldn't have triggered the issue on its own.
IMHO, blaming a misconception of time always going forward is just convenient here. The fix was changing this bit:
...though it probably would have been better to just log that somewhere and set it to the DefaultTimeout.
if rttMax == 0 {
rttMax = DefaultTimeout
}
They just changed "==" to "<=". There was no reason not to have it as "<=" to begin with, even if one ignores where rttMax comes from. Any time I check if something is == to something else, and I don't have else conditions covering the other cases, I ask myself what should happen in those other else cases and ensure I'm covered. That may still have caused it to break, but it could have done:
if rttMax == 0 {
rttMax = DefaultTimeout
} else if rttMax < 0 {
panic("What the fuck happened to rttMax to make it negative!?!")
}
Anyway, I think it's a great example of a one character bug that only triggers on very obscure events under significant load.
That's a terrible analogy. When you go to a restaurant you only "own" the food on your plate and your body gets rid of it a few hours later.
I think you're on to a better analogy.
For TV, you "own" the show content until you're done processing it (ie. you rent it, or have a temporary license to it). So, where's the TV/software/restaurant fall in there?
* food = content/shows/etc
* salt/pepper = ff/pause/rewind/mute
* restaurant = network (abc/nbc/etc)
* chef/waiters/staff = content producers (director/actor/etc)
* tv = plate/fork/knife/etc
* tv OS = Miss Manners rules for using your utensils and how to properly consume your food
If you eat at a restaurant, it's then like going out to the movie theater.
If you order delivery (and thus supply your own dishes and your own house rules for the handling thereof), it's like watching the content on your TV.
As such, they have no business enforcing Judith Martin's rules while in the privacy of your own home.
If you went to their site, they already know your IP address. They let you type in other IP addresses in addition to your own, which every server knows whenever a client connects (barring tor or proxy usage).
This is a double blessing then. I could never wrap my brain around the name cyanogenmod, could never remember it, had few friends that ever even heard of it, etc. People would just say (incorrectly), "root your phone", rather than something like, "install cyanogenmod or another alternative OS". LineageOS sounds like a great name.
The one thing that bothers me with this drive is that it requires a somewhat large amount of energy to work, ... plus they add a lot of weight to the probes, which would nullify the EM drive's point to some extent.
I don't see the connection to those points from TFS. This drive seems plenty interesting, but from a power to weight ratio, it sucks. Loads of energy in, and micronewtons of thrust out? It's still very interesting, but I don't see how this looks viable for small satellites. Hopefully I'm overlooking something obvious.
Why in the world are they saying this is, "a software category it essentially gave birth to"? There were loads of others with far more features for years before git, let alone github. It's like saying facebook gave birth to social networks.
I can't give tickets as gifts?
Yes, you can. Put in their info, and have them bring their ID.
I can't let someone use my season pass ticket if I can't go or don't want to go?
Completely optional. Season passes could easily be exempted (they don't sell an unlimited number of those). If the pass is something re-usable, then you wouldn't be selling it to just anyone anyway, so that's not a problem.
You're saying that an event ticket isn't really my property but only a license to a particular seat?
Are you saying you think you own that seat? Of course not. The ticket grants YOU access (or could, under that setup).
If they want to allow and facilitate resale, as they do now, then they should simple auction them off. If they want to prevent resale, they can easily do so without this stupid bot law. I don't really care which way it's done; I don't think this has any place as a law.