My guess is that they will go back further than they need to in order to cover all their products under warranty. Anything beyond that is them just being nice.
Well, except that in the US Government, the Judicial branch is over the Executive branch, not the other way around. Federal judges don't answer to trump. Their appointments are for life. Trump's term is only 4 or 8 years.
Unfortunately, neither can I. Either it is someone who finally is saying how they really feel, or someone making fun of how some really feel. In either case, I see it as someone's "truth", whether it be the poster's or the anti-trump crowd.
Then again, I remember working 60 hours straight as well. Prepping for a major deployment during regular work hours, spending all night trying to deploy it, then at the last possible moment, rolling it all back again. Spending the next day debugging what went wrong, fixing it, prepping for another deployment attempt, spending all night re-deploying it, and then spending the following day monitoring the system and ensuring there were no issues and finally going home exhausted. I doubt I could still pull that off today.
Wow. "Hundreds" are not in most common usage the equivalent number to 50,000. Once "hundreds" become "thousands" or "tens of thousands" usually it is reported as such instead of 500 "hundreds".
Or we can wait for 499 more articles just like this one, then they will be equivalent.
Well you have me beat by about 5 years if you consider programming "for fun" -- started in 1978 myself. I didn't actually start getting paid for programming until 1981 or so, however, and I was in Junior High school at the time as well.
That is a good approach to take for some projects. Unfortunately, not others. Like a certain company I know who has 20+ clients, all running what should be the same code, but it's 20 different COPIES of code in various stages of development. Try getting that under control, lol.
The code base isn't very TDD friendly. Like there are routines that are 50,000+ lines long, which basically does everything. And there are 20+ copies of that routine, that are all slightly different, most of the time for no reason other than 3 different people wrote the same logic 3 different ways -- all of which are terrible... 20-30 lines of code that is hard to read with edge case bugs, and can be rewritten in 1-3 easy to read line with no edge case bugs, but trying to find out if those edge cases are really bugs or just bad code is nerve wracking. One out of 20 times, yes, there really was a edge case that needed that not so easy to see side effect. The entire code base is that way. So we're refactoring what we can, modularizing it, and then breaking those modules out into standard libraries and then re-integrating those libraries into each of the 20 clients. Long, and painful.
It is a LOT more painful than just starting over, and then attempting to migrate all the clients to the new code base one at a time. Starting with a code base that is TDD/Unit Testing friendly. However, that is a 6-9 month project before you see any tangible results vs the long and painful way which will likely take us the next 5-6 years and when we are done it won't be nearly as nice. But... managers and owners like to see stuff NOW. Even if it's crap. I just hope that support isn't dropped from our old frameworks before we get a chance to get off them.
I suspect there's been 15-20 years of programmers telling the higher ups that they need to rewrite this stuff and the bosses saying they can't afford it/don't have to time to do it/won't hire the additional people to do it. All the while creating new non-standard tweaks to the system that has turned it into a big hairy ball of mess.
Sucks to be them now.
Source: I also work in IT, and deal with financial clients (Most of the big names).
While I can attest to just how bad seagate's DM-series drives actually are (I had a dozen or so 3TB DM-series myself, and they typically died in less than a year). That said, the DM-series isn't meant to be used in that way. They aren't designed to put in a box connector facing down, with little heat dissipation and with high vibration. It does appear the 8TB DM model actually holds up really well, even in this environment . The NM-series (Enterprise class) did fair above average. I just wish that backblaze would have a bigger sampling of their actual NAS-class drives (VN-series).
I suggest you go back and re-read the entire spec, because it completely supports the Anonymous Cowards' perspective.
It goes into very very specific details that not only suggest he is correct, but actually says this it even suggests doing so "In particular, even user agents that do implement the suggested default rendering are encouraged to offer settings that override this default to improve the experience for the user, e.g., changing the color contrast, using different focus styles, or otherwise making the experience more accessible and usable to the user."
Sorry, but the spec is almost screaming to please do this. LazyLoading (intelligently) will improve the experience for the user.
Problem solved. Prohibit car dealers from selling cars, period. You order the car online, and you can have it delivered to your house (for fee), or to a dealership (for a slightly higher fee).
On the other end, none of my media players (Xbox One, Nvidia Shield, Chromecast, AppleTV, etc) support it or my 4k projector (All support MPEG2/h.264/HEVC).
HEVC is already entrenched in the UHD standard, so that isn't going away any time soon. So unless they can get the content creators (TV Shows and movies) to switch, then AV-1 will be a "web-only" technology unfortunately or we will have to wait for a UHD 2.0 standard.
I think GPU manufacturers might drop MPEG support soon.
Oh hell no. The CPU isn't powerful enough. The difference between CPU and GPU encode/decode is my CPU (3930k I believe - 6 core/12 thread) struggling to keep up with 1 stream (and sucking down a ton of power) and having no issues streaming 4 streams, and only streaming 1 or 2 using next to no power.
Backdoors are bad for everyone. I recommend building a front door instead. Faster to get in and out too since you don't have to walk around the building.
I like rotten tomatos, but I ignore the tomatometer and look at the audience score. I find it much more accurate to whether I will enjoy the movie or not. Then again, I'm looking to be entertained, not oogle over how the director is using cutting edge lighting techniques to a boring story.
I think the disconnect here is you think that multicast packets are just simple routes and they aren't. There are many types of multicast packets, and you are probably thinking of the 224.x.x.x types, which isn't what most people use, and specifically google chromecast uses the DIAL protocol in the 239.x.x.x range (239.255.255.250 for server location discovery).
Many multicast groups function under a subscribe/unsubscribe type model, and the router has to be able to track who is subscribed to what multicast group (Bad if your router has very little memory and your network tries to spin up a ton of logical mutlicast groups...). It also means that routing for those packets won't be handled in the hardware and instead CPU routed, and quite frankly most home routers just can't handle routing a large percentage of packets through the CPU. You are quite mistaken in thinking that the CPUs in most routers are capable of "easily" routing a few hundred mbps of tiny packets. Try it, even with the latest and greatest home routers available and you'll quickly find they will get overwhelmed.
As an example, my current home router is a Asus AC-RT88U (Dual core 1.4Ghz CPU), which when I bought it was about the fastest you could find in a home router at any price. Turn on some QoS functions and it'll quickly die because the packets are being routed through the CPU. It also has 8 ports, but logically it's two 4-ports bridged via software, and if you have a large amount of data that has to go from one of the 4-port switches to the other... Again, it'll die.
And before you go on about how that router/brand sucks, I can't point at a shelf with all the big name routers on it (netgear, asus, linksys, etc) of all makes and models running stock and non-stock (tomato, dd-wrt, etc).. All of which perform worse.
So, as I stated before... Home routers just weren't designed to handle wirespeed traffic of unusual types whether it be large amounts of open TCP connections through NAT (this is what killed most routers in peer-to-peer) or having to do most of the routing through the CPU. They handle the traffic patterns that 99% of households require, and if you step outside of that, they fall flat -- quickly.
Yes. Windows 10 Pro has the following options for managing updates: Change active hours -> You can select 18 hours per day in which updates will not happen. Restart Options (different from applying the update, this affects when it will restart the machine to finish applying the update if it needs to) -> Schedule time pick any time in the next x days (set in advanced options) in which the PC will restart to apply the update if you don't manually restart before then. Restart Options -> Show more notifications (on/off) -> if you want to see more or less notifications about restarting. Advanced -> Pick how many days you can delay the update for feature updates from 0 to 365 days. Advanced -> Pick how many days you can delay the update if it contains security updates from 0 to 30 days. Advanced -> Pause all updates for up to 35 days (on/off) Advanced -> Auto download updates on metered connections (on/off) Advanced -> Delivery Optimization (many options in here) Advanced -> Privacy settings (Here you can turn off the telemetry stuff)
#2 is a bigger problem, but realistically home routers were never designed to do those types of things. They aren't capable of doing full wirespeed packet inspection, and for 99+% of the homes this works fine at a fraction of the cost.
The obvious solution if you need to do wire speed packet inspections, have huge routing tables, and rock solid stability, then buy a router designed for business use. Unfortunately those are MUCH more expensive.
Slashdot Mirror
My guess is that they will go back further than they need to in order to cover all their products under warranty. Anything beyond that is them just being nice.
Well, except that in the US Government, the Judicial branch is over the Executive branch, not the other way around. Federal judges don't answer to trump. Their appointments are for life. Trump's term is only 4 or 8 years.
Unfortunately, neither can I. Either it is someone who finally is saying how they really feel, or someone making fun of how some really feel. In either case, I see it as someone's "truth", whether it be the poster's or the anti-trump crowd.
Then again, I remember working 60 hours straight as well. Prepping for a major deployment during regular work hours, spending all night trying to deploy it, then at the last possible moment, rolling it all back again. Spending the next day debugging what went wrong, fixing it, prepping for another deployment attempt, spending all night re-deploying it, and then spending the following day monitoring the system and ensuring there were no issues and finally going home exhausted. I doubt I could still pull that off today.
Wow. "Hundreds" are not in most common usage the equivalent number to 50,000. Once "hundreds" become "thousands" or "tens of thousands" usually it is reported as such instead of 500 "hundreds".
Or we can wait for 499 more articles just like this one, then they will be equivalent.
If the greybeards wrote the code for slashdot, it would support unicode.
Well you have me beat by about 5 years if you consider programming "for fun" -- started in 1978 myself. I didn't actually start getting paid for programming until 1981 or so, however, and I was in Junior High school at the time as well.
That is a good approach to take for some projects. Unfortunately, not others. Like a certain company I know who has 20+ clients, all running what should be the same code, but it's 20 different COPIES of code in various stages of development. Try getting that under control, lol.
The code base isn't very TDD friendly. Like there are routines that are 50,000+ lines long, which basically does everything. And there are 20+ copies of that routine, that are all slightly different, most of the time for no reason other than 3 different people wrote the same logic 3 different ways -- all of which are terrible... 20-30 lines of code that is hard to read with edge case bugs, and can be rewritten in 1-3 easy to read line with no edge case bugs, but trying to find out if those edge cases are really bugs or just bad code is nerve wracking. One out of 20 times, yes, there really was a edge case that needed that not so easy to see side effect. The entire code base is that way. So we're refactoring what we can, modularizing it, and then breaking those modules out into standard libraries and then re-integrating those libraries into each of the 20 clients. Long, and painful.
It is a LOT more painful than just starting over, and then attempting to migrate all the clients to the new code base one at a time. Starting with a code base that is TDD/Unit Testing friendly. However, that is a 6-9 month project before you see any tangible results vs the long and painful way which will likely take us the next 5-6 years and when we are done it won't be nearly as nice. But... managers and owners like to see stuff NOW. Even if it's crap. I just hope that support isn't dropped from our old frameworks before we get a chance to get off them.
I figured it was appropriate as long as we were on the topic of telling others what they should do with their assets, it was fair game.
I suspect there's been 15-20 years of programmers telling the higher ups that they need to rewrite this stuff and the bosses saying they can't afford it/don't have to time to do it/won't hire the additional people to do it. All the while creating new non-standard tweaks to the system that has turned it into a big hairy ball of mess.
Sucks to be them now.
Source: I also work in IT, and deal with financial clients (Most of the big names).
Feel free to donate $245 million of your own money to charity if you feel so strongly.
While I can attest to just how bad seagate's DM-series drives actually are (I had a dozen or so 3TB DM-series myself, and they typically died in less than a year). That said, the DM-series isn't meant to be used in that way. They aren't designed to put in a box connector facing down, with little heat dissipation and with high vibration. It does appear the 8TB DM model actually holds up really well, even in this environment
.
The NM-series (Enterprise class) did fair above average. I just wish that backblaze would have a bigger sampling of their actual NAS-class drives (VN-series).
Yes, like the destination charge you already pay -- and you currently even pay it if you pick it up at the factory.
Except irregardless is actually a word, and been in use since 1795. Please check your dictionary.
I suggest you go back and re-read the entire spec, because it completely supports the Anonymous Cowards' perspective.
It goes into very very specific details that not only suggest he is correct, but actually says this it even suggests doing so "In particular, even user agents that do implement the suggested default rendering are encouraged to offer settings that override this default to improve the experience for the user, e.g., changing the color contrast, using different focus styles, or otherwise making the experience more accessible and usable to the user."
Sorry, but the spec is almost screaming to please do this. LazyLoading (intelligently) will improve the experience for the user.
Problem solved. Prohibit car dealers from selling cars, period. You order the car online, and you can have it delivered to your house (for fee), or to a dealership (for a slightly higher fee).
On the other end, none of my media players (Xbox One, Nvidia Shield, Chromecast, AppleTV, etc) support it or my 4k projector (All support MPEG2/h.264/HEVC).
HEVC is already entrenched in the UHD standard, so that isn't going away any time soon. So unless they can get the content creators (TV Shows and movies) to switch, then AV-1 will be a "web-only" technology unfortunately or we will have to wait for a UHD 2.0 standard.
I think GPU manufacturers might drop MPEG support soon.
Oh hell no. The CPU isn't powerful enough. The difference between CPU and GPU encode/decode is my CPU (3930k I believe - 6 core/12 thread) struggling to keep up with 1 stream (and sucking down a ton of power) and having no issues streaming 4 streams, and only streaming 1 or 2 using next to no power.
Backdoors are bad for everyone. I recommend building a front door instead. Faster to get in and out too since you don't have to walk around the building.
I like rotten tomatos, but I ignore the tomatometer and look at the audience score. I find it much more accurate to whether I will enjoy the movie or not. Then again, I'm looking to be entertained, not oogle over how the director is using cutting edge lighting techniques to a boring story.
Sexist much? Sometimes it is 'that woman' too. And sometimes it is 'that guy who dresses like a woman'.
by "supported by all browsers" you mean "supported only by chrome and chrome derivitives (opera, samsung)". No other browsers support it.
I think the disconnect here is you think that multicast packets are just simple routes and they aren't. There are many types of multicast packets, and you are probably thinking of the 224.x.x.x types, which isn't what most people use, and specifically google chromecast uses the DIAL protocol in the 239.x.x.x range (239.255.255.250 for server location discovery).
Many multicast groups function under a subscribe/unsubscribe type model, and the router has to be able to track who is subscribed to what multicast group (Bad if your router has very little memory and your network tries to spin up a ton of logical mutlicast groups...). It also means that routing for those packets won't be handled in the hardware and instead CPU routed, and quite frankly most home routers just can't handle routing a large percentage of packets through the CPU. You are quite mistaken in thinking that the CPUs in most routers are capable of "easily" routing a few hundred mbps of tiny packets. Try it, even with the latest and greatest home routers available and you'll quickly find they will get overwhelmed.
As an example, my current home router is a Asus AC-RT88U (Dual core 1.4Ghz CPU), which when I bought it was about the fastest you could find in a home router at any price. Turn on some QoS functions and it'll quickly die because the packets are being routed through the CPU. It also has 8 ports, but logically it's two 4-ports bridged via software, and if you have a large amount of data that has to go from one of the 4-port switches to the other... Again, it'll die.
And before you go on about how that router/brand sucks, I can't point at a shelf with all the big name routers on it (netgear, asus, linksys, etc) of all makes and models running stock and non-stock (tomato, dd-wrt, etc).. All of which perform worse.
So, as I stated before... Home routers just weren't designed to handle wirespeed traffic of unusual types whether it be large amounts of open TCP connections through NAT (this is what killed most routers in peer-to-peer) or having to do most of the routing through the CPU. They handle the traffic patterns that 99% of households require, and if you step outside of that, they fall flat -- quickly.
Yes. Windows 10 Pro has the following options for managing updates:
Change active hours -> You can select 18 hours per day in which updates will not happen.
Restart Options (different from applying the update, this affects when it will restart the machine to finish applying the update if it needs to) -> Schedule time pick any time in the next x days (set in advanced options) in which the PC will restart to apply the update if you don't manually restart before then.
Restart Options -> Show more notifications (on/off) -> if you want to see more or less notifications about restarting.
Advanced -> Pick how many days you can delay the update for feature updates from 0 to 365 days.
Advanced -> Pick how many days you can delay the update if it contains security updates from 0 to 30 days.
Advanced -> Pause all updates for up to 35 days (on/off)
Advanced -> Auto download updates on metered connections (on/off)
Advanced -> Delivery Optimization (many options in here)
Advanced -> Privacy settings (Here you can turn off the telemetry stuff)
That's pretty comprehensive.
#2 is a bigger problem, but realistically home routers were never designed to do those types of things. They aren't capable of doing full wirespeed packet inspection, and for 99+% of the homes this works fine at a fraction of the cost.
The obvious solution if you need to do wire speed packet inspections, have huge routing tables, and rock solid stability, then buy a router designed for business use. Unfortunately those are MUCH more expensive.