I'm sure it's hard to pin down the cost of a z13 mainframe, but from what I can tell a high end one is in the several million range at *least*.
Say $5M for one, that runs 5000 really anemic VMs (like 1/10 core each, so they are useless for real load). You can get literally thousands of servers for that price, or more realistically a reasonable number of servers that will still be able to provide multiple cores per VM, making them actually useful under load.
Seems like the only advantage of the mainframe is it's highly fault tolerant. ie. solve the problem with hardware and money instead of good software design. That works to a degree, but it's still vertical scaling. Most Internet companies have long abandoned that model for good software architecture and horizontal scaling - which if done right provides much better fault tolerance, of course.
Well then how is that at all relevant or useful? I bet you can run 2000 VM instances on a high end commodity Linux server if you don't want to them to *do* anything.
The newer mainframes can run 2,000+ instances of Linux.
Citation?
The most I have seen is is on the z13 which has ~140 cores - you can run a bunch of separate instances of Linux but it's nowhere NEAR the same performance as running Linux on fairly cheap dedicated "standard" servers. Not to say there aren't reasons to use a mainframe - just that maximizing virtual Linux instances is rarely one of them.
Wha? It's either using a neural network or it isn't. Whether it's a neural network or not doing the computing has nothing to do with the results or the efficiency of the search vs more traditional computing methods.
And I wasn't commenting on the methods or even the effectiveness of the program, just the (probably clueless non-tech) editor's assertion that it applied to "almost any image".
I'm not sure the author was using the phrase in a mathematical sense. More likely it was in the probabilistic sense of a very high percentage of successes in a number of finite trials.
Not really... the author was mostly using it in the Internet clickbait sense.
If you try it out, it's basically guessing the location of a Google street view image on Google maps. Hardly "determining the location of any image".
Still, it is a lot of fun. Somehow I scored > 10000 points on the "Paris" map even though I was there my first (and only) time last year. Guess it helps to have walked the shit out of a city...
That may have been true in the distant past, but the median net worth of a US senator is now $2.9M.
Ironically the "poorest" member of Congress may be Rep. Alcee Hastings, who is millions of dollars in debt to law firms from fighting a bribery charge back when he was a Federal judge (also amazing is that he was impeached and removed of his judgeship by Congress, than later elected to it!)
His H1B reform bill wants to set a minimum salary on H1B's to make sure it's used to fill vacancies and not cut costs, as well as outlawing "disparagement clauses" in severance agreements that prevent employees from saying anything bad about their ex-employer.
I am pro H1B in general, and the majority of companies in the Bay Area are not using it to replace US workers (there are hundreds of thousands of openings companies are trying to fill, and the starting salaries of many of those are easily in the 6 figures). But clearly there are some companies in the US that abuse the program and there are things that can be done to fix that without throwing it out altogether.
Trump is on record as being... for Planned Parenthood.
This statement is totally incorrect. Here's an exact quote from Trump in the debate on 2/26:
"So you can say whatever you want, but they have millions of women going through Planned Parenthood that are helped greatly. And I wouldn’t fund it. I would defund it because of the abortion factor, which they say is 3 percent. I don’t know what percentage it is. They say it’s 3 percent."
He freely admits that 97%+ of the services are not related to abortion, it HELPS WOMEN GREATLY, and still is ok with throwing it under the bus. CERTAINLY not "for" PP. And certainly not even for women's rights and health in general.
The difference is Pol Pot, in addition to being a genocidal maniac, was actually very well educated himself...
Whereas Donald Trump probably thinks Khmer Rouge is a type of makeup. ("I know Khmer Rouge, I love Khmer Rouge. They sponsored several of my beauty pageants. My wife Melania is a loyal customer.")
Obviously. I know the history as well as anyone. But ARPANET != Internet. The whole point of the Internet is that every node is globally addressable from any other node. That wasn't true until TCP/IP was made the (sole) standard.
The "Internet" was not created in 1969. By most counts you could probably trace it to 1983, when TCP/IP was officially made the protocol for all routing on ARPANET.
And seriously, calling the Internet dead in 1993? If that's true, why are you still posting on/. instead of living offline in your cave?
It is a major release. Some breakage is expected. As long as you stay on the same major it should be stable.
Hah. Good luck with that. FFmpeg is an amazing collection of codecs wrapped in a horrible and ever-changing API. I have used it on a number of projects and it seems every time I upgrade something breaks.
That would be true if the HA market wasn't already like 30+ years old. The actual "players" and innovators already have plenty of patents or prior art.
There are some of these "Alliances" that actually build momentum and make sense (see: UHD Alliance). This does not look like one of those.
So, I suppose I should take the high road and ignore your assholish tone, but oh well, not today.
I know exactly how standards groups work. I have been on a few. How about you? And I am not as naive as you to think that when a bunch of companies who have as of yet missed the boat get together, it somehow ends up magically setting standards.
Your comment was a bunch of hot air, with nothing new. Much like this new "Alliance".
At the moment, there are several firms pushing their own platforms with respect to the IoT and none of these things interact with things from other companies.
It's funny, looking at your other/. comments the theme seems to be that using ad hominem is much preferred to actual knowledge or research. Have you tried SmartThings? Or any other home automation controller, for that matter? The whole point of the original ST project was to work with as many devices as possible. It uses a combination of existing HA standards (IP, Z-Wave, Zigbee) with an open API for developers to write their own plugin/protocol support. Which is why I mentioned it. For fuck's sake, next time at least do a LITTLE bit of research before you spout off.
Actually, they don't have much of an "IoT" API, just an API to do things in the cloud with Echo/Alexa. Their IoT integration requires them to integrate with the client (Echo) firmware to do things like discover Philips Hue controllers, etc. (and it's not a matter of "invitation" - if Amazon wanted to be a part of that Alliance they others would be falling all over themselves).
The big company with the "best" IoT API is probably Apple's HomeKit. But I put "best" in quotes because it's only a good API if you agree to drink every last drop of Apple Koolaid - as usual, they have a lot of really onerous restrictions and licensing to make hardware products using it, and there was ZERO thought of compatibility with any of the current home automation protocols like Z-Wave or Zigbee. (Note Google did essentially the same thing with "Works with Nest", an awful cloud-based design to interface with their products).
Having done a lot of home automation, Samsung's SmartThings is probably one of the best products out there (from a major company, at least). It's currently plagued by a horribly buggy software release and overloaded servers, but the technology (a home controller with hardware support for most HA protocols, plus good support for user-created plugins) is at least fairly good, and OPEN. There are a bunch of less well known products (Vera, Wink, HomeSeer, etc) that are as good or better, but given their lack of funding and vision they will probably always be niche...
Except these companies getting together to "lock" up the IoT / Home Automation market is like Hasbro and LEGO getting together to lock up the home construction market.
Samsung is a valid player, IMO (since they bought SmartThings). The rest of them are a really motley crew desperately trying to figure out a way to get their chips or OS into as-of-yet uninvented IoT devices, or get their networks to work with said uninvented devices.
But there is no power grab here, since no power WAS grabbed. Unless Google, Apple, or Amazon pays any attention to them it will be mostly irrelevant.
Actually Samsung bought SmartThings a while ago, so they are at least in the game. The others (excepting maybe GE, which has a fair number of Z-Wave devices) are currently on the sidelines trying to figure out a way to get in it.
Of course there is, that's a stupid statement. There are many zero sum games. It's a huge branch of game theory.
And while long term investing may not be one of those, much of the short term investing of Wall Street (options, futures, many of the strategies of hedge funds) are for all practical purposes zero sum games. And of course usually the professionals (i.e. the already rich people) win, and the average middle class investors convinced to try their "luck" or forced to invest in retirement accounts, lose.
Slashdot Mirror
I'm sure it's hard to pin down the cost of a z13 mainframe, but from what I can tell a high end one is in the several million range at *least*.
Say $5M for one, that runs 5000 really anemic VMs (like 1/10 core each, so they are useless for real load). You can get literally thousands of servers for that price, or more realistically a reasonable number of servers that will still be able to provide multiple cores per VM, making them actually useful under load.
Seems like the only advantage of the mainframe is it's highly fault tolerant. ie. solve the problem with hardware and money instead of good software design. That works to a degree, but it's still vertical scaling. Most Internet companies have long abandoned that model for good software architecture and horizontal scaling - which if done right provides much better fault tolerance, of course.
Well then how is that at all relevant or useful? I bet you can run 2000 VM instances on a high end commodity Linux server if you don't want to them to *do* anything.
The newer mainframes can run 2,000+ instances of Linux.
Citation?
The most I have seen is is on the z13 which has ~140 cores - you can run a bunch of separate instances of Linux but it's nowhere NEAR the same performance as running Linux on fairly cheap dedicated "standard" servers. Not to say there aren't reasons to use a mainframe - just that maximizing virtual Linux instances is rarely one of them.
But that's not really a neural network then
Wha? It's either using a neural network or it isn't. Whether it's a neural network or not doing the computing has nothing to do with the results or the efficiency of the search vs more traditional computing methods.
And I wasn't commenting on the methods or even the effectiveness of the program, just the (probably clueless non-tech) editor's assertion that it applied to "almost any image".
Which is also totally untrue. It should have said "it can determine the location of almost any image ON GOOGLE STREET VIEW".
I'm not sure the author was using the phrase in a mathematical sense. More likely it was in the probabilistic sense of a very high percentage of successes in a number of finite trials.
Not really... the author was mostly using it in the Internet clickbait sense.
If you try it out, it's basically guessing the location of a Google street view image on Google maps. Hardly "determining the location of any image".
Still, it is a lot of fun. Somehow I scored > 10000 points on the "Paris" map even though I was there my first (and only) time last year. Guess it helps to have walked the shit out of a city...
WiiU, PS3, and PS4 all run freebsd as the OS so why not.
PS4, yes. PS3 and WiiU, no.
That may have been true in the distant past, but the median net worth of a US senator is now $2.9M.
Ironically the "poorest" member of Congress may be Rep. Alcee Hastings, who is millions of dollars in debt to law firms from fighting a bribery charge back when he was a Federal judge (also amazing is that he was impeached and removed of his judgeship by Congress, than later elected to it!)
but I actually agree with Ted Cruz on something.
http://www.computerworld.com/a...
His H1B reform bill wants to set a minimum salary on H1B's to make sure it's used to fill vacancies and not cut costs, as well as outlawing "disparagement clauses" in severance agreements that prevent employees from saying anything bad about their ex-employer.
I am pro H1B in general, and the majority of companies in the Bay Area are not using it to replace US workers (there are hundreds of thousands of openings companies are trying to fill, and the starting salaries of many of those are easily in the 6 figures). But clearly there are some companies in the US that abuse the program and there are things that can be done to fix that without throwing it out altogether.
Trump is on record as being ... for Planned Parenthood.
This statement is totally incorrect. Here's an exact quote from Trump in the debate on 2/26:
"So you can say whatever you want, but they have millions of women going through Planned Parenthood that are helped greatly. And I wouldn’t fund it. I would defund it because of the abortion factor, which they say is 3 percent. I don’t know what percentage it is. They say it’s 3 percent."
He freely admits that 97%+ of the services are not related to abortion, it HELPS WOMEN GREATLY, and still is ok with throwing it under the bus. CERTAINLY not "for" PP. And certainly not even for women's rights and health in general.
The difference is Pol Pot, in addition to being a genocidal maniac, was actually very well educated himself...
Whereas Donald Trump probably thinks Khmer Rouge is a type of makeup. ("I know Khmer Rouge, I love Khmer Rouge. They sponsored several of my beauty pageants. My wife Melania is a loyal customer.")
I suppose you are one of the Trump supporters he loves so much? I mean he said it best himself, "I love the poorly educated."
Your post makes zero sense... neither of those have anything to do with the Internet, they are just applications/features that may use it.
Obviously. I know the history as well as anyone. But ARPANET != Internet. The whole point of the Internet is that every node is globally addressable from any other node. That wasn't true until TCP/IP was made the (sole) standard.
The "Internet" was not created in 1969. By most counts you could probably trace it to 1983, when TCP/IP was officially made the protocol for all routing on ARPANET.
And seriously, calling the Internet dead in 1993? If that's true, why are you still posting on /. instead of living offline in your cave?
"do encoding"?? Encoding of WHAT? FFmpeg has a lot of codecs. Your question makes no sense unless you provide some context of what you want to encode.
It is a major release. Some breakage is expected. As long as you stay on the same major it should be stable.
Hah. Good luck with that. FFmpeg is an amazing collection of codecs wrapped in a horrible and ever-changing API. I have used it on a number of projects and it seems every time I upgrade something breaks.
That would be true if the HA market wasn't already like 30+ years old. The actual "players" and innovators already have plenty of patents or prior art.
There are some of these "Alliances" that actually build momentum and make sense (see: UHD Alliance). This does not look like one of those.
So, I suppose I should take the high road and ignore your assholish tone, but oh well, not today.
I know exactly how standards groups work. I have been on a few. How about you? And I am not as naive as you to think that when a bunch of companies who have as of yet missed the boat get together, it somehow ends up magically setting standards.
Your comment was a bunch of hot air, with nothing new. Much like this new "Alliance".
At the moment, there are several firms pushing their own platforms with respect to the IoT and none of these things interact with things from other companies.
It's funny, looking at your other /. comments the theme seems to be that using ad hominem is much preferred to actual knowledge or research. Have you tried SmartThings? Or any other home automation controller, for that matter? The whole point of the original ST project was to work with as many devices as possible. It uses a combination of existing HA standards (IP, Z-Wave, Zigbee) with an open API for developers to write their own plugin/protocol support. Which is why I mentioned it. For fuck's sake, next time at least do a LITTLE bit of research before you spout off.
Actually, they don't have much of an "IoT" API, just an API to do things in the cloud with Echo/Alexa. Their IoT integration requires them to integrate with the client (Echo) firmware to do things like discover Philips Hue controllers, etc. (and it's not a matter of "invitation" - if Amazon wanted to be a part of that Alliance they others would be falling all over themselves).
The big company with the "best" IoT API is probably Apple's HomeKit. But I put "best" in quotes because it's only a good API if you agree to drink every last drop of Apple Koolaid - as usual, they have a lot of really onerous restrictions and licensing to make hardware products using it, and there was ZERO thought of compatibility with any of the current home automation protocols like Z-Wave or Zigbee. (Note Google did essentially the same thing with "Works with Nest", an awful cloud-based design to interface with their products).
Having done a lot of home automation, Samsung's SmartThings is probably one of the best products out there (from a major company, at least). It's currently plagued by a horribly buggy software release and overloaded servers, but the technology (a home controller with hardware support for most HA protocols, plus good support for user-created plugins) is at least fairly good, and OPEN. There are a bunch of less well known products (Vera, Wink, HomeSeer, etc) that are as good or better, but given their lack of funding and vision they will probably always be niche...
Except these companies getting together to "lock" up the IoT / Home Automation market is like Hasbro and LEGO getting together to lock up the home construction market.
Samsung is a valid player, IMO (since they bought SmartThings). The rest of them are a really motley crew desperately trying to figure out a way to get their chips or OS into as-of-yet uninvented IoT devices, or get their networks to work with said uninvented devices.
But there is no power grab here, since no power WAS grabbed. Unless Google, Apple, or Amazon pays any attention to them it will be mostly irrelevant.
Actually Samsung bought SmartThings a while ago, so they are at least in the game. The others (excepting maybe GE, which has a fair number of Z-Wave devices) are currently on the sidelines trying to figure out a way to get in it.
Biased much against Google?
Cook posted a letter yesterday, Pichai responded today. OH MY GOOD HOW COULD IT TAKE SO LONG!?
There is no such thing as a zero sum game.
Of course there is, that's a stupid statement. There are many zero sum games. It's a huge branch of game theory.
And while long term investing may not be one of those, much of the short term investing of Wall Street (options, futures, many of the strategies of hedge funds) are for all practical purposes zero sum games. And of course usually the professionals (i.e. the already rich people) win, and the average middle class investors convinced to try their "luck" or forced to invest in retirement accounts, lose.