Right. I remember those days. It was back when everyone was on Yahoo!What made Yahoo! chat rooms work was that you could go there and find people
I really doubt that. Newsgroups are as old as the internet itself. And talkd and IRC are also quite old. Waaaay before Yahoo.
What makes Facebook work is that is where everyone is now.
True.
This is called "network effect"
This is EXACTLY what you want from a social network. Or you'd have skype, google talk and the remaining crap to talk with people. A niche network - the shit G+ was catering to when it was launched (by arrogant nerds for nerds) - will not fly. Internet IS NOT for the elite. And Facebook understands that.
ou realize that people were talking online and sharing pictures, personal updates, etc, on the internet long, long before Facebook ever existed, right?
Yeah. Except geocities, gopher, thematic webrings, personal home pages, personal gnome pages (wink wink gifs) and whatnot. And all of that required a browser, some html knowledge, a domain name (optional), some blink tags and the actual content. So you could congratulate your colleague in his newborn son (6 months old by the time you find out) as he's leaving the company (that's how you found out, he has a own domain email!). That appeals to the working class, why not? "Here, have some tech - in 6 months, you'll be able to barely update your friends on your cat pictures AND your cooking recipes by installing this 25 pieces of software"
Refusal to use them is not the same as not having them.
I'd assume that you, AC, are followed with utter relevance.
no current system/application/internet thing exists to make the experience of 'you and all of your friends visiting the same varying virtual destinations a social experience'
Where the f*** did you read that in my comment?
Facebook only offers you your info for sale, and a way to consistently communicate in recognizable patterns globally through a single identity, not a real social experience
So does your provider, and everybody else. What do I care that my music listening preferences are blasted into the world and data mined? Do you know what trends are? They are data-mined:). And as "social experience", I choose to have a life offline. Like most users do.
modern research (http://www.scientificamerican.com/
I'd argue that scientific american is neither a modern and an acccurate reference when talking about social networks. But I get a bit bitchy about details.
I'd be willing to bet that what our brains are currently tuned to interpret as a truly "social"
Brains process patterns. Proximity (specially multi-sensorial) is a pattern, much more fullfilling than online experiences.
i shouldn't post slashdot comments on St. Patty's Day.
Waaaay ahead of you:D Happy St Patrick's day from Portugal;)
A social network is only relevant if it is "relevant" (aka if it appeases your social needs). Usually this requires a clear market winner. That's how we got CD vs DAT, VHS vs BetaMax and BluRay vs HD-DVD.
. There's a slight possibility that both drives fail at the same time but that's a very remote chance.
Well, its not. Unless you go server-grade, the chance you will find flipping bits and CRC errors is quite high. If the drive is full, syncing the drive may easily cause stress failure on high-capacity disks (consumer-grade disks are not designed to spit 2TB at maximum data rate nor the IOPS this envolves). This is one of the reasons why RAID10 is prefered over RAID5 - the rebuilt effort is smaller, and the stress on the disk drives will be much lower.
Add the cost of putting meaningful bytes on them. Platform, electricity, etc.
Instead of 1 year of service I can expect 5 years out of a SATA drive typically
Except if your house is robbed, you drop the drives, bad luck, fire, flood, short circuit, bad electronics, etc. All those things people don't care about until its too late.
So if nothing goes wrong, I've saved myself $480, if something does go wrong with both drives, I've saved myself $360
So, how do you backup those 2 drives to make sure you don't lose data? At that point, this kind of service seems really cheap.
As an example, I run my own online storage system, and have roughly 3TB of ZFS, redundant storage I use to backup my critical data. It's still around 0.02Eur/GB, plus the time to maintain it.
Eh? Can't you just throttle it at your router? How hard is that? Hand in your/. credentials plz...
So, please care to explain how to throttle a specific upload on an HTTPS connection using a home router, please? So lets say you're uploading to Gdrive and want to keep using GMail...
I am sorry, but I don't understand: why is it okay for Microsoft to use to BSD code, without giving BSD any credit, but not Linux?
Where did you get that idea? Microsoft used the BSD-based TCP stack on previous versions of windows, and the disclaimers are fairly well documented, even on the header files. The BSD clause is let intact, as required. The Microsoft Services For Unix was (AFAIK) based on OpenBSD tools and some GPL stuff, all also in compliance with the license (an old version is described in http://technet.microsoft.com/e...)
y understanding is: the ISC, MIT and BSD-licenses allow for sublicensing without making any modifications so the Linux devs are perfectly within the license when they sublicense the original code under the GPL
You CANNOT rip the BSD disclaimer. Its right there on the license. Theo is right.
And tecnically, you CANNOT dual-license a BSD file with GPL without any change. If you do it, the less restrictive license takes precedence. You can change the file and have your own modifications under GPL if you want, but for the rest of the code, the GPL clauses are void because BSD is less restrictive and the content is ALSO licensed under BSD.
Also, I am not sure that Theo is justified in ranting about "the Linux people" when this was one incident that happened about 15 years ago, and was corrected.
Another guy already replied to this. If a guy chooses GPL because he thinks its a better license, he should at least have the same respect for other licenses. More often than not, this doesn't happen. The fallacy of repeating RMS GPL bullshit as facts doesn't make them come to reality, and it hurts the OSS ecosystem as a whole.
Its not. You retain full authorship, so deleting the license and pasting a new one directly violates the license.
I can take BSD code, and relicense it any way I please
No. You cannot remove the BSD disclaimer from the source and/or claim it as your own.
If you want kernel improvements to be implemented back, why license your code under the BSD to begin with?
One of the reasons TCP/IP is a huge success is because it was BSD licensed. As many fundamental daemons that gave name to some pretty well-known services that we now call "internet" collectively.
When you release your code BSD, you allow relicensing. That's why MS prefers the BSD license.
Also Apple. And every other sane company that doesn't do business selling "open source", but products with added value. If you really look into it, very few relevant OSS projects use GPL, aside from Linux infrastructure and the whole FSF ecosystem.
If BSD advocates want to thump their chests about their licence being so free, then why do they bitch when Linux - but not Microsoft - takes them up on their offer?
No one bitched about "taking the offer". "Taking the offer" means respecting the license - hey, even Microsoft does it, right? AFAIK (your link isn't opening) the bitching was about ignoring the licensing terms.
I can take one of his questions:D They are all quite relevant and interesting, but I'd add HammerFS to the list also, since its one of the projects listed in Google SoC.
I should also add that, when OpenBSD added randomness to the ld.so and mmap(), a non-trivial amount of bugs was discovered in some well-known OSS projects. Probably every OSS user has benefited - either directly or indirectly - from OpenBSD; that doesn't mean we should build altars to worship Theo when - after all - it is a collective effort.
The point the parent was making is that catering to actual security and catering to egos are different things. And while it has been proven that Theo was wrong a whole bunch of times (and right a whole bunch of other times), it still has no effect on him. Smart people will often realize they were wrong and be happy with it (after all, they fixed a problem AND learned a new thing). Ego maniacs will silently ignore the fact and accept new change with smugness.
I'm an OpenBSD user since 2.9. I stopped using it in most professional setups around 4.2-4.4, because I find the maintenance cycle unnaceptable. However, I still buy both CDs and assorted merch from the project when I can, because I see real value in the team. But the truth is, while some of the side projects are quite alive (OpenSSH, PF stuff, OpenBGPD, etc), OpenBSD itself hasn't aged quite well. The VFS layer is a mess. Thread support is subpar. No container support whatsoever. No ACL support, no MAC. No virtualization support. Crappy SMP support. And this is so obvious, that some ex-developers decided to fork it and create https://www.bitrig.org./ And lets face it, it seems like local attacks aren't even considered vulnerabilities. The whole remote exploit stuff is such a bullshit - I remember one release that was antecipated 1-2 weeks so some Apache hole would not count. And while no one really cares about OpenBSD anymore, their subprojects are beneficial to almost every other operating system - if it wasn't for them (specially OpenSSH), I think OpenBSD would be long gone.
Deflation isn't a problem. Only in your head, and the heads of the Keynesians.
Deflation isn't always a problem. As inflation isn't a problem, then balance exists. And there is a whole world in my head.
Investment becomes attractive in a deflationary environment when a risk-assessed 5% loan returns more than a 2% "sit on cash".
So, no more buying houses on credit, then. And no more long-term renting on developing areas (because the rent will increase in real value according to the increase in real value of the property).
In case you haven't noticed, in your central-bank engineered inflationary system we're in right now
Not really, no. I'm a EU citizen, and I actually lived the effects of deflation the last couple of years. I've also lived the effects of some periods of excess inflation, so its not only ideology.
banks are sitting on over $2trillion in excess reserves and the Fed is paying them interest NOT TO INVEST THE MONEY.
As I said, I'm a EU citizen, and I don't follow up internal US economics (aside from the usual major announcements). I would say there are probably good reasons to do that (specially considering the country increase in debt the last years and the money emissions made), but I would also point out that - while US is a central pillar in the world economy, the US is not the world.
If you want to go literal, maybe. Other way of looking at it is decreasing of percepted value. While high inflation is bad, some inflation is necessary for the economy to work. The internet is riddled with articles about this, but check eg. http://business.time.com/2011/...
Deflation is creation of wealth
Well, its not. Is the increase of perception of value caused (usually) by scarcity or hoarding of resources/currency. Eg. diamonds, while having no intrinsic value by themselves, are expensive because the diamond market works based on scarcity (google "diamond market fake scarcity" for a ton of articles about it, and other markets that work as that).
Now imagine your day-to-day life based on scarcity - you are going to buy bread, and it costs X in absolute value, but tomorrow will be 10% more expensive. Now imagine you were borrowing from your bread provider - you pick the bread today, but will pay next week. Next week the real bread value will be much higher, and what you borrowed with absolute value X may have doubled in absolute value. No wealth there - in fact, you'll probably be paying twice as much in absolute value, plus the profit the bread provider expects to allow you to borrow.
Wait until the people who bought them there realize that a 50% gain after a 50% loss puts you in the hole.
There is also another point *against* Bitcoin. Because it is deflationary by nature, it promotes hoarding of the currency (eg. if Bitcoin is low now, and if you believe in it, just hang on to it until hopefully will gain more value). Current capitalist systems are inflationary, so if you sit on your pile of cash it will decrease value over time, not increase. This gives a huge incentive to invest that money - like deposit in a bank, buy some stock or bonds, do riskier high-return investments, etc. Thats how you increase your capital - you just don't sit on it, you make things happen that will generate profit. And this keeps the world turning, regardless of personal preferences of who's in charge of it. With deflationary currencies, eg. companies don't need to invest to have profit (eg. you would not have Internet, because why spend 300 billion on undersea cables when you can just sit on that money until it increases in value?). And with unregulated currencies, the loan system cannot work reliably - and this is essential to make the world as we know work.
My opinion is, Bitcoin is a ponzi scheme. Even if it gets adopted officially. (Yeah I know, fiat currency isn't that far from a ponzi scheme, either)
Most of what is "modern x86" today is akin to RISC processors - CISC instructions are decomposed into micro-ops (similar to RISC) instructions, and fed into the execution pipelines. Some of those micro-ops are executed by following a set of rules defined by the cpu's own firmware (microcode), others are not. Some of the micro-ops that aren't have additional complex circuitry, and frequently are legacy instructions. You may not be aware of this, but Intel has been doing (rather unsucessfully, I must add) RISC processors at least since the beginning of the 90's, and that all x86 cpu's since Pentium Pro borrow concepts from the ix60 line.
(...) and imaginary Intel roadmap.
I would be quite surprised if they didn't deliver. This is not a startup pulling data out of their ass, and you already have simple circuits done with 10nm from competitors (http://www.samsung.com/global/business/semiconductor/news-events/press-releases/detail?newsId=12281). In 8 years, Intel went from 65nm with the first Core CPUs, to 22nm last year. Odds are, the roadmap is right.
CPU cache OVERWHELMINGLY dominates the x86/x64 die.
Yeah, so? Are you saying that 4MB on-die cache is somehow smaller in an ARM chip? And if you look at it closely, you don't see a huge increase in cache sizes in the past 10 years, do you? The tendency is to actually shrink slightly in the future, because of latency issues. But hey, lets focus on what we know today - 4MB of cache in 32nm (common today for ARM) is not the same area as in 22nm (used *today* on Haswell). Or in 5nm. And if you can free an additional 20% of real estate in the chip by cleaning seldom used parts, you do get a smaller die.
Regarding pricing, memory manufacturing and testing is a somewhat cheap process, when compared to testing the functions of the CPU. Cutting parts and shrinking the design would produce less defects, and it would probably have a direct impact on price. I know, utterly ridiculous.
Yeah, sure. So you know every binary your current compiler toolchain has, AND additionally you developed a fancy way for a local user to be forbidden of downloading executable programs. Its one of those things that sound easy to do/use until you have to use them.
You are right. I wasn't aware that Intel stopped manufacturing them in 2007 - but I haven't read anything about expired licensing. The VHDL for the cores is somewhat widely available, but haven't find any info about if they are "free". I would doubt they have expired, but they may have opened it to everyone - after all, the early Beatles catalog is not public domain, is it?
Most security vulnerabilities are in the applications themselves, eg buffer overflows, or on the client side
This adds nothing to the discussion. Just like VM's. If you can tamper with data via security vulnerabilities, there is nothing virtualization can do to help you.
Let's see some evidence where weakness in concern-separation from VMware instances or sudo glitches is a major contributor to malware mishaps these days.
Well, lets see some evidence that virtualization somehow magically reduces or mitigates security problems, as you suggested. Servers are servers, regardless of the physical infrastructure. If they get compromised, you may be screwed - regardless of it is a logic instance on some blade server or a good ol' physical server.
I suppose the main vulnerability is a bit less control against insider malfeasance, and those are mostly due to configuration errors or corrupt admins.
The most common mistake I see (and I'm not a DevOps, so it may be biased) is authentication - once a single server is compromised, it may be possible to access other servers, due to proliferation of key-based authentication without password. It is a lot harder to properly secure a bunch of isolated servers that somehow work as a single service than a single machine, without sacrificing ease of administration. Good chances are, on a given cluster, all webservers have the same credentials - and this gets even more common when using management tools like puppet or cfengine.
I'd say you're off the curve of reasonable expectations if you're asking for mainframe-style "trusted" isolation on a setup of only a few (or just one) PC-grade servers in which you have all applications and services running together along with a variety of login access from different categories of users who may be potential attackers.
I'm not, You suggested that, with VM. I just pointed out how unrealistic that is - in the real world.
I say "PC-grade" because your scenario sounds economically uninteresting -- important enough to protect as you want (with excessive apps & users), but not important enough that there's budget to do hardware separation.
On the contrary. You were promoting VMs as a tool for security - its not. More often than not, it actually extends the surface of attack. VMs do have huge advantages, not only to consolidate resources, but also as an easy way of providing fault-tolerance and high-availability. Those are some of the key benefits of virtualization. Security is not one of them.
I agree that BSD Capsicums (etc) may be a good fit for these outlier use-cases, or special situations, but mostly if your establishment is willing to make a heavy technology investment in going that route.
the parent thread was about the broken unix security model. Not that some of the available tools on the BSDs would require considerable effort - having an application xyz running on 1 machine/vm or in 10.000 machines/vm is roughly the same thing, from an administrative point of view. I could argue that, academically, Windows has a far superior security model than anything you can find on unix systems, but that would be off-topic. The same way considerations about edge cases and cluster sizes are - its not what is being discussed. VMs add nothing - directly - to the security discussion.
For anything designed the last 5 years it's more than likely that those pesky old 8051s have been replaced by ARMs, Coretex M0s, M3s and M4s
Actually, its not. For many applications, this would require rewriting of the software stack, for a chip with roughly the same die size and possibly less funcionality. 8051 is a microcontroller, not a microprocessor - and it is 8bit. It does not distinguish between IO and memory access. However, it does have a semi-complete serial interface, and 8 lines of digital I/O - and this is the base version. A ton of variations exist, with extra ports, A/D and D/A funcionality, extra timers, assorted controllers in-chip and multiple RAM and ROM configurations. It would never be replaced by a powerful 16-bit or 32-bit processor, because its not even on the same league.
Slashdot Mirror
FB will be "lame" sooner or later. Give it 2-5 years and everyone will be jumping on that next best thing.
I'd generally agree with this, but I was the guy saying that 5 years ago. Everything has a peak, but it seems to be too soon to tell.
Right. I remember those days. It was back when everyone was on Yahoo!What made Yahoo! chat rooms work was that you could go there and find people
I really doubt that. Newsgroups are as old as the internet itself. And talkd and IRC are also quite old. Waaaay before Yahoo.
What makes Facebook work is that is where everyone is now.
True.
This is called "network effect"
This is EXACTLY what you want from a social network. Or you'd have skype, google talk and the remaining crap to talk with people. A niche network - the shit G+ was catering to when it was launched (by arrogant nerds for nerds) - will not fly. Internet IS NOT for the elite. And Facebook understands that.
ou realize that people were talking online and sharing pictures, personal updates, etc, on the internet long, long before Facebook ever existed, right?
Yeah. Except geocities, gopher, thematic webrings, personal home pages, personal gnome pages (wink wink gifs) and whatnot. And all of that required a browser, some html knowledge, a domain name (optional), some blink tags and the actual content. So you could congratulate your colleague in his newborn son (6 months old by the time you find out) as he's leaving the company (that's how you found out, he has a own domain email!). That appeals to the working class, why not? "Here, have some tech - in 6 months, you'll be able to barely update your friends on your cat pictures AND your cooking recipes by installing this 25 pieces of software"
Refusal to use them is not the same as not having them.
I'd assume that you, AC, are followed with utter relevance.
no current system/application/internet thing exists to make the experience of 'you and all of your friends visiting the same varying virtual destinations a social experience'
Where the f*** did you read that in my comment?
Facebook only offers you your info for sale, and a way to consistently communicate in recognizable patterns globally through a single identity, not a real social experience
So does your provider, and everybody else. What do I care that my music listening preferences are blasted into the world and data mined? Do you know what trends are? They are data-mined :). And as "social experience", I choose to have a life offline. Like most users do.
modern research (http://www.scientificamerican.com/
I'd argue that scientific american is neither a modern and an acccurate reference when talking about social networks. But I get a bit bitchy about details.
I'd be willing to bet that what our brains are currently tuned to interpret as a truly "social"
Brains process patterns. Proximity (specially multi-sensorial) is a pattern, much more fullfilling than online experiences.
i shouldn't post slashdot comments on St. Patty's Day.
Waaaay ahead of you :D Happy St Patrick's day from Portugal ;)
That is a long living anomaly, no?
A social network is only relevant if it is "relevant" (aka if it appeases your social needs). Usually this requires a clear market winner. That's how we got CD vs DAT, VHS vs BetaMax and BluRay vs HD-DVD.
Lack of options. And no, G+ doesn't cut it.
Transfering files over SSH is slow as f****, specially with high-speed links.
. There's a slight possibility that both drives fail at the same time but that's a very remote chance.
Well, its not. Unless you go server-grade, the chance you will find flipping bits and CRC errors is quite high. If the drive is full, syncing the drive may easily cause stress failure on high-capacity disks (consumer-grade disks are not designed to spit 2TB at maximum data rate nor the IOPS this envolves). This is one of the reasons why RAID10 is prefered over RAID5 - the rebuilt effort is smaller, and the stress on the disk drives will be much lower.
I can pickup a 1 TB drive right now for ~$60
Add the cost of putting meaningful bytes on them. Platform, electricity, etc.
Instead of 1 year of service I can expect 5 years out of a SATA drive typically
Except if your house is robbed, you drop the drives, bad luck, fire, flood, short circuit, bad electronics, etc. All those things people don't care about until its too late.
So if nothing goes wrong, I've saved myself $480, if something does go wrong with both drives, I've saved myself $360
So, how do you backup those 2 drives to make sure you don't lose data? At that point, this kind of service seems really cheap.
As an example, I run my own online storage system, and have roughly 3TB of ZFS, redundant storage I use to backup my critical data. It's still around 0.02Eur/GB, plus the time to maintain it.
Eh? Can't you just throttle it at your router? How hard is that? Hand in your /. credentials plz...
So, please care to explain how to throttle a specific upload on an HTTPS connection using a home router, please? So lets say you're uploading to Gdrive and want to keep using GMail...
I am sorry, but I don't understand: why is it okay for Microsoft to use to BSD code, without giving BSD any credit, but not Linux?
Where did you get that idea? Microsoft used the BSD-based TCP stack on previous versions of windows, and the disclaimers are fairly well documented, even on the header files. The BSD clause is let intact, as required. The Microsoft Services For Unix was (AFAIK) based on OpenBSD tools and some GPL stuff, all also in compliance with the license (an old version is described in http://technet.microsoft.com/e...)
y understanding is: the ISC, MIT and BSD-licenses allow for sublicensing without making any modifications so the Linux devs are perfectly within the license when they sublicense the original code under the GPL
You CANNOT rip the BSD disclaimer. Its right there on the license. Theo is right.
And tecnically, you CANNOT dual-license a BSD file with GPL without any change. If you do it, the less restrictive license takes precedence. You can change the file and have your own modifications under GPL if you want, but for the rest of the code, the GPL clauses are void because BSD is less restrictive and the content is ALSO licensed under BSD.
Also, I am not sure that Theo is justified in ranting about "the Linux people" when this was one incident that happened about 15 years ago, and was corrected.
Another guy already replied to this. If a guy chooses GPL because he thinks its a better license, he should at least have the same respect for other licenses. More often than not, this doesn't happen. The fallacy of repeating RMS GPL bullshit as facts doesn't make them come to reality, and it hurts the OSS ecosystem as a whole.
As I understand it, BSD is almost public domain.
Its not. You retain full authorship, so deleting the license and pasting a new one directly violates the license.
I can take BSD code, and relicense it any way I please
No. You cannot remove the BSD disclaimer from the source and/or claim it as your own.
If you want kernel improvements to be implemented back, why license your code under the BSD to begin with?
One of the reasons TCP/IP is a huge success is because it was BSD licensed. As many fundamental daemons that gave name to some pretty well-known services that we now call "internet" collectively.
When you release your code BSD, you allow relicensing. That's why MS prefers the BSD license.
Also Apple. And every other sane company that doesn't do business selling "open source", but products with added value. If you really look into it, very few relevant OSS projects use GPL, aside from Linux infrastructure and the whole FSF ecosystem.
If BSD advocates want to thump their chests about their licence being so free, then why do they bitch when Linux - but not Microsoft - takes them up on their offer?
No one bitched about "taking the offer". "Taking the offer" means respecting the license - hey, even Microsoft does it, right? AFAIK (your link isn't opening) the bitching was about ignoring the licensing terms.
I can take one of his questions :D They are all quite relevant and interesting, but I'd add HammerFS to the list also, since its one of the projects listed in Google SoC.
I should also add that, when OpenBSD added randomness to the ld.so and mmap(), a non-trivial amount of bugs was discovered in some well-known OSS projects. Probably every OSS user has benefited - either directly or indirectly - from OpenBSD; that doesn't mean we should build altars to worship Theo when - after all - it is a collective effort.
The point the parent was making is that catering to actual security and catering to egos are different things. And while it has been proven that Theo was wrong a whole bunch of times (and right a whole bunch of other times), it still has no effect on him. Smart people will often realize they were wrong and be happy with it (after all, they fixed a problem AND learned a new thing). Ego maniacs will silently ignore the fact and accept new change with smugness.
I'm an OpenBSD user since 2.9. I stopped using it in most professional setups around 4.2-4.4, because I find the maintenance cycle unnaceptable. However, I still buy both CDs and assorted merch from the project when I can, because I see real value in the team. But the truth is, while some of the side projects are quite alive (OpenSSH, PF stuff, OpenBGPD, etc), OpenBSD itself hasn't aged quite well. The VFS layer is a mess. Thread support is subpar. No container support whatsoever. No ACL support, no MAC. No virtualization support. Crappy SMP support. And this is so obvious, that some ex-developers decided to fork it and create https://www.bitrig.org./ And lets face it, it seems like local attacks aren't even considered vulnerabilities. The whole remote exploit stuff is such a bullshit - I remember one release that was antecipated 1-2 weeks so some Apache hole would not count. And while no one really cares about OpenBSD anymore, their subprojects are beneficial to almost every other operating system - if it wasn't for them (specially OpenSSH), I think OpenBSD would be long gone.
Deflation isn't a problem. Only in your head, and the heads of the Keynesians.
Deflation isn't always a problem. As inflation isn't a problem, then balance exists. And there is a whole world in my head.
Investment becomes attractive in a deflationary environment when a risk-assessed 5% loan returns more than a 2% "sit on cash".
So, no more buying houses on credit, then. And no more long-term renting on developing areas (because the rent will increase in real value according to the increase in real value of the property).
In case you haven't noticed, in your central-bank engineered inflationary system we're in right now
Not really, no. I'm a EU citizen, and I actually lived the effects of deflation the last couple of years. I've also lived the effects of some periods of excess inflation, so its not only ideology.
banks are sitting on over $2trillion in excess reserves and the Fed is paying them interest NOT TO INVEST THE MONEY.
As I said, I'm a EU citizen, and I don't follow up internal US economics (aside from the usual major announcements). I would say there are probably good reasons to do that (specially considering the country increase in debt the last years and the money emissions made), but I would also point out that - while US is a central pillar in the world economy, the US is not the world.
Inflation is theft of value.
If you want to go literal, maybe. Other way of looking at it is decreasing of percepted value. While high inflation is bad, some inflation is necessary for the economy to work. The internet is riddled with articles about this, but check eg. http://business.time.com/2011/...
Deflation is creation of wealth
Well, its not. Is the increase of perception of value caused (usually) by scarcity or hoarding of resources/currency. Eg. diamonds, while having no intrinsic value by themselves, are expensive because the diamond market works based on scarcity (google "diamond market fake scarcity" for a ton of articles about it, and other markets that work as that).
Now imagine your day-to-day life based on scarcity - you are going to buy bread, and it costs X in absolute value, but tomorrow will be 10% more expensive. Now imagine you were borrowing from your bread provider - you pick the bread today, but will pay next week. Next week the real bread value will be much higher, and what you borrowed with absolute value X may have doubled in absolute value. No wealth there - in fact, you'll probably be paying twice as much in absolute value, plus the profit the bread provider expects to allow you to borrow.
Wait until the people who bought them there realize that a 50% gain after a 50% loss puts you in the hole.
There is also another point *against* Bitcoin. Because it is deflationary by nature, it promotes hoarding of the currency (eg. if Bitcoin is low now, and if you believe in it, just hang on to it until hopefully will gain more value). Current capitalist systems are inflationary, so if you sit on your pile of cash it will decrease value over time, not increase. This gives a huge incentive to invest that money - like deposit in a bank, buy some stock or bonds, do riskier high-return investments, etc. Thats how you increase your capital - you just don't sit on it, you make things happen that will generate profit. And this keeps the world turning, regardless of personal preferences of who's in charge of it. With deflationary currencies, eg. companies don't need to invest to have profit (eg. you would not have Internet, because why spend 300 billion on undersea cables when you can just sit on that money until it increases in value?). And with unregulated currencies, the loan system cannot work reliably - and this is essential to make the world as we know work.
My opinion is, Bitcoin is a ponzi scheme. Even if it gets adopted officially. (Yeah I know, fiat currency isn't that far from a ponzi scheme, either)
(...) and imaginary Intel roadmap.
I would be quite surprised if they didn't deliver. This is not a startup pulling data out of their ass, and you already have simple circuits done with 10nm from competitors (http://www.samsung.com/global/business/semiconductor/news-events/press-releases/detail?newsId=12281). In 8 years, Intel went from 65nm with the first Core CPUs, to 22nm last year. Odds are, the roadmap is right.
CPU cache OVERWHELMINGLY dominates the x86/x64 die.
Yeah, so? Are you saying that 4MB on-die cache is somehow smaller in an ARM chip? And if you look at it closely, you don't see a huge increase in cache sizes in the past 10 years, do you? The tendency is to actually shrink slightly in the future, because of latency issues. But hey, lets focus on what we know today - 4MB of cache in 32nm (common today for ARM) is not the same area as in 22nm (used *today* on Haswell). Or in 5nm. And if you can free an additional 20% of real estate in the chip by cleaning seldom used parts, you do get a smaller die.
Regarding pricing, memory manufacturing and testing is a somewhat cheap process, when compared to testing the functions of the CPU. Cutting parts and shrinking the design would produce less defects, and it would probably have a direct impact on price. I know, utterly ridiculous.
Yeah, sure. So you know every binary your current compiler toolchain has, AND additionally you developed a fancy way for a local user to be forbidden of downloading executable programs. Its one of those things that sound easy to do/use until you have to use them.
You are right. I wasn't aware that Intel stopped manufacturing them in 2007 - but I haven't read anything about expired licensing. The VHDL for the cores is somewhat widely available, but haven't find any info about if they are "free". I would doubt they have expired, but they may have opened it to everyone - after all, the early Beatles catalog is not public domain, is it?
Most security vulnerabilities are in the applications themselves, eg buffer overflows, or on the client side
This adds nothing to the discussion. Just like VM's. If you can tamper with data via security vulnerabilities, there is nothing virtualization can do to help you.
Let's see some evidence where weakness in concern-separation from VMware instances or sudo glitches is a major contributor to malware mishaps these days.
Well, lets see some evidence that virtualization somehow magically reduces or mitigates security problems, as you suggested. Servers are servers, regardless of the physical infrastructure. If they get compromised, you may be screwed - regardless of it is a logic instance on some blade server or a good ol' physical server.
I suppose the main vulnerability is a bit less control against insider malfeasance, and those are mostly due to configuration errors or corrupt admins.
The most common mistake I see (and I'm not a DevOps, so it may be biased) is authentication - once a single server is compromised, it may be possible to access other servers, due to proliferation of key-based authentication without password. It is a lot harder to properly secure a bunch of isolated servers that somehow work as a single service than a single machine, without sacrificing ease of administration. Good chances are, on a given cluster, all webservers have the same credentials - and this gets even more common when using management tools like puppet or cfengine.
I'd say you're off the curve of reasonable expectations if you're asking for mainframe-style "trusted" isolation on a setup of only a few (or just one) PC-grade servers in which you have all applications and services running together along with a variety of login access from different categories of users who may be potential attackers.
I'm not, You suggested that, with VM. I just pointed out how unrealistic that is - in the real world.
I say "PC-grade" because your scenario sounds economically uninteresting -- important enough to protect as you want (with excessive apps & users), but not important enough that there's budget to do hardware separation.
On the contrary. You were promoting VMs as a tool for security - its not. More often than not, it actually extends the surface of attack. VMs do have huge advantages, not only to consolidate resources, but also as an easy way of providing fault-tolerance and high-availability. Those are some of the key benefits of virtualization. Security is not one of them.
I agree that BSD Capsicums (etc) may be a good fit for these outlier use-cases, or special situations, but mostly if your establishment is willing to make a heavy technology investment in going that route.
the parent thread was about the broken unix security model. Not that some of the available tools on the BSDs would require considerable effort - having an application xyz running on 1 machine/vm or in 10.000 machines/vm is roughly the same thing, from an administrative point of view. I could argue that, academically, Windows has a far superior security model than anything you can find on unix systems, but that would be off-topic. The same way considerations about edge cases and cluster sizes are - its not what is being discussed. VMs add nothing - directly - to the security discussion.
For anything designed the last 5 years it's more than likely that those pesky old 8051s have been replaced by ARMs, Coretex M0s, M3s and M4s
Actually, its not. For many applications, this would require rewriting of the software stack, for a chip with roughly the same die size and possibly less funcionality. 8051 is a microcontroller, not a microprocessor - and it is 8bit. It does not distinguish between IO and memory access. However, it does have a semi-complete serial interface, and 8 lines of digital I/O - and this is the base version. A ton of variations exist, with extra ports, A/D and D/A funcionality, extra timers, assorted controllers in-chip and multiple RAM and ROM configurations. It would never be replaced by a powerful 16-bit or 32-bit processor, because its not even on the same league.