The first rule of the tautology club is the first rule of the tautology club:)
And coincidentally, I do have a fair amount of experience with ATMs (probably outdated by now, though). Boot time, at least in my country, isn't a factor: nowadays, most of them actually run a stripped down version of Windows 2000...
That statement shows a severe lack of understanding on IT Technology... The PIN number doesn't need a boot of the whole OS system to be processed, all relevant information should either be in RAM memory, or contained in the card itself...
You're not. I'm also almost blind on one eye (perfect peripheral vision, so no depth perception problems, but forward vision is around 20%), and 3D simply doesn't work for me, regardless of the technology used.
I'm not exactly concerned about exclusion, yet. I've watched the 2D versions of most recent 3D movies (Avatar excluded), and thought they all sucked. The only thing going for them was the 3D gimmick, so most people (my wife included) liked the visuals and felt they got their money's worth; for me, not so much.
I know at least 2 other people with the same condition I have (1 relative, 1 co-worker), and several others with various eyesight problems that prevent them from fully seeing 3D movies. I doubt (at least hope) we'll ever get to a point where 3D stops being a gimmick and becomes an integral part of the movie (like... needing a 3D view to fully understand a scene), at least until holograms (actual 3D projections instead of visual tricks) come along...
In the last 5 years, I've had to take the laptop out in most of Europe, including the UK (which you're listing), Belgium, Netherlands, France (there I even had to turn it on! what if I had had a dead battery?), Spain, Portugal, and Italy. I don't remember having to do it in Asia, but wherever I stop in Europe, it's "laptop out, sir".
We've always had bandwidth caps where I live, ADSL included.
The cap for the mobile access my company sells is currently at 25 GB/month, but there are frequent promotions of the "join until the end of August and get unlimited downloads forever" type, so we have a lot of "unlimited" customers. And no, they're not an issue.
There are many European countries selling 3G based "broadband Internet", with dedicated devices (not phones, just simple USB HSDPA modems that take a SIM card).
In those countries, the basestations should be more than capable of handling hogs (either by allowing the traffic, or by throttling abusers so they don't mess up everybody else's connections).
As for tethering... That's the main reason I gave away my iPhone 3G (I work for a mobile operator, got one at launch day, and gave it away to the person in my team that pulled the longest straw. Yes, literally.). Every phone I've had for the last 4 years has allowed me unrestricted bluetooth connectivity to the internet, starting with GPRS and now with HSDPA), and it is something I use *a lot*. I don't have to take my phone out of the pocket, I just fire up the laptop, push a button, and I'm done.
Sorry, but I'll have to disagree on the daemontools issue. Yes, it's a completely different way of thinking, but once you get used to it (and to ucspi-tcp), there's a lot of stuff that can be done with it...
All servers under my direct or indirect control have them installed as a rule (and we're talking a few hundred servers here), and most of them are actively using them.
Need a quick TCP-based daemon for menial tasks (like outputting a status line for monitorization?) tcpserver + supervise. A looping task? supervise that task, and add a sleep at the end of the run script. Don't want to write a logging function for some app? print to stdout/stderr, supervise it, and use multilog. ("free" log rotation and filtering included). Some sort of program that wasn't supposed to run as a daemon? fghack, supervise, multilog, and you're set. There's a lot of ways you can use those tools, you just have to remember they're there...
I see this coming up every time there's a gmail discussion...
IT'S NOT TRUE!
If you register "foobar", any dotted variation of it is yours, from "foo.bar" to "f.o.o.b.a.r". Likewise, if you register "foo.bar", "foobar" is automatically yours. There are no "other people owning the same address without dots". All of "them" are you.
Having seen the whole thing as well, I disagree. I don't think anyone would dispute it started awfully (well, maybe Berman), but from about the middle of season 3, it started to feel like Trek again.
I really, really liked season 4 (could have lived without the "oh my baby/clone/hybrid/whatever" stuff, though). And the Mirror episodes were great. (however, I always had a thing for the mirror universe, so I'm biased)
How does that make it okay to equate rape with justice?
It doesn't. But as a person who also managed e-mail services for about 8 years, I see where they're coming from. SPAM (and spammers) are amazingly frustrating to deal with, not to mention expensive (especially when you're managing a few million mailboxes in a few thousand domains). Whenever a new SPAM wave gets you off guard, there's a lot of emergency mail pattern analysis, filter writing, sender blocking and similar stuff to do.
If you're too slow or simply unable to filter it out and get to the point where it causes service degradation, you have to deal with the people doing helpdesk, because they got a lot of angry customers wanting to know why their e-mail was delayed. (and others wanting to know why they suddenly got a lot of weird messages). And then there's the marketing people asking why the SPAM filtering capabilities they're touting aren't working. And the sheer labor of going through the mail queues and siphoning out the trash. Oh, and the traditional "you sold my e-mail address to spammers, you bastards!"
For those in non-English countries (such as myself), there's the additional burden of trying to get some customers to comprehend that "no, that's not a legitimate message, there's no exiled Nigerian prince wanting to talk to you and give you a lot of money." and "no, sir, that wasn't a case of mistaken identity, it's just junk." and "well sir, your english-reading skills suck, please believe us when we're telling you to delete and forget about it, it's trash".
Well, first of all, decoding application layer protocol information is too slow to be done on really large border routers.
Agreed. It's perfectly feasible on data-center edge switches, though. Why use 30 public IP addresses in a set of MXs when a single one will do? Why use multiple IPs to split websites by technology (say, one IP for the IIS server that hosts the ASPs, one for the apache that hosts the PHP, and one for the plain-jane HTML hosting) when the switch can look into the URI, see the.asp/.php/.html extension and deliver the connection to the right server? Unless the requirements are ridiculously high, it can be done.
Second, if you do that, all routers will have to know all IP-based protocols (and even many TCP-based ones if they want to support connection initiation in both directions) they want to be able to route, which is basically impossible, kills end-to-end connectivity and (and this point is very important) prevents newcomers/startup companies from freely inventing new, innovative services and protocols on the net.
Again... Agreed. Protocol-based "routing" isn't a magic bullet, and there's a lot of situations where it doesn't apply. However, it _can_ be used to solve a lot of the current address-space waste, especially for simpler and well-established protocols. Webservers, mail servers, and DNS servers come to mind. In ISPs, there's no need to keep DHCP and RADIUS servers, BAS, NAS, or DSLAM in public IP spaces, yet many do so.
OTOH, the major consumption, at least where I work, doesn't come from these areas: it's the increase in customers, and the habit of giving each and every user connection a public IP address; why do UMTS-enabled cell-phones need a public IP? Or GSM-enabled tracking devices? Or even most of the residential-market ADSL / Fiber / dial-up connections? We're getting around 4 new/16s every year (and using them), and we're not that big, so I can easily understand where the quoted 17/8s went to. What I don't understand is why ISPs and telcos keep using public addresses everywhere...
I have deployed similar solutions for SMTP, POP3, and DNS. ssh, not really (but in the case of SSH, you usually want to reach a specific machine, instead of accessing a certain service).
In the case of DNS (resolvers), we split internal domains to a server group, the most common ccTLDs to another specific server group, and the rest of them, which represent lower loads, to a weaker set of servers.
SMTP and POP3 both require the load-balancers to have a minimal knowledge of the protocol. They respond to the initial part of the conversation on their own, up until the point where they identify the correct server farm (by domain in the RCPT TO/USER commands). Once that's been caught, the connection is passed to the final server, and the previous parts of the conversation replayed; from that point on, it's the real server who's answering.
This doesn't really scale well, though, as it requires large (for a switch) amounts of memory. We ended up doing this sort of split in the servers themselves; even then, from the Internet's perspective, it's still a single IP acting as the SMTP server.
if you had specific URIs in that site, or any unique header in the request (say, a specific user-agent) it could still be done. Other than that, just opening the IP address would fail (or it would be sent to a default "unidentifiable site" server)
It's been a long time since I've seen a legitimate website (other than small and usually personal stuff) without a proper hostname, though.
Who modded this "Insightful"? You CAN forward ports to multiple servers, easily. There's plenty of equipment to do that.
Any half-decent load-balancer is minimally L7-aware, to the point of being able to send specific hostnames in HTTP requests to specific servers (or server groups). The ones I primarily use go to the point of allowing me to distribute traffic based on arbitrary headers, cookies, URIs, you name it. Plenty of sites and distinct server farms behind a single public IP address.
I suppose that really depends on the SCM and how the commits are structured or processed. I, for example, have commit access to one of the Linux kernel subsystems; about half the changes I commit, maybe a bit less, are contributed by users, and they appear properly attributed (to the original patch author) in the kernel git changelog and in ohloh.
As for SCM changes, care must be taken during the conversion... We started out with CVS, and are currently using mercurial; when the change was made, the whole history was imported to bootstrap the new repo, and the mercurial tree includes the full changelog from the CVS era; if ohloh started tracking the HG repository, it would get the whole history (it won't, since it looking at the kernel tree, but you get the idea)
I can hear it too (always could), and I'm in my 30s. I can even hear the capacitors in some more high-tech pieces of equipment (like some cheap laptops or portable media players). This was especially "fun" when I had to return a laptop because I couldn't use it for more than 5 minutes without getting a headache, and couldn't demonstrate the problem...
Anyway... During college, I knew at least 2 other people who could hear high-pitched noise from electronic devices at those frequencies, so it's probably not that rare.
It's "norte-americano" (literally "north-american"). Still innacurate (last time I looked, our new Canadian Overlords were located in North America, too)
Five isn't all that much, and neither are multiple laptops...
I have 2 media-centers (in separate divisions), 2 desktops (1 for me, 1 for my wife) and 3 laptops (wife's work laptop, my work laptop, and my personal 12" laptop that follows me everywhere), plus 1 "server" (actually a desktop that's used as the home gateway and file server).
So, overall... that's 5 desktop computers and 3 laptop computers usually inside the house, for 2 people. Not average, I admit, but they're all actually used.
RHEL (like Fedora) does NOT include or support XGL. They support AIGLX, another accelerated desktop mechanism. They do support and ship compiz (the Window Manager that does the cube thingy), though. (compiz works on both AIGLX and XGL)
Think of it like - you don't have to believe in gravity, it's still just there.
Ha! Gravity is only a theory! I, myself, believe in Intelligent Falling towards Heavy Bodies
V4L1 has been marked as deprecated a few months ago, and is currently being phased out. the V4L1 ioctls that still work will probably go away soon.
I do have a version of saa7134-alsa patched to properly mute/unmute, but will be unable to test it until next weekend... Therefore my previous question.
As for your patches, the best place to send them would be the v4l list, second best place the v4l-maintainers list (the address is listed at the kernel MAINTAINERS file.). Since you mention a "he", I assume you're talking about Gerd; it's been a while since he handed over the v4l maintenance to another group of people (of which I am one. modinfo saa7134-alsa, compare to my nickname).
You're right, I completely forgot about that problem... I placed a call to "v4lctl unmute" at my rc.local and never thought about it again. Would you mind rolling back that Myth patch and trying a version of saa7134-alsa that does the mute/unmute itself?
I was actually referring to the ALSA interface, it's been available since 2.6.15. I'm using it just fine with MythTV (via OSS emulation, but it's saa7134-alsa); what problem do you have, exactly?
Slashdot Mirror
The first rule of the tautology club is the first rule of the tautology club :)
And coincidentally, I do have a fair amount of experience with ATMs (probably outdated by now, though). Boot time, at least in my country, isn't a factor: nowadays, most of them actually run a stripped down version of Windows 2000...
That statement shows a severe lack of understanding on IT Technology... The PIN number doesn't need a boot of the whole OS system to be processed, all relevant information should either be in RAM memory, or contained in the card itself...
You're not. I'm also almost blind on one eye (perfect peripheral vision, so no depth perception problems, but forward vision is around 20%), and 3D simply doesn't work for me, regardless of the technology used.
I'm not exactly concerned about exclusion, yet. I've watched the 2D versions of most recent 3D movies (Avatar excluded), and thought they all sucked. The only thing going for them was the 3D gimmick, so most people (my wife included) liked the visuals and felt they got their money's worth; for me, not so much.
I know at least 2 other people with the same condition I have (1 relative, 1 co-worker), and several others with various eyesight problems that prevent them from fully seeing 3D movies.
I doubt (at least hope) we'll ever get to a point where 3D stops being a gimmick and becomes an integral part of the movie (like... needing a 3D view to fully understand a scene), at least until holograms (actual 3D projections instead of visual tricks) come along...
I've just checked my work's logs (an ISP). The number of hits in the spam taggers fell from 12/sec to 3/sec earlier this week.
So either we're identifying less spam, or there is in fact less of it.
How long ago was that?
In the last 5 years, I've had to take the laptop out in most of Europe, including the UK (which you're listing), Belgium, Netherlands, France (there I even had to turn it on! what if I had had a dead battery?), Spain, Portugal, and Italy. I don't remember having to do it in Asia, but wherever I stop in Europe, it's "laptop out, sir".
We've always had bandwidth caps where I live, ADSL included.
The cap for the mobile access my company sells is currently at 25 GB/month, but there are frequent promotions of the "join until the end of August and get unlimited downloads forever" type, so we have a lot of "unlimited" customers. And no, they're not an issue.
There are many European countries selling 3G based "broadband Internet", with dedicated devices (not phones, just simple USB HSDPA modems that take a SIM card).
In those countries, the basestations should be more than capable of handling hogs (either by allowing the traffic, or by throttling abusers so they don't mess up everybody else's connections).
As for tethering... That's the main reason I gave away my iPhone 3G (I work for a mobile operator, got one at launch day, and gave it away to the person in my team that pulled the longest straw. Yes, literally.). Every phone I've had for the last 4 years has allowed me unrestricted bluetooth connectivity to the internet, starting with GPRS and now with HSDPA), and it is something I use *a lot*. I don't have to take my phone out of the pocket, I just fire up the laptop, push a button, and I'm done.
Sorry, but I'll have to disagree on the daemontools issue. Yes, it's a completely different way of thinking, but once you get used to it (and to ucspi-tcp), there's a lot of stuff that can be done with it...
All servers under my direct or indirect control have them installed as a rule (and we're talking a few hundred servers here), and most of them are actively using them.
Need a quick TCP-based daemon for menial tasks (like outputting a status line for monitorization?) tcpserver + supervise. A looping task? supervise that task, and add a sleep at the end of the run script. Don't want to write a logging function for some app? print to stdout/stderr, supervise it, and use multilog. ("free" log rotation and filtering included). Some sort of program that wasn't supposed to run as a daemon? fghack, supervise, multilog, and you're set. There's a lot of ways you can use those tools, you just have to remember they're there...
I see this coming up every time there's a gmail discussion...
IT'S NOT TRUE!
If you register "foobar", any dotted variation of it is yours, from "foo.bar" to "f.o.o.b.a.r". Likewise, if you register "foo.bar", "foobar" is automatically yours. There are no "other people owning the same address without dots". All of "them" are you.
Having seen the whole thing as well, I disagree. I don't think anyone would dispute it started awfully (well, maybe Berman), but from about the middle of season 3, it started to feel like Trek again.
I really, really liked season 4 (could have lived without the "oh my baby/clone/hybrid/whatever" stuff, though). And the Mirror episodes were great. (however, I always had a thing for the mirror universe, so I'm biased)
It doesn't. But as a person who also managed e-mail services for about 8 years, I see where they're coming from. SPAM (and spammers) are amazingly frustrating to deal with, not to mention expensive (especially when you're managing a few million mailboxes in a few thousand domains). Whenever a new SPAM wave gets you off guard, there's a lot of emergency mail pattern analysis, filter writing, sender blocking and similar stuff to do.
If you're too slow or simply unable to filter it out and get to the point where it causes service degradation, you have to deal with the people doing helpdesk, because they got a lot of angry customers wanting to know why their e-mail was delayed. (and others wanting to know why they suddenly got a lot of weird messages). And then there's the marketing people asking why the SPAM filtering capabilities they're touting aren't working. And the sheer labor of going through the mail queues and siphoning out the trash. Oh, and the traditional "you sold my e-mail address to spammers, you bastards!"
For those in non-English countries (such as myself), there's the additional burden of trying to get some customers to comprehend that "no, that's not a legitimate message, there's no exiled Nigerian prince wanting to talk to you and give you a lot of money." and "no, sir, that wasn't a case of mistaken identity, it's just junk." and "well sir, your english-reading skills suck, please believe us when we're telling you to delete and forget about it, it's trash".
Well, first of all, decoding application layer protocol information is too slow to be done on really large border routers.
.asp/.php/.html extension and deliver the connection to the right server? Unless the requirements are ridiculously high, it can be done.
/16s every year (and using them), and we're not that big, so I can easily understand where the quoted 17 /8s went to. What I don't understand is why ISPs and telcos keep using public addresses everywhere...
Agreed. It's perfectly feasible on data-center edge switches, though. Why use 30 public IP addresses in a set of MXs when a single one will do? Why use multiple IPs to split websites by technology (say, one IP for the IIS server that hosts the ASPs, one for the apache that hosts the PHP, and one for the plain-jane HTML hosting) when the switch can look into the URI, see the
Second, if you do that, all routers will have to know all IP-based protocols (and even many TCP-based ones if they want to support connection initiation in both directions) they want to be able to route, which is basically impossible, kills end-to-end connectivity and (and this point is very important) prevents newcomers/startup companies from freely inventing new, innovative services and protocols on the net.
Again... Agreed. Protocol-based "routing" isn't a magic bullet, and there's a lot of situations where it doesn't apply. However, it _can_ be used to solve a lot of the current address-space waste, especially for simpler and well-established protocols. Webservers, mail servers, and DNS servers come to mind. In ISPs, there's no need to keep DHCP and RADIUS servers, BAS, NAS, or DSLAM in public IP spaces, yet many do so.
OTOH, the major consumption, at least where I work, doesn't come from these areas: it's the increase in customers, and the habit of giving each and every user connection a public IP address; why do UMTS-enabled cell-phones need a public IP? Or GSM-enabled tracking devices? Or even most of the residential-market ADSL / Fiber / dial-up connections? We're getting around 4 new
I have deployed similar solutions for SMTP, POP3, and DNS. ssh, not really (but in the case of SSH, you usually want to reach a specific machine, instead of accessing a certain service).
In the case of DNS (resolvers), we split internal domains to a server group, the most common ccTLDs to another specific server group, and the rest of them, which represent lower loads, to a weaker set of servers.
SMTP and POP3 both require the load-balancers to have a minimal knowledge of the protocol. They respond to the initial part of the conversation on their own, up until the point where they identify the correct server farm (by domain in the RCPT TO/USER commands). Once that's been caught, the connection is passed to the final server, and the previous parts of the conversation replayed; from that point on, it's the real server who's answering.
This doesn't really scale well, though, as it requires large (for a switch) amounts of memory. We ended up doing this sort of split in the servers themselves; even then, from the Internet's perspective, it's still a single IP acting as the SMTP server.
if you had specific URIs in that site, or any unique header in the request (say, a specific user-agent) it could still be done. Other than that, just opening the IP address would fail (or it would be sent to a default "unidentifiable site" server)
It's been a long time since I've seen a legitimate website (other than small and usually personal stuff) without a proper hostname, though.
Who modded this "Insightful"? You CAN forward ports to multiple servers, easily. There's plenty of equipment to do that.
Any half-decent load-balancer is minimally L7-aware, to the point of being able to send specific hostnames in HTTP requests to specific servers (or server groups). The ones I primarily use go to the point of allowing me to distribute traffic based on arbitrary headers, cookies, URIs, you name it. Plenty of sites and distinct server farms behind a single public IP address.
I suppose that really depends on the SCM and how the commits are structured or processed.
I, for example, have commit access to one of the Linux kernel subsystems; about half the changes I commit, maybe a bit less, are contributed by users, and they appear properly attributed (to the original patch author) in the kernel git changelog and in ohloh.
As for SCM changes, care must be taken during the conversion... We started out with CVS, and are currently using mercurial; when the change was made, the whole history was imported to bootstrap the new repo, and the mercurial tree includes the full changelog from the CVS era; if ohloh started tracking the HG repository, it would get the whole history (it won't, since it looking at the kernel tree, but you get the idea)
I can hear it too (always could), and I'm in my 30s. I can even hear the capacitors in some more high-tech pieces of equipment (like some cheap laptops or portable media players). This was especially "fun" when I had to return a laptop because I couldn't use it for more than 5 minutes without getting a headache, and couldn't demonstrate the problem...
Anyway... During college, I knew at least 2 other people who could hear high-pitched noise from electronic devices at those frequencies, so it's probably not that rare.
It's "norte-americano" (literally "north-american"). Still innacurate (last time I looked, our new Canadian Overlords were located in North America, too)
Five isn't all that much, and neither are multiple laptops...
I have 2 media-centers (in separate divisions), 2 desktops (1 for me, 1 for my wife) and 3 laptops (wife's work laptop, my work laptop, and my personal 12" laptop that follows me everywhere), plus 1 "server" (actually a desktop that's used as the home gateway and file server).
So, overall... that's 5 desktop computers and 3 laptop computers usually inside the house, for 2 people. Not average, I admit, but they're all actually used.
*grin*
:)
Ok, what did you think the continent Australia is in was called?
RHEL (like Fedora) does NOT include or support XGL. They support AIGLX, another accelerated desktop mechanism. They do support and ship compiz (the Window Manager that does the cube thingy), though. (compiz works on both AIGLX and XGL)
Think of it like - you don't have to believe in gravity, it's still just there. Ha! Gravity is only a theory! I, myself, believe in Intelligent Falling towards Heavy Bodies
OK, by parts:
V4L1 has been marked as deprecated a few months ago, and is currently being phased out. the V4L1 ioctls that still work will probably go away soon.
I do have a version of saa7134-alsa patched to properly mute/unmute, but will be unable to test it until next weekend... Therefore my previous question.
As for your patches, the best place to send them would be the v4l list, second best place the v4l-maintainers list (the address is listed at the kernel MAINTAINERS file.). Since you mention a "he", I assume you're talking about Gerd; it's been a while since he handed over the v4l maintenance to another group of people (of which I am one. modinfo saa7134-alsa, compare to my nickname).
You're right, I completely forgot about that problem... I placed a call to "v4lctl unmute" at my rc.local and never thought about it again. Would you mind rolling back that Myth patch and trying a version of saa7134-alsa that does the mute/unmute itself?
I was actually referring to the ALSA interface, it's been available since 2.6.15. I'm using it just fine with MythTV (via OSS emulation, but it's saa7134-alsa); what problem do you have, exactly?