Linguistic prescription is dead. Language means whatever people agree it means. You get extra points for arguing AGAINST linguistic prescription in your signature though.
If I make an app where I can throw birds at pigs that happens to use a GPL'ed JSON library, it doesn't mean that the whole app has to be open-sourced does it?
According to the GPL, the answer is yes, you need to make the whole app Free Software. In terms of copyright law it is a bit more of a gray area, so if you never distribute the JSON library or statically link with it, you can possibly get away with it, in at least some jurisdictions. You need to be able to say that you never accepted the GPL and therefore any distribution of the GPL'd code is right out. Even that may not be enough, but there is AFAIK no clear case law yet.
Spammers don't give a shit about US laws because most of them don't live or operate here anyways.
10 worst spam havens. Currently the US in first place is almost 3 times as bad as second place China. Even worse, a lot of the spam coming out of China is sent on behalf of spammers residing in the US.
As to the RouterBoard, mentioned by RichiH, I doubt that consists of an out-of-the-box solution for most people.
As long as you get the RB750 or the RB750G, it certainly counts as an out-of-the-box solution. It comes default with DHCP on the WAN interface, the rest bridged as LAN, DHCP server on the LAN, and IPv4 NAT. No assembly or configuration required.
I have no idea which IPv6 features are enabled by default though.
You can solve TSP for 1 million cities if you're willing to wait a few billion years
No you can't. You can't solve TSP for 1 million cities before the expected heat death of the universe. You can't even solve TSP for 500 cities before the expected heat death of the universe (assuming you can do less than 10^100 instructions per second).
The problem with switches is that most switches have not merely small buffers, which would be ok, but microscopic ones. E.g. Cisco 3560G loses traffic on a gigabit port when faced with 50Mbps of bursty traffic in total coming from two ports. 10ms of buffer at 1Gbps is ~1MB, and most switches have nothing near that per port.
The 10-prefixes are falling out of use. There are a few combinations which survive. Cm, dl, and dB are probably the only ones with real life left in them. Otherwise it's 1000-prefixes only.
Please don't do this unless you really need to. It may only cost you $100 a year, but TCAM space in most core routers is quite limited. If you announce a PI route, you take up a spot in practically EVERY core router on the entire Internet. You will likely announce an IPv6 route as well, and those take up even more resources.
Re:Another great Python 3.x series release
on
Python 3.2 Released
·
· Score: 2
I write something in Java, I know it's gonna work in 5 years
More like you write something in Java, and it breaks with equal probability in an update from 1.6.0.0 to 1.6.0.1 as in an update from 1.6 to 1.7... At least that is my experience as a user of Java Applets.
Server-side this does not matter much because few people upgrade Java anyway.
IIRC it was reasonably common advice in Unix books from the 80's (I can't provide citations because I borrowed from the library). Reboot at least weekly with a full fsck. Supposedly file systems weren't as stable back then.
Since the myth was well and truly dead by the time I managed to touch a Unix box for the first time (1993), it seems a bit late to try to kill it.
In theory we can develop anything. In practice, we are far off from being able to send anything to even the nearest stars with a travel time shorter than thousands of years. It is unlikely that we can make anything which can last thousands of years and still be functional enough to find a planet and brake to a reasonable speed before landing.
We cannot really send containers anywhere useful except Mars, Venus, and Europa. Neither of those are particularly good hopes for sustaining intelligent life one day. Except Venus perhaps, but that is going to take more than some microbes.
Half an hour without redundancy is no fun. Especially if it was an "unscheduled service window" and you are hoping that the secondary won't fail too.
From experience working with critical systems, a low boot time is very helpful when things are not working as they should and the stress level is high.
Boot time is proportional to system cost. This is fairly stupid by the way, since costly systems tend to be critical systems where even service windows should be kept short, but I know of no vendor who cares.
There are a lot of very low hanging fruit in this area, but no vendors care, and I dare not go with coreboot.
Slashdot Mirror
Your statement correctly applies to the median.
Not so! Consider the set of integers {1, 1, 1, 37}. How many of these are below the median? Sorry for nitpicking.
And sorry for quote-fucking, it was just too tempting.
We do not do enough, but we can still applaud that we are actually doing something.
Emacs is a nice OS, but it lacks a decent browser.....
An OS with cooperative multitasking didn't cut it in 2000, and certainly doesn't cut it now.
These days much of the real value from standards is the process which creates them. The finished standard is almost just a diploma to put on the wall.
If we can find a superior way to get competitors working together, perhaps standards can just go away.
Linguistic prescription is dead. Language means whatever people agree it means. You get extra points for arguing AGAINST linguistic prescription in your signature though.
If I make an app where I can throw birds at pigs that happens to use a GPL'ed JSON library, it doesn't mean that the whole app has to be open-sourced does it?
According to the GPL, the answer is yes, you need to make the whole app Free Software. In terms of copyright law it is a bit more of a gray area, so if you never distribute the JSON library or statically link with it, you can possibly get away with it, in at least some jurisdictions. You need to be able to say that you never accepted the GPL and therefore any distribution of the GPL'd code is right out. Even that may not be enough, but there is AFAIK no clear case law yet.
Spammers don't give a shit about US laws because most of them don't live or operate here anyways.
10 worst spam havens. Currently the US in first place is almost 3 times as bad as second place China. Even worse, a lot of the spam coming out of China is sent on behalf of spammers residing in the US.
As to the RouterBoard, mentioned by RichiH, I doubt that consists of an out-of-the-box solution for most people.
As long as you get the RB750 or the RB750G, it certainly counts as an out-of-the-box solution. It comes default with DHCP on the WAN interface, the rest bridged as LAN, DHCP server on the LAN, and IPv4 NAT. No assembly or configuration required.
I have no idea which IPv6 features are enabled by default though.
If I get to put all the cities on a straight line, I can solve TSP in linear time. Of course you can solve special cases fast.
You can solve TSP for 1 million cities if you're willing to wait a few billion years
No you can't. You can't solve TSP for 1 million cities before the expected heat death of the universe. You can't even solve TSP for 500 cities before the expected heat death of the universe (assuming you can do less than 10^100 instructions per second).
You would need 60km/s delta-v. Currently infeasible.
The problem with switches is that most switches have not merely small buffers, which would be ok, but microscopic ones. E.g. Cisco 3560G loses traffic on a gigabit port when faced with 50Mbps of bursty traffic in total coming from two ports. 10ms of buffer at 1Gbps is ~1MB, and most switches have nothing near that per port.
The 10-prefixes are falling out of use. There are a few combinations which survive. Cm, dl, and dB are probably the only ones with real life left in them. Otherwise it's 1000-prefixes only.
Please don't do this unless you really need to. It may only cost you $100 a year, but TCAM space in most core routers is quite limited. If you announce a PI route, you take up a spot in practically EVERY core router on the entire Internet. You will likely announce an IPv6 route as well, and those take up even more resources.
I write something in Java, I know it's gonna work in 5 years
More like you write something in Java, and it breaks with equal probability in an update from 1.6.0.0 to 1.6.0.1 as in an update from 1.6 to 1.7... At least that is my experience as a user of Java Applets.
Server-side this does not matter much because few people upgrade Java anyway.
IIRC it was reasonably common advice in Unix books from the 80's (I can't provide citations because I borrowed from the library). Reboot at least weekly with a full fsck. Supposedly file systems weren't as stable back then.
Since the myth was well and truly dead by the time I managed to touch a Unix box for the first time (1993), it seems a bit late to try to kill it.
In theory we can develop anything. In practice, we are far off from being able to send anything to even the nearest stars with a travel time shorter than thousands of years. It is unlikely that we can make anything which can last thousands of years and still be functional enough to find a planet and brake to a reasonable speed before landing.
We cannot really send containers anywhere useful except Mars, Venus, and Europa. Neither of those are particularly good hopes for sustaining intelligent life one day. Except Venus perhaps, but that is going to take more than some microbes.
A modern cpu will just translate the x86 instructions into risc internally.
That is what microcoded CPU's have been doing all the time. The main point of RISC is to avoid that translation layer.
Plants are actually vastly more efficient than any solar process we've got right now for collecting solar energy
This is not true. The best plants can do less than 9% solar-to-sugar conversion. A 9% efficiency for a solar cell is laughable.
There is nowhere near enough left-overs from crops to fuel current cars, and then you have all the other things we use oil for.
Sure there is: Most priests have doubts about their beliefs at least once.
Half an hour without redundancy is no fun. Especially if it was an "unscheduled service window" and you are hoping that the secondary won't fail too.
From experience working with critical systems, a low boot time is very helpful when things are not working as they should and the stress level is high.
Boot time is proportional to system cost. This is fairly stupid by the way, since costly systems tend to be critical systems where even service windows should be kept short, but I know of no vendor who cares.
There are a lot of very low hanging fruit in this area, but no vendors care, and I dare not go with coreboot.
Careful with using modern vi for binary files, it isn't safe anymore! vim needs the -b option.
Even EMACS does not break binary files...
With the audience of this site, it wouldn't surprise me if Lynx is a test case when the design is modified.
With the history of this site, I would be greatly surprised if it involves "design", never mind "test cases".