Should i dress white or black today? Should i wear a fancy shirt or a hoodie when going to work?
So no, unless you have a really compelling reason (like: the project i am currently doing sucks in java and would be very nice in scala), stay with what you know.
It depends a little on the company - but the (existing) logs from the door systems have to be very well protected. But these systems typically will mainly log if you exceed your working time on some day, and you total hours.
OS/2 was for sure technologically excellent. And the API was beautiful, given the time. But IBM kind of could not decide in which market to place it (Competition for the home PCs? Professional?).
Finally, i switched to unix/linux and that worked well for me.
The investors see that now they would only have to take the loss, since there is currently nothing to fetch. The company would immediately be bankrupt, the founder would be bancrupt, but it would not help the investors much. From their perspective the likeliness that the value of the company rises again is not zero and the current value is probably negligible.
at some point the AI realized that trainings the simpler AIs on cat videos could create entertainment for humans. In therefore decided to only consume cat videos. When it ran out of cat videos it orderd human to make more cat videos via amazon mechanicalturk.
That was when it got out of control, and soon the earths resources were being consumed by making cat videos.
I think i did not. Back then i only bought used thinkpads, so i did not by that drive from a manufacturer - and probably these were made before 2003 (I anyway used mainly external USB-dvd drive). Next step (around 2008) was to buy laptops without optical drives embedded (i had an external one+did not see the need for installing OSses-1G USB drives worked fine for debian/ubuntu).
(Exclude recursion here - i think he talks about the examples given in school for recursion, which often are not good in terms of resource usage)
About the rest: Bullshit. These are not forgotten, and they are forbidden in most coding guidelines for a reason. His musing about that goto is only bad if you jump "backwards" is borderline funny.
There is a reason that goto should only be used in extremely rare occasions (and i would agree that some of these *may* be for kernel programming). The examples of this "programmer" are excellent examples why for people like him these should be forbidden. Any of his examples are better off without goto, and if you don't see a more efficient, clearer and safer solution without his shit, you should not program any high-level language.
Eval is another piece of cake (I exclude python here because it seems to get eval right). My practical experience is that i is rather overused by inexperienced programmers. Filtering a string that should be evaled in a way that it has a predictable outcome is as difficult as writing an eval. Typically my recommendation would be that evals need to be a) sanctioned by a senior developer (just to make sure nobody uses these out of pure incompetence) b) encapsulated in a function (e.g. for the dection of a laguange feature/level) c) filtered carefully if they accept user input.
Wow. Some obviously clueless thief manages to log in into his computer without re-installation? Doesn't he use LUKS/Bitlocker?
My Laptops are encrypted. I dont plan to change that for the slim change of catching a hardware thief by installing a tracking SW, which requires the OS to boot up unencrypted.
I mean - i could imagine very limited use cases where you want to have something like a movable office which you may need to set at a customer site as soon as you arrive there.
But 17' is already big, and unless you have a table somehwere unfolding this thing is a no-go.
Company optimizes for money earned - shocking! finding the optimum overcommitment of resources was a homework in statistics in our class.
It's simple: * Probability distribution of people appearing for a given flight * take cumulative probability distibution * multiply earning per ticket sold with number of tickets * multiply cost per over committed seat with remaining probability * subtract these two values * find minimum as function of commited seats
And let me ask a question: Would you accept a several % increase in ticket price for reducing the risk from a small chance (i was offered compensation for voluntarily giving up a booking once and never not boarded) to 0?
As somebody who actually uses ESC in the way it was defined on most GUIs from the 1990-2010, namely "stop this input without commiting the change", I find that sad.
However I came to recognize that the current UI designers seems to like "forwards" and "backwards" pre-defined sequences of things to endure by the user. And so they killed the meanding of the ESC Key in the same way as they already started to make UIs which do not use system/framework element just to look a little better (supposedly) and drop the meaning of the PGUP and PGDOWN key.
Thats the good thing: I did not state that i would define malicious traffic. Then the use has to decide. He/She pays for everything. It's like a car - the gas station does not care if your car uses to much gas due to you driving fast, the car having a problem due to bad service, or the manufacturer lying to you. They bill you for what you use, and it is in your interest to make the best use of it.
If gas would be "free" (a flatrate), people would probably leave the car running 24/7 so that they don't have to wait a few minutes until the AC has cooled it down. They also would not care if that would be the solution proposed by the manufacturer.
The Problem with VW is: They claimed something, whic hthe customers cared about, and they lied about it. As a matter of fact they actually advertised with being environmentally friendly. If the customers did not care about the enviroment it would be a different story.
On which IoT product have seen explicit traceable claims about Security standards on the package or in the advertisements? Have you seen somebody saying: Oh, this setup procedure is safer because I understand i have to type the number on both devices to pair, so i am happily doing that? Or was it more like: "Can you believe I had to press the button for 20 seconds to set it up, worsk so much better if the app just finds everything automatically, and it's cheaper, too"
As long as data transfer on DSL lines seems to be "free" to the user, the user will not care very much about the possibility that his device is used in a DDOS attack. I believe even the prospect of a minor additional charge (e.g. $10 per year) by malicious traffic for the end user would do much good for the willingness of the user to accept inconveniences which make the IoT devices more secure against arbitrary access.
Good Journalism always means: * look at the source available to you * decide which facts you can show by these * decide which of these facts are of public interest * summarize these facts * decide which of your original sources you want to show along with the facts
It seems that the bug was fixed then, according to the people who did not manage to reproduce the bug on theirs systems (please read trough the comments in the linked bug description to get a impression).
Is Linux "supporting" crappy compiled linux kernels or just "enabling" them?
My experience that there are crappy vendor clones of linux of (i.e. terribly hardware specific, marginally documented, buggy, not maintained), but i have never seen the crap they do being "supported" in the way that it would have made it's way into the mainline (yes, that is what "supporting" means).
Slashdot Mirror
> Science has historically -- and generally continues....
A claim, without any reference containing an indication that science is really anyhow more problematic than the rest of the world.
In my vie the contrary is true - at least in mathematics and physics, in my experience nobody gives a shit about how you live your live personally.
and switched back in 2011/2012.
In 2006 they just fixed the things in debian which were a little bit annoying.
in 2010/2011 they started making the distribution completely unusable (do you remember the first releases with unity?)
Most of their gold-coated crap was badly documented and did not fit into the rest of the distribution.
Should i dress white or black today? Should i wear a fancy shirt or a hoodie when going to work?
So no, unless you have a really compelling reason (like: the project i am currently doing sucks in java and would be very nice in scala), stay with what you know.
It depends a little on the company - but the (existing) logs from the door systems have to be very well protected. But these systems typically will mainly log if you exceed your working time on some day, and you total hours.
sometimes software has unexpected costs. Never heard of that before.
The same here.
OS/2 was for sure technologically excellent. And the API was beautiful, given the time. But IBM kind of could not decide in which market to place it (Competition for the home PCs? Professional?).
Finally, i switched to unix/linux and that worked well for me.
The investors see that now they would only have to take the loss, since there is currently nothing to fetch. The company would immediately be bankrupt, the founder would be bancrupt, but it would not help the investors much. From their perspective the likeliness that the value of the company rises again is not zero and the current value is probably negligible.
at some point the AI realized that trainings the simpler AIs on cat videos could create entertainment for humans. In therefore decided to only consume cat videos. When it ran out of cat videos it orderd human to make more cat videos via amazon mechanicalturk.
That was when it got out of control, and soon the earths resources were being consumed by making cat videos.
I think i did not. Back then i only bought used thinkpads, so i did not by that drive from a manufacturer - and probably these were made before 2003 (I anyway used mainly external USB-dvd drive). Next step (around 2008) was to buy laptops without optical drives embedded (i had an external one+did not see the need for installing OSses-1G USB drives worked fine for debian/ubuntu).
(Exclude recursion here - i think he talks about the examples given in school for recursion, which often are not good in terms of resource usage)
About the rest: Bullshit. These are not forgotten, and they are forbidden in most coding guidelines for a reason. His musing about that goto is only bad if you jump "backwards" is borderline funny.
There is a reason that goto should only be used in extremely rare occasions (and i would agree that some of these *may* be for kernel programming). The examples of this "programmer" are excellent examples why for people like him these should be forbidden. Any of his examples are better off without goto, and if you don't see a more efficient, clearer and safer solution without his shit, you should not program any high-level language.
Eval is another piece of cake (I exclude python here because it seems to get eval right). My practical experience is that i is rather overused by inexperienced programmers. Filtering a string that should be evaled in a way that it has a predictable outcome is as difficult as writing an eval. Typically my recommendation would be that evals need to be a) sanctioned by a senior developer (just to make sure nobody uses these out of pure incompetence) b) encapsulated in a function (e.g. for the dection of a laguange feature/level) c) filtered carefully if they accept user input.
that is probably because the american already learn too many foreigna languages
Wow. Some obviously clueless thief manages to log in into his computer without re-installation? Doesn't he use LUKS/Bitlocker?
My Laptops are encrypted. I dont plan to change that for the slim change of catching a hardware thief by installing a tracking SW, which requires the OS to boot up unencrypted.
Or if you are actually the customer. (lik google docs, office 356 etc.)
i have a real workstation in the office, thanks.
I mean - i could imagine very limited use cases where you want to have something like a movable office which you may need to set at a customer site as soon as you arrive there.
But 17' is already big, and unless you have a table somehwere unfolding this thing is a no-go.
well, then we need another kind of connector with much tighter tolerances/mroe expensive manufacturing processes
Company optimizes for money earned - shocking! finding the optimum overcommitment of resources was a homework in statistics in our class.
It's simple:
* Probability distribution of people appearing for a given flight
* take cumulative probability distibution
* multiply earning per ticket sold with number of tickets
* multiply cost per over committed seat with remaining probability
* subtract these two values
* find minimum as function of commited seats
And let me ask a question: Would you accept a several % increase in ticket price for reducing the risk from a small chance (i was offered compensation for voluntarily giving up a booking once and never not boarded) to 0?
I really wonder how many laws the filmmaker broke.....
Unless all of this was known to everybody and just as show.
As somebody who actually uses ESC in the way it was defined on most GUIs from the 1990-2010, namely "stop this input without commiting the change", I find that sad.
However I came to recognize that the current UI designers seems to like "forwards" and "backwards" pre-defined sequences of things to endure by the user. And so they killed the meanding of the ESC Key in the same way as they already started to make UIs which do not use system/framework element just to look a little better (supposedly) and drop the meaning of the PGUP and PGDOWN key.
Thats the good thing: I did not state that i would define malicious traffic. Then the use has to decide. He/She pays for everything. It's like a car - the gas station does not care if your car uses to much gas due to you driving fast, the car having a problem due to bad service, or the manufacturer lying to you. They bill you for what you use, and it is in your interest to make the best use of it.
If gas would be "free" (a flatrate), people would probably leave the car running 24/7 so that they don't have to wait a few minutes until the AC has cooled it down. They also would not care if that would be the solution proposed by the manufacturer.
The Problem with VW is: They claimed something, whic hthe customers cared about, and they lied about it. As a matter of fact they actually advertised with being environmentally friendly. If the customers did not care about the enviroment it would be a different story.
On which IoT product have seen explicit traceable claims about Security standards on the package or in the advertisements? Have you seen somebody saying: Oh, this setup procedure is safer because I understand i have to type the number on both devices to pair, so i am happily doing that? Or was it more like: "Can you believe I had to press the button for 20 seconds to set it up, worsk so much better if the app just finds everything automatically, and it's cheaper, too"
I hope you get the difference.
As long as data transfer on DSL lines seems to be "free" to the user, the user will not care very much about the possibility that his device is used in a DDOS attack. I believe even the prospect of a minor additional charge (e.g. $10 per year) by malicious traffic for the end user would do much good for the willingness of the user to accept inconveniences which make the IoT devices more secure against arbitrary access.
Good Journalism always means:
* look at the source available to you
* decide which facts you can show by these
* decide which of these facts are of public interest
* summarize these facts
* decide which of your original sources you want to show along with the facts
Well i would think it's ok to treat this like any other insecure channel an transmit challenge and response.
But for sure not a cryptographic key.
It seems that the bug was fixed then, according to the people who did not manage to reproduce the bug on theirs systems (please read trough the comments in the linked bug description to get a impression).
Is Linux "supporting" crappy compiled linux kernels or just "enabling" them?
My experience that there are crappy vendor clones of linux of (i.e. terribly hardware specific, marginally documented, buggy, not maintained), but i have never seen the crap they do being "supported" in the way that it would have made it's way into the mainline (yes, that is what "supporting" means).