I can't tell whether the writer means that incident "X" did not make the landing hazardous,
In the F-15's case that the former :
- The plane has body lift too. Thus there are some safety margins (even with missing bits it can generate enough lift).
- The onboard avionics(*) are able to compensate for quite a lot of situations.
So even with a wing missing, although it couldn't probably perform complex acrobatic maneuvers, could still land safely provided that the pilot is experienced and know how to handle the plane too.
TL;DR: brilliant pilot + bad-ass airplane = can still fly "almost normally" even with missing bits.
Just don't try this it at home, you have drastically reduced quite some safety margins.
---
I've read somewhere that Airbus has some software routine in their avionics that are able to compensate for missing / damamged tail bits. I'll have to track down the reference.
On Firefox 60.0.1 (my distro (OpenSUSE Tumbleweed) package) :
- By default, it isn't checked.
- But right click, and toggle "Check Spelling", and it gets checked by the internal spell checker of Firefox (Spellbound)
- That still doesn't trigger any proof-reader/grammar (such as LanguageTool(*) or Grammarly), thus you're still not covered against homophones or agreements/declension (depending on your language. Here it's English, so only the former).
---
(*) In my opinion, a much better plug-in than grammarly :
- It's open-source
- You can also run a daemon locally.
- So you don't need to necessarily pipe all your text fields into some untrusted 3rd party web server.
But aren't the things which are banned from esports *macro programmable* ? As in, you program a whole complete sequence of button press commands (preferrably a complex or difficult one) to a single physical trigger ? (e.g.: whenever I push button W5, please replay the complete Konami code, followed by that difficult special move from Zangieff on Street Fighter II)
This thing is "simply" a *remapable* controller : "left foot pedal" = "button B" (because the disabled player is hemiplegic and lacks control on the right side of the body, and thus can't push buttons with the right thumb and thus could get past the first pit in a Mario platformer).
If those computers have two-button mice, then replacing the mice with one-button mice might work.
For an extremely long time on PCs : "Main button" : gives you a "left"-click "2nd button" OR "main button" + "shift" : gives you a "right"-click "3rd button" OR "2nd button + main button" OR "main button" + "ctrl" : gives you a "middle" click. (I even have manuals of DOS games explaining this).
I saddly don't have a single button mouse handy to test if that still works in 2018 OSes. But the "two button" for "middle click" definitely works on my laptop on Linux.
On mac : Most modern single bouton Apple mice have capacitive surface and will emit a click depending on either the finger position or number of fingers.
So the student could have some way around input device screw-up. (It takes just one kid to discover it, then after half an hour, the whole school knows the trick).
The best strategy would be to NOT put the sentence in a <TEXTAREA> to begin with. (avoid triggering the spell checker all together). They should instead have had a first round to point to errors by clicking words.
and only ask for correct spelling in a second round of the exercices.
WARNING!!! If you are tempted to try to see how they see the world and are tempted to turn of the ad-blocking in your browser off, be prepared : have a bottle of your favorite eye-bleach within reach, and have the number of a psychological support ready on fast dial on your phone.
We're/. geeks, we still see the "old Microsoft" whenever we talk about that company.
The parent poster isn't refering to the recent gaming business of MS. The comment was about the numerous time Bill Gate was on stage / on live TV trying to demo some Microsoft software product, and the product blew him in the face with a crash/BSOD/whatever. You can say tons of bad things about Microsoft back-then (with EEE appreaing very oftenin the discussions), but at least they weren't faking their demo.
1.) Run your own little server behind your DSL Modem. Starts at $30, called Raspberry PI.
A littlebit under powered, but still good for many non intensive use. (Been there, still doing that sometimes).
2.) Store your data on your private server. Control access/encipher via SSH/SCP.
Oh, common. It's 2018. at least use rsync. (Also, if you want good resilience against damage/corruption, you'd want a base with a tad bit more sata ports, and running RAID5. Just saying)
3.) Run your own discussion forums.
...but be prepared to not have the richness of interactions currently found on the Social Networking Site du jour/whatever Zuck has bought the most recently.
You'll basically have your mom' writing you a message, every now and then, to remind you to call her.
4.) Run Linux on your PC instead of the Snoop-Ware of MSFT.
But be prepared to spend time on stack exchange just to learn how to make your touch pad work* ~~~
5.) Run your own Jabber server instead of Corporate-Controlled Whatsapp and the like.
And you'll have exactly 1 single user registered on it : you.
Because most of the people you know won't bother installing yet another client (but it's opensource) just to send messages to you, when all their friends are on Skype/FBMessenger/WhatsApp/Snapchat/whatever is popular next (hint: look at what Zuck is trying to buy next).
---
* - Disclaimer: Proudly written from a Linux running laptop
Google and Youtube already "nudge users into alignment with their goals"
In that context, *their goals* = Google's goals. You know the "feelgood" piece at the end about solving poverty, disease and world peace. So Google can justify what they're doing by deluding themselves it will lead to "greater good".
by manipulating search results, pushing sites/producers with opinions they prefer and hiding those they disagree with.
in your context *they prefer* and *they disagree* are the users : - algorithms are optimized for one single target : bring more clicks in (because that's what makes more money going in by providing more eyeballs to sell to advertisers) - and the machine "learning/AI/NN/whatever is powering google now" automatically learns that the best way to achieve that is to keep the users in their bubble, preferably with even more extreme and emotion-appealing content.
end result: - if you're a tiny bit left leaning you'd get endless stream of video about social injustice. - if you're a tiny bit on the right you'd get a huge barrage of video about all the ultra-violent rapist-thieving-thugs immigrant that are about to invade the country and it all ends up spiraling toward conspiracy theory revealing to you that the government is lying by making you think the earth is round and hiding the fact that it's actually flat... (because ?...)
TL;DR: the current think that might look like censorship is just a incidental side-effect of the algorithm trying to give you whatever make you stay for more content, and makes you a better recipient for ads. There's no conscious censorship going on.
All my mission critical windows systems are now quarantined from the internet and other networks
Which by itself is a good approach as it drastically decrease *nearly all* security and safety risk for mission critical equipement. (Basically only a few very advanced threats geared toward air gapped targets as stuxnet) No mater the insanity of Windows 10 updates.
The next step would be to enclose them inside VMs and use a slightly saner host OS (something unix-y, e.g.: Linux) to handle the VMs. Including snapshotting for safe roll back (which among other could help mitigate the buggy updates) While of course still keeping them network isolated (they are mission critical after all).
If your Samsung is 840 or later, you can get its own measure of temperature out of SMART attribute "190 Airflow_Temperature_Cel". (I just don't know how to do this in Windows 10. I mostly use Linux and smartmontools. Maybe Samsung's Wizard tools can do it ? Maybe SpeedFan ? Have a look here.)
On Samsung 830, I haven't seen such an attribute.
If that unknown sensors is nearby the SSD, it and the SSD's sensor should evolve similarly over time (spike at roughly the same time, with more or less similar top temperatures)
The results of this review indicate detrimental health consequences associated with high level occupational physical activity in men, even when adjusting for relevant factors (such as leisure time physical activity). These findings suggest that research and physical activity guidelines may differentiate between occupational and leisure time physical activity.
They directly state that they have observed "association". They are only suggesting that physical activity guideline should take into account leisure vs. occupation.
Plus, the hypothesis evoked here on/. (exposure to toxic chemimcals) and the hypothesis from the summary (more trivially, less rest whenever wanted) both boil down to "in occupation, the physical exercie cannot be done while removing negatively implacting factors, unlike when doing it as a leisure".
Answer trashed by miss-typing, I'll try to rewrite my reply again.
Brypt is indeed designed as a *password* hashing function, so it's better for passwords than sha-2 is. I think sha-2 is also acceptable.
Bcrypt, Scrypt, Argon2 (and the older PBKDF2, for completeness) were all designed for password and are designed with resistance to bruteforce in mind. sha{n} family were designed for speed (and sha-3/keccak also because a sponge function is an interesting new concept) and are only acceptable if you don't mind brute forcing.
Definitely don't use a common password. Using "password" as your password will suck no matter what else you do. Therefore it's a mistake to say: Doing X won't help if the passwords suck. Therefore don't do X.
The problem isn't simply people using "password" or "123456" for password. The problem is general approach to security. We human suck at picking strong purely random password.
We/. geeks, will probably pipe a decent (more than 256) amount from/dev/random (not even the 'u' one! with several true random source mixed into the kernel entropy pool !!!) through base64, and use the output for only 1 specific site, and store it into a decent opensource password manager, that backs up on our own cloud. We also activate two-factor authentication whenever possible (using an OTP app, not insecure channels like SMS). That thing is pretty much secure. Even with a fast hash, it's not realistic to brute force. And in case of a database leak, an attacker wouldn't gain much, until the leak is discovered, they'll only could try to log into that specific site and they'll still will be missing the 2FA.
Normal people don't do that. They seldom activate two-factor authentication (it's too cumbersome). And when asked to pick a password, they tend to follow pattern. Most frequently, they will put the required capital letter at the beginning, use a 5-6 letter combo, followed by 2-4 of the required digit, and put an "!" at the end to the mandatory punctuation sign. Such a password follows the most stringent rules nearly everywhere. It will show a green light on most password strength evaluation. But do the math : such a password, in best case (6 letter, 4 digits. The position of the capitalisation at the beginning and the "!" at the end don't bring much as they are predictable) has less than 42 bits of entropy. Collect all the other common patterns, allow for a few extra substitution, and you'd probably still be within 56bits, something that isn't considered secure at all nowadays. Using these pattern, if a database using only slat+fast hashes is leaked, you can recover a very sizeable fraction of all the password within. (Not all password. There's bound to be two or three of/. geeks crowd using/dev/random. But almost all the others).
Worse, lots of people have a tendency to re-use passwords (out of convenience, it's hard to remember "good" passwords). Once a database is leaked and passwords recovered, a sizeable amount of these password could be used to open other stuff (Access the e-mail account associated with that user's database record ? (and once logged in, try to use the "reset password" function of any website associated with this e-mail ?) Try to log into several social network website using the same e-mail and password ? etc.)
In short, it's not only people using "password" and "123456", it's every body who is not using/dev/[u]random (the biggest fraction of your userbase)
Because we know weak passwords will be weak no matter what you do, it probably makes sense to try to make the system as secure as possible +for users who use good passwords+.
On the other hand because we know very few users will use purely random password + a password manager, and activate two-factor authentication, it makes sense to try to make a potential database leak as resistant to brute-forcing as possible. Specially because it only costs us a few hundreds of milliseconds and a little bit more RAM once at the beginning of a session, during the log-in phase.
Maybe if they can get Windows 10 to run Android apps, like ChromeOS, they might stand a chance.
Well, that was the whole initial purpose of WSL / BashOnWindows. Except it's currently still limited to file I/O and some network sockets. So they pivoted WSL into the current "tool for devs to run and test linux dirsto's userspace", and Windows 10 Mobile is still without any access to the 2 main relevant app ecosystem.
Meaning that due to networking effect, their platform is more or less useless, except for the couple of weird user who are only interested into phone functionality and are happy with the one or two apps that got ported on the platform. So basically all the former Zune users. All five of them.
So basically they've been getting a taste of their own monopoly/network effect that they've used on the desktop.
Hey Microsoft, how's being on the receiving end of your own bullcrap ?
why don't you just say that this poor system will never be able to run well
Assuming* it's using x86 and not ARM, you could always install Linux on it to have it run well:-P
--- *: and thus UEFI Secure Boot must mandatory enable the end-user to boot into something else than the microsoft-signed windows. (by disable secure boot and/or by enabling the user to add any extra signing key).
Yes, I get your point. - Parent poster points about rainbow table (tables that point hashes back to strings that can generate the same hash). - You point that a well designed (=non borked design) hashing function should give two different hash for two dissimilar short passwords. Thus you would need a giant rainbow table that gives a password for *every single possible 160bit hash* (that's ~10^48 entries, i.e.: within an order or two from the number of atoms on earth). Fat chance.
BUT! Even if the hash->password direction is *hard*. password->hash direction is easy.
All the algorithm you mention (MD5, SHA1, SHA2 and let's throw SHA3/KECCAK in the mix too) are all extremely *fast hasing functions* (They are ultra fast, and have very low ressource requirement by design). Meaning you can take a GPU running a special compute shader/OpenCL/Cuda code that can process millions of them in a second. So you could scan through ALL the common password (based on frequent leaked passwords and/or on frequent paterns, etc. and their substitutions) within a reasonable time until you find a match. As the summary points out, we humans are bad at picking-up password, we definitely use less than 2^ ${whatever bits used by current popular hash} different passwords. Even if you use salt (so your hash doesn't match any other precedent hash in any rainbow table), and even if you use the latest *hashing* function (SHA3 - well okay, it's a sponge function, but basically works the same), it's definitely within the reach of a reasonable budget to loan GPU compute time on the cloud and brute force the passwords.
So if a database containing SHA{n}-hashed (and optionally salted) passwords get leaked, you can consider that all except the most unusual passwords can be brute-forced.
So in short DO NOT USE HASHES. USE KEY DERIVATION FUNCTIONS.
Things like bcrypt, scrypt or the current competition winner argon2, are on purpose designed to be slow and resource intensive. (By iterating multiple rounds, by require significant memory, etc.) For you, it doesn't change much if loging in take now a third of a second - you only log once, after all, it won't kill you to wait for 300ms just once at the beginning of your session. But for potential brute-forcers, not being able to quickly go through million of tests is suddenly a huge blocker.
So in short : do NOT use SHA2 for your password database. use bcrypt/scrypt/argon2 instead.
I immediately though "well, basically static vs. dynamic friction", too.
On the other hand, electric motors are specially good at having decent torque at very low speed. (You don't need to shift a different gear to start them, unlike ICE. They use a fixed transmission ratio). So it's about the best type of motors you could be using for "just pulling".
If this is the top speed,
I suspect that's more for braking safety. If anything goes wrong, you'll need to brake. - there's only so much kinetic energy that you can shed with the car's brakes/regenerative braking. - there's the reaction time of the guys in the cockpit hitting the landing gear's brakes, whenever alerted over comms about something out of their field of view.
That could be an explanation for the czech being among the most indoor people : historically a former eastern bloc country, with some country regions having significant industrial development (coal mining, iron smelting, etc.) but not so much pollution control. thus people will percieve the outdoor air being more polluted and will develop the habit of staying indoor. (among tons of other different cultural reasons).
America has minimal government interference in transactions between consenting adults,
Except when it concerns what's going in the bedroom (sexual orientation, sex workers, etc.) Then suddenly it's the government business to interfere legally as much as possible.
which is proper.
The idea behind the various European governments is to balance long term risk and costs. The thing to which the adults might consenting could come with tons of long-term risk. Health (both physical and mental) and safety risks, that the public healthcare system could end-up paying. By putting some limitations on health and safety hazards, the government is limiting the money that they would have to chip-in in the long-term.
Better make sure to have a healthy population that has enough opportunity for out-door activities, rest, healthy diet, etc. than having to support tons of people with burn-outs, depression, work-place accidents, work-place chronic disease, obesity by being on constant fast-food diet, etc.
Oh, yeah. I forgot. That requires to actually something resembling a public health-care system.
(Note that the same idea goes regarding drugs or prostitution. People will be doing it anyway, so it better be done within normal, legal (and tax paying) respectable business or self-employment, rather than enriching criminal gangs and putting a high unnecessary burden on the legal and prison systems).
If you want extra vacation rather than higher pay, then that should be between you and your employer, not something imposed on every worker by the government.
And when you ask, you're employer will say no. If you try to shop around you'll find most other employer are think why should *they* if nobody else is doing it.
Thinking of leveraging your exclence/skills/competencies/relevance as a key employee? Haha... The employer would dream to replace you with someone less experienced, less competent, but incredibly cheaper, and then give himselve a bonus to celebrate the decreased costs with absolutely no long-term thinking.
Seems to us European that in the US, the negotiations between employee and corporations is strongly dis-balanced in favour of the later. But yeah, unions, more social-leaning polical parties and anything that could work a little bit more toward better situations for employ is "evil communism" and should be fought off.
Solar and wind are network connections, the former giving a relatively fixed amount of data over the day which changes by the hour, and the latter shoving random amounts of data down the pipe.
So shared connection over TV-cable network vs. 3G connection in an area with bad coverage, resp. ?
No, cgroups & co are there and functional. Namespacing, what I think you really meant, is not yet globally enabled, so no containers.
Namespacing is what I meant by "& co" - i.e.: and all the other kernel bits that make container isolation possible but written shorter.
Yes, this is because as a guest,
No, it's not a "guest". There's no virtualization going on, it's not Hyper-V based. It's supposed to be just a different set of API made available by the NT kernel, in addition to win32. The same way a OS/2-like API was available a long time ago.
it's not supposed to be interfacing directly with hardware, this includes block devices.
WSL's "linux-like" API is just one of the API exposed by the NT kernel.
in WSL mode, ELF are just regular applications running on top of the kernel. But they don't get access to block devices. Neither in kernel-mode (forget about installing a linux filesystem driver) nor in user-mode (no FUSE for you).
They don't get access to USB (libusb works on Linux, work in Win32, but not in WSL). They don't get access to tons of things.
(We could add : no DRM gfx stack, you're limited to X forwarding over SSH, so no wayland compositing either)
Again..... hardware.
Well okay, I was excessively condesning this one (the normal path is DRM(kernel API) -> DRI(user-land component) -> Mesa(OpenGL and co) -> Wayland) But, compositing is *entirely* user-space on Linux.
It requires OpenGL, and a few extensions. (NVidia wasn't providing them, until recently, when they started providing a different set than anyone else, forcing Wayland to feature different paths depending on the hardware.)
It's a normal userland library, whose call are availble to Win32 application (say, Quake3), but not in WSL. Same goes for Vulkan.
Android userspace requires, among direct hardware access, non-mainline kernel extensions....
Android userspace is the whole reason Microsoft started the thing that is WSL now. - Direct hardware access : not in any way more than what any other Windows application would ask access for. - And whether Binder IPC is part of the mainline kernel or not has absolutely nothing to do with WSL. WSL does not use any bit of the Linus kernel. It just tries to expose the same API to applications. Binder IPC not being present in upstream vanilla kernel has nothing to do with microsoft not providing the API in WSL.
You have a very strong opinion about something that you apparently have very little knowledge about.
I've been successfully using WSL to run scientific computing software on Windows. I've been following the various blog post of Microsoft on the subject. I might have a tiny bit more experience and knowledge than you presume.
I'm just point out that currently WSL is just basically some high-level file I/O and some network socket binding and that's about it.
Which covers lots of devs' use-case, and end-user simple use-case (As said above, it worked for us). But that's about it.
Good that you managed to solve the base Thunderbird problem ! (POP3 ought to be abandoned !)
No server. I just have one computer. She's got slow internet. She's not nearby so I only visit every few months.
Then setting up a *local* server ?
Like a low-power single-board computer (you could go to a Raspberry Pi for the popular solution, though beware of later models requiring good supplies in order to not trigger under-voltage CPU throttling).
If you go for a slightly more expensive solution (something that has directly SATA port(s), or at least support good transfer speed over USB3) you could also plugin a disk and install a file server for backups (e.g.: her photo collection, her documents). (With snapshotting cronjobs on the linux side of things, only accessible over SSH. Samba/CIFS only see the topmost snapshot. If mom's laptop ever catches a Ransomware with networking abilities, the virus will only be able to fuck up the latest backup over CIFS, not the older time-line over SSH-only)
Do not try to save money by picking up an excessively cheap SD boot card. Decent UPS solution for SBC boil down usually to a small daughter board with a smartphone's battery management chip and some relatively cheap LiFePo battery plugged to it.
If you're not much into funny home brewed solutions, you could go for a Synology server box which could let you settup most of the above (mail server, backups).
Plus the existence of a linux machine within her network that you can SSH into could let you do some minimal remote admin. (restrict SSH access to public-keys only, no password allowed. optionally install fail2ban)
Slashdot Mirror
I can't tell whether the writer means that incident "X" did not make the landing hazardous,
In the F-15's case that the former :
- The plane has body lift too. Thus there are some safety margins (even with missing bits it can generate enough lift).
- The onboard avionics(*) are able to compensate for quite a lot of situations.
So even with a wing missing, although it couldn't probably perform complex acrobatic maneuvers, could still land safely provided that the pilot is experienced and know how to handle the plane too.
TL;DR: brilliant pilot + bad-ass airplane = can still fly "almost normally" even with missing bits.
Just don't try this it at home, you have drastically reduced quite some safety margins.
---
I've read somewhere that Airbus has some software routine in their avionics that are able to compensate for missing / damamged tail bits. I'll have to track down the reference.
On Firefox 60.0.1 (my distro (OpenSUSE Tumbleweed) package) :
- By default, it isn't checked.
- But right click, and toggle "Check Spelling", and it gets checked by the internal spell checker of Firefox (Spellbound)
- That still doesn't trigger any proof-reader/grammar (such as LanguageTool(*) or Grammarly), thus you're still not covered against homophones or agreements/declension (depending on your language. Here it's English, so only the former).
---
(*) In my opinion, a much better plug-in than grammarly :
- It's open-source
- You can also run a daemon locally.
- So you don't need to necessarily pipe all your text fields into some untrusted 3rd party web server.
Pardon me if I'm wrong.
But aren't the things which are banned from esports *macro programmable* ? As in, you program a whole complete sequence of button press commands (preferrably a complex or difficult one) to a single physical trigger ?
(e.g.: whenever I push button W5, please replay the complete Konami code, followed by that difficult special move from Zangieff on Street Fighter II)
This thing is "simply" a *remapable* controller :
"left foot pedal" = "button B" (because the disabled player is hemiplegic and lacks control on the right side of the body, and thus can't push buttons with the right thumb and thus could get past the first pit in a Mario platformer).
If those computers have two-button mice, then replacing the mice with one-button mice might work.
For an extremely long time on PCs :
"Main button" : gives you a "left"-click
"2nd button" OR "main button" + "shift" : gives you a "right"-click
"3rd button" OR "2nd button + main button" OR "main button" + "ctrl" : gives you a "middle" click.
(I even have manuals of DOS games explaining this).
I saddly don't have a single button mouse handy to test if that still works in 2018 OSes.
But the "two button" for "middle click" definitely works on my laptop on Linux.
On mac :
Most modern single bouton Apple mice have capacitive surface and will emit a click depending on either the finger position or number of fingers.
So the student could have some way around input device screw-up. (It takes just one kid to discover it, then after half an hour, the whole school knows the trick).
The best strategy would be to NOT put the sentence in a <TEXTAREA> to begin with. (avoid triggering the spell checker all together).
They should instead have had a first round to point to errors by clicking words.
and only ask for correct spelling in a second round of the exercices.
Okay, I know that you like to watch sattelite falling into the ocean from very close, but maybe you should start mooring your observering barge a little bit further away from Point Nemo.
I'm sure it's going to solve your speakers problem.
Just saying.
People don't ad block ?
Yep, such people exist out of our /. crowd.
WARNING!!! If you are tempted to try to see how they see the world and are tempted to turn of the ad-blocking in your browser off, be prepared : have a bottle of your favorite eye-bleach within reach, and have the number of a psychological support ready on fast dial on your phone.
We're /. geeks, we still see the "old Microsoft" whenever we talk about that company.
The parent poster isn't refering to the recent gaming business of MS.
The comment was about the numerous time Bill Gate was on stage / on live TV trying to demo some Microsoft software product, and the product blew him in the face with a crash/BSOD/whatever.
You can say tons of bad things about Microsoft back-then (with EEE appreaing very oftenin the discussions), but at least they weren't faking their demo.
There's an old saying about democracy being "two wolves and a sheep voting on what to have for lunch".
That what representative democracy is. (And is huge chunk of the reasons why your system is so screwed in the US).
Meanwhile, direct democracy works. You should try this sometimes.
1.) Run your own little server behind your DSL Modem. Starts at $30, called Raspberry PI.
A littlebit under powered, but still good for many non intensive use.
(Been there, still doing that sometimes).
2.) Store your data on your private server. Control access/encipher via SSH/SCP.
Oh, common. It's 2018. at least use rsync.
(Also, if you want good resilience against damage/corruption, you'd want a base with a tad bit more sata ports, and running RAID5.
Just saying)
3.) Run your own discussion forums.
...but be prepared to not have the richness of interactions currently found on the Social Networking Site du jour/whatever Zuck has bought the most recently.
You'll basically have your mom' writing you a message, every now and then, to remind you to call her.
4.) Run Linux on your PC instead of the Snoop-Ware of MSFT.
But be prepared to spend time on stack exchange just to learn how to make your touch pad work* ~~~
5.) Run your own Jabber server instead of Corporate-Controlled Whatsapp and the like.
And you'll have exactly 1 single user registered on it :
you.
Because most of the people you know won't bother installing yet another client (but it's opensource) just to send messages to you, when all their friends are on Skype/FBMessenger/WhatsApp/Snapchat/whatever is popular next (hint: look at what Zuck is trying to buy next).
---
* - Disclaimer: Proudly written from a Linux running laptop
Google and Youtube already "nudge users into alignment with their goals"
In that context, *their goals* = Google's goals.
You know the "feelgood" piece at the end about solving poverty, disease and world peace.
So Google can justify what they're doing by deluding themselves it will lead to "greater good".
by manipulating search results, pushing sites/producers with opinions they prefer and hiding those they disagree with.
in your context *they prefer* and *they disagree* are the users :
- algorithms are optimized for one single target : bring more clicks in (because that's what makes more money going in by providing more eyeballs to sell to advertisers)
- and the machine "learning/AI/NN/whatever is powering google now" automatically learns that the best way to achieve that is to keep the users in their bubble, preferably with even more extreme and emotion-appealing content.
end result:
- if you're a tiny bit left leaning you'd get endless stream of video about social injustice.
- if you're a tiny bit on the right you'd get a huge barrage of video about all the ultra-violent rapist-thieving-thugs immigrant that are about to invade the country
and it all ends up spiraling toward conspiracy theory revealing to you that the government is lying by making you think the earth is round and hiding the fact that it's actually flat... (because ?...)
TL;DR: the current think that might look like censorship is just a incidental side-effect of the algorithm trying to give you whatever make you stay for more content, and makes you a better recipient for ads.
There's no conscious censorship going on.
All my mission critical windows systems are now quarantined from the internet and other networks
Which by itself is a good approach as it drastically decrease *nearly all* security and safety risk for mission critical equipement.
(Basically only a few very advanced threats geared toward air gapped targets as stuxnet)
No mater the insanity of Windows 10 updates.
The next step would be to enclose them inside VMs and use a slightly saner host OS (something unix-y, e.g.: Linux) to handle the VMs.
Including snapshotting for safe roll back (which among other could help mitigate the buggy updates)
While of course still keeping them network isolated (they are mission critical after all).
If your Samsung is 840 or later, you can get its own measure of temperature out of SMART attribute "190 Airflow_Temperature_Cel".
(I just don't know how to do this in Windows 10. I mostly use Linux and smartmontools.
Maybe Samsung's Wizard tools can do it ? Maybe SpeedFan ? Have a look here.)
On Samsung 830, I haven't seen such an attribute.
If that unknown sensors is nearby the SSD, it and the SSD's sensor should evolve similarly over time (spike at roughly the same time, with more or less similar top temperatures)
From the abstract :
They directly state that they have observed "association".
They are only suggesting that physical activity guideline should take into account leisure vs. occupation.
Plus, the hypothesis evoked here on /. (exposure to toxic chemimcals) and the hypothesis from the summary (more trivially, less rest whenever wanted) both boil down to "in occupation, the physical exercie cannot be done while removing negatively implacting factors, unlike when doing it as a leisure".
Answer trashed by miss-typing, I'll try to rewrite my reply again.
Brypt is indeed designed as a *password* hashing function, so it's better for passwords than sha-2 is. I think sha-2 is also acceptable.
Bcrypt, Scrypt, Argon2 (and the older PBKDF2, for completeness) were all designed for password and are designed with resistance to bruteforce in mind.
sha{n} family were designed for speed (and sha-3/keccak also because a sponge function is an interesting new concept) and are only acceptable if you don't mind brute forcing.
Definitely don't use a common password. Using "password" as your password will suck no matter what else you do. Therefore it's a mistake to say:
Doing X won't help if the passwords suck. Therefore don't do X.
The problem isn't simply people using "password" or "123456" for password. The problem is general approach to security.
We human suck at picking strong purely random password.
We /. geeks, will probably pipe a decent (more than 256) amount from /dev/random (not even the 'u' one! with several true random source mixed into the kernel entropy pool !!!) through base64, and use the output for only 1 specific site, and store it into a decent opensource password manager, that backs up on our own cloud.
We also activate two-factor authentication whenever possible (using an OTP app, not insecure channels like SMS).
That thing is pretty much secure. Even with a fast hash, it's not realistic to brute force. And in case of a database leak, an attacker wouldn't gain much, until the leak is discovered, they'll only could try to log into that specific site and they'll still will be missing the 2FA.
Normal people don't do that. They seldom activate two-factor authentication (it's too cumbersome). /. geeks crowd using /dev/random. But almost all the others).
And when asked to pick a password, they tend to follow pattern.
Most frequently, they will put the required capital letter at the beginning, use a 5-6 letter combo, followed by 2-4 of the required digit, and put an "!" at the end to the mandatory punctuation sign.
Such a password follows the most stringent rules nearly everywhere. It will show a green light on most password strength evaluation.
But do the math : such a password, in best case (6 letter, 4 digits. The position of the capitalisation at the beginning and the "!" at the end don't bring much as they are predictable) has less than 42 bits of entropy. Collect all the other common patterns, allow for a few extra substitution, and you'd probably still be within 56bits, something that isn't considered secure at all nowadays.
Using these pattern, if a database using only slat+fast hashes is leaked, you can recover a very sizeable fraction of all the password within. (Not all password. There's bound to be two or three of
Worse, lots of people have a tendency to re-use passwords (out of convenience, it's hard to remember "good" passwords).
Once a database is leaked and passwords recovered, a sizeable amount of these password could be used to open other stuff (Access the e-mail account associated with that user's database record ? (and once logged in, try to use the "reset password" function of any website associated with this e-mail ?) Try to log into several social network website using the same e-mail and password ? etc.)
In short, it's not only people using "password" and "123456", it's every body who is not using /dev/[u]random (the biggest fraction of your userbase)
Because we know weak passwords will be weak no matter what you do, it probably makes sense to try to make the system as secure as possible +for users who use good passwords+.
On the other hand because we know very few users will use purely random password + a password manager, and activate two-factor authentication, it makes sense to try to make a potential database leak as resistant to brute-forcing as possible.
Specially because it only costs us a few hundreds of milliseconds and a little bit more RAM once at the beginning of a session, during the log-in phase.
Maybe if they can get Windows 10 to run Android apps, like ChromeOS, they might stand a chance.
Well, that was the whole initial purpose of WSL / BashOnWindows.
Except it's currently still limited to file I/O and some network sockets.
So they pivoted WSL into the current "tool for devs to run and test linux dirsto's userspace", and Windows 10 Mobile is still without any access to the 2 main relevant app ecosystem.
Meaning that due to networking effect, their platform is more or less useless, except for the couple of weird user who are only interested into phone functionality and are happy with the one or two apps that got ported on the platform.
So basically all the former Zune users. All five of them.
So basically they've been getting a taste of their own monopoly/network effect that they've used on the desktop.
Hey Microsoft, how's being on the receiving end of your own bullcrap ?
why don't you just say that this poor system will never be able to run well
Assuming* it's using x86 and not ARM, :-P
you could always install Linux on it to have it run well
---
*: and thus UEFI Secure Boot must mandatory enable the end-user to boot into something else than the microsoft-signed windows. (by disable secure boot and/or by enabling the user to add any extra signing key).
Yes, I get your point.
- Parent poster points about rainbow table (tables that point hashes back to strings that can generate the same hash).
- You point that a well designed (=non borked design) hashing function should give two different hash for two dissimilar short passwords. Thus you would need a giant rainbow table that gives a password for *every single possible 160bit hash* (that's ~10^48 entries, i.e.: within an order or two from the number of atoms on earth). Fat chance.
BUT!
Even if the hash->password direction is *hard*.
password->hash direction is easy.
All the algorithm you mention (MD5, SHA1, SHA2 and let's throw SHA3/KECCAK in the mix too) are all extremely *fast hasing functions* (They are ultra fast, and have very low ressource requirement by design).
Meaning you can take a GPU running a special compute shader/OpenCL/Cuda code that can process millions of them in a second.
So you could scan through ALL the common password (based on frequent leaked passwords and/or on frequent paterns, etc. and their substitutions) within a reasonable time until you find a match.
As the summary points out, we humans are bad at picking-up password, we definitely use less than 2^ ${whatever bits used by current popular hash} different passwords.
Even if you use salt (so your hash doesn't match any other precedent hash in any rainbow table), and even if you use the latest *hashing* function (SHA3 - well okay, it's a sponge function, but basically works the same), it's definitely within the reach of a reasonable budget to loan GPU compute time on the cloud and brute force the passwords.
So if a database containing SHA{n}-hashed (and optionally salted) passwords get leaked, you can consider that all except the most unusual passwords can be brute-forced.
So in short DO NOT USE HASHES. USE KEY DERIVATION FUNCTIONS.
Things like bcrypt, scrypt or the current competition winner argon2, are on purpose designed to be slow and resource intensive.
(By iterating multiple rounds, by require significant memory, etc.)
For you, it doesn't change much if loging in take now a third of a second - you only log once, after all, it won't kill you to wait for 300ms just once at the beginning of your session.
But for potential brute-forcers, not being able to quickly go through million of tests is suddenly a huge blocker.
So in short :
do NOT use SHA2 for your password database.
use bcrypt/scrypt/argon2 instead.
Yeah, Caterpillar, where's my P-5000 Work Loader ?
Anything bigger uses big diesel tugs that are built like tanks.
Sometime *literally* like tanks (Aliens)
(or the soviet reversal thereof).
I immediately though "well, basically static vs. dynamic friction", too.
On the other hand, electric motors are specially good at having decent torque at very low speed.
(You don't need to shift a different gear to start them, unlike ICE. They use a fixed transmission ratio).
So it's about the best type of motors you could be using for "just pulling".
If this is the top speed,
I suspect that's more for braking safety.
If anything goes wrong, you'll need to brake.
- there's only so much kinetic energy that you can shed with the car's brakes/regenerative braking.
- there's the reaction time of the guys in the cockpit hitting the landing gear's brakes, whenever alerted over comms about something out of their field of view.
That could be an explanation for the czech being among the most indoor people :
historically a former eastern bloc country, with some country regions having significant industrial development (coal mining, iron smelting, etc.) but not so much pollution control.
thus people will percieve the outdoor air being more polluted and will develop the habit of staying indoor.
(among tons of other different cultural reasons).
America has minimal government interference in transactions between consenting adults,
Except when it concerns what's going in the bedroom (sexual orientation, sex workers, etc.)
Then suddenly it's the government business to interfere legally as much as possible.
which is proper.
The idea behind the various European governments is to balance long term risk and costs.
The thing to which the adults might consenting could come with tons of long-term risk.
Health (both physical and mental) and safety risks, that the public healthcare system could end-up paying.
By putting some limitations on health and safety hazards, the government is limiting the money that they would have to chip-in in the long-term.
Better make sure to have a healthy population that has enough opportunity for out-door activities, rest, healthy diet, etc.
than having to support tons of people with burn-outs, depression, work-place accidents, work-place chronic disease, obesity by being on constant fast-food diet, etc.
Oh, yeah. I forgot. That requires to actually something resembling a public health-care system.
(Note that the same idea goes regarding drugs or prostitution. People will be doing it anyway, so it better be done within normal, legal (and tax paying) respectable business or self-employment, rather than enriching criminal gangs and putting a high unnecessary burden on the legal and prison systems).
If you want extra vacation rather than higher pay, then that should be between you and your employer, not something imposed on every worker by the government.
And when you ask, you're employer will say no.
If you try to shop around you'll find most other employer are think why should *they* if nobody else is doing it.
Thinking of leveraging your exclence/skills/competencies/relevance as a key employee?
Haha... The employer would dream to replace you with someone less experienced, less competent, but incredibly cheaper, and then give himselve a bonus to celebrate the decreased costs with absolutely no long-term thinking.
Seems to us European that in the US, the negotiations between employee and corporations is strongly dis-balanced in favour of the later.
But yeah, unions, more social-leaning polical parties and anything that could work a little bit more toward better situations for employ is "evil communism" and should be fought off.
Solar and wind are network connections, the former giving a relatively fixed amount of data over the day which changes by the hour, and the latter shoving random amounts of data down the pipe.
So shared connection over TV-cable network vs. 3G connection in an area with bad coverage, resp. ?
No, cgroups & co are there and functional. Namespacing, what I think you really meant, is not yet globally enabled, so no containers.
Namespacing is what I meant by "& co" - i.e.: and all the other kernel bits that make container isolation possible but written shorter.
Yes, this is because as a guest,
No, it's not a "guest". There's no virtualization going on, it's not Hyper-V based.
It's supposed to be just a different set of API made available by the NT kernel, in addition to win32.
The same way a OS/2-like API was available a long time ago.
it's not supposed to be interfacing directly with hardware, this includes block devices.
WSL's "linux-like" API is just one of the API exposed by the NT kernel.
in Win32 API mode, the kernel *does* expose to application some hardware including block devices (used by several user-space utilities. Including disk-image writers, and including user-mode tools to access Ext2/3,ReiserFS, etc.)
in WSL mode, ELF are just regular applications running on top of the kernel. But they don't get access to block devices.
Neither in kernel-mode (forget about installing a linux filesystem driver)
nor in user-mode (no FUSE for you).
They don't get access to USB (libusb works on Linux, work in Win32, but not in WSL).
They don't get access to tons of things.
(We could add : no DRM gfx stack, you're limited to X forwarding over SSH, so no wayland compositing either)
Again..... hardware.
Well okay, I was excessively condesning this one (the normal path is DRM(kernel API) -> DRI(user-land component) -> Mesa(OpenGL and co) -> Wayland)
But, compositing is *entirely* user-space on Linux.
It requires OpenGL, and a few extensions.
(NVidia wasn't providing them, until recently, when they started providing a different set than anyone else, forcing Wayland to feature different paths depending on the hardware.)
It's a normal userland library, whose call are availble to Win32 application (say, Quake3), but not in WSL.
Same goes for Vulkan.
Android userspace requires, among direct hardware access, non-mainline kernel extensions....
Android userspace is the whole reason Microsoft started the thing that is WSL now.
- Direct hardware access : not in any way more than what any other Windows application would ask access for.
- And whether Binder IPC is part of the mainline kernel or not has absolutely nothing to do with WSL. WSL does not use any bit of the Linus kernel. It just tries to expose the same API to applications. Binder IPC not being present in upstream vanilla kernel has nothing to do with microsoft not providing the API in WSL.
You have a very strong opinion about something that you apparently have very little knowledge about.
I've been successfully using WSL to run scientific computing software on Windows.
I've been following the various blog post of Microsoft on the subject.
I might have a tiny bit more experience and knowledge than you presume.
I'm just point out that currently WSL is just basically some high-level file I/O and some network socket binding and that's about it.
Which covers lots of devs' use-case, and end-user simple use-case (As said above, it worked for us). But that's about it.
Good that you managed to solve the base Thunderbird problem !
(POP3 ought to be abandoned !)
No server. I just have one computer. She's got slow internet. She's not nearby so I only visit every few months.
Then setting up a *local* server ?
Like a low-power single-board computer (you could go to a Raspberry Pi for the popular solution, though beware of later models requiring good supplies in order to not trigger under-voltage CPU throttling).
If you go for a slightly more expensive solution (something that has directly SATA port(s), or at least support good transfer speed over USB3) you could also plugin a disk and install a file server for backups (e.g.: her photo collection, her documents).
(With snapshotting cronjobs on the linux side of things, only accessible over SSH. Samba/CIFS only see the topmost snapshot.
If mom's laptop ever catches a Ransomware with networking abilities, the virus will only be able to fuck up the latest backup over CIFS, not the older time-line over SSH-only)
Do not try to save money by picking up an excessively cheap SD boot card.
Decent UPS solution for SBC boil down usually to a small daughter board with a smartphone's battery management chip and some relatively cheap LiFePo battery plugged to it.
If you're not much into funny home brewed solutions, you could go for a Synology server box which could let you settup most of the above (mail server, backups).
Plus the existence of a linux machine within her network that you can SSH into could let you do some minimal remote admin.
(restrict SSH access to public-keys only, no password allowed. optionally install fail2ban)