One might as well complain about all the basic utilities under the GNU project umbrella.
I can use ls without having to use info, but I can't use systemd-networkd without using systemd. Conversely, there is no logging system other than systemd-journald that works with systemd.
In other words, each individual program that makes up the "systemd brand" must all be installed and running or else none of them work. This is completely different from the current init system, which doesn't care which system logger (for example) you use, and doesn't even require you to use one at all.
So, even though the "systemd brand" is many separate applications, the net result is no different from one monolithic application with many shared libraries.
(2) Most importantly, I can hack that bash script to do whatever I damn well please.
And, although you can hack the "sendmail.service" file shown in that link, your hack will be overwritten the next time sendmail is updated, since the file is in/usr/lib/systemd/system. Also,/usr could be a read-only filesystem.
Instead, to make your hack effective, you'll need to create a file somewhere under/etc/systemd (your guess is as good as mine...the documentation sucks) that will do what you want. Since there is also no documentation about what, exactly, must be in the file (do you need every entry that was in the original, or can you just override what you want to change?), you'll have to play around for a while to see what works.
The right thing to do is for the sendmail package maintainer to place an example user file in the right place, and comment out everything so it doesn't actually overwrite the default, but the comments will let a sysadmin know what to change if they need to. But this is yet another major problem with systemd...it has so damn many config files, that if every package maintainer did this, you'd have hundreds of files in the override directory, even though you only need a few for the changes you want to make.
So, the really right thing to do is to not keep config files of any sort in/usr/lib, but instead put them only in/etc, and then any changes the user makes there are applied as the should be. This is not the systemd way, though, as the systemd maintainers know much better how your system should be run than you do.
Considering this ship is designed to hold however many thousand cargo containers and the crew to support it, it needs a lot of empty space inside.
According to TFA, the crew while at sea is 15 people, which is nothing compared to the 5,000 or so on a fully-populated-for-war aircraft carrier.
When you add in the fact that a warship is supposed to be able to go for at least weeks at a time without any replenishment, needs a much stronger hull for its size, has a lot more electronics that need special cabling and conduits, etc., commercial vessels are actually quite easy to build in comparison.
The mongolian clusterfuck is the result of the byzantine cofiguration rules each vendor has for determining a blade's NIC or FC mapping with the blade center's (overpriced) internal switch bays.
Cisco's blades do all of this through software...you can add and delete NICs and fiber channel cards with a couple of mouse clicks on the Java applet that runs in the browser.
your cars, in your televisions, in your game systems, in your embedded devices...
Now your fanboyism is showing. None of those are significantly Linux-driven, which is mostly because of the GPL.
What are you talking about? Many of those devices do use the Linux kernel, which is all "Linux" really is. The userspace many people think of as "Linux" is really a collection of other software, much of it GNU.
For example, I have two media players, an eBook reader, a TV, a surveillance camera, and an A/V receiver that all use the Linux kernel. On the first three, I can get a shell prompt, and can cross-compile apps to run on them. None of the devices I'm talking about run Android, either.
I've read the agreements. The cable guy lied to you. They can't unbundle what they pay, but they can unbundle what you pay.
Apparently, you haven't read the agreements in detail. Not only does the cable company have to pay the content provider for the full package, the agreement also says that the cable company has to provide all those channels on some particular tier (like "basic", "expanded basic", etc.) so that the content provider gets to say "available in X% of homes" to keep advertising dollars higher.
So, yes, the cable company could unbundle the channels and only charge you for the channels you want, but you'd still get all the other channels from that content provider based on your "tier". The only way to avoid this would be to have a completely "a la carte" tier. If that happened, though, the content providers would have lower advertising dollars, and would thus have to charge more for their content, which would mean that a la carte would end up at similar prices to what you pay now, unless you truly do just watch one or two channels.
Yeah, you say that now, but when we get more power, you can all but guarantee we'll use more power.
There's a limit to that before the Earth becomes Venus or Mercury.
Basically, the vast majority of the power on Earth right now came from the sun at some point in time. The exception is nuclear reactions (either in power plants or in the rocks in the ground). Right now, these exceptions are a very small part of the total, but cheap fusion would change that dramatically.
Eventually, all the power turns to heat, and too much of it is a very bad thing. Perhaps reducing CO2 (no <sub> tag support...really Slashdot?) emissions by cutting back dramatically on fossil fuel usage would help balance out and allow radiation of heat to space to be more efficient, but it's hard to say when everything is still speculation.
If you use a commonly used phrase, its the same as using a commonly used password.
As I said, if you "pick the words for your pass phrase from a small, well-known dictionary", you're going to be in trouble. A "commonly used phrase" would be the extreme version of using a limited pool of words.
Though, if it is enough words, and you use some kind of mutator (vary spacing, capitalization & punctuation, use l33t, etc.), even knowing the exact phrase would take a few thousand guesses to find which mutators were used in which position.
Commonly used passwords are vulnerable to dictionary attacks, that doesn't change when you use passphrases.
Yes, it does, unless you do all the following:
Pick the words for your pass phrase from a small, well-known dictionary.
Follow the spacing expected by the attacker.
Use only the case the attacker expects (all upper, all lower, proper caps, etc).
Use only letters and spaces...no punctuation or special characters.
Don't do any substitution of characters (no l33t, etc.)
Spell every word correctly.
It's easy to create a phrase that is personal to you and won't appear in any Google search. But, even if it does, if you don't just use lowercase letters with the words run together, it will take a long time for the attacker to run through all the permutation tricks on a 40+ character phrase.
And here's a really good one...the part of your post that I quoted would make an excellent pass phrase, since it contains one word that isn't in the *nix words list. Something as simple as making a compound out of "pass phrase" is enough to cause an attacker pain if they use the wrong dictionary. And, when attackers start including every single "word" in their dictionary, it gets even closer to brute force. When you use "Tatooine" and "Mordor" in your pass phrase that doesn't in any other way reference "Star Wars" or LoTR, it's pretty secure: Tucson is hot, but it's no Mordor or Tatooine. Easy to remember, easy to type, but painful to crack.
It's also a potential DOS for the server if a bunch of people start submitting preposterously long "passwords" anywhere they have a password box.
Nobody's asking for sites to allow you to use your favorite novel as a password, but limiting to some insanely short value is not the right way to solve the problem.
Set a limit of 255 characters for the password, and you won't get any complaints about too short a limit while keeping the computing requirements for the hash creation reasonable.
this strain of ebola appears to have a 70% mortality rate.
Mortality rates for hemorrhagic fevers are often inversely proportional to the level of medical care available.
Keeping a patient cool and hydrated reduces mortality rate dramatically. Having antibiotics on hand to battle secondary infections is also a big plus. Even a supply of more powerful fever-reducing drugs than aspirin would be considered a luxury in many of the places where Ebola has a high mortality rate.
For any truly infectious disease, proper procedure would have health care workers walk into a disinfectant shower (and possibly UV light) before removing protective clothes. Any disease that can survive that sort of thing is going to kill us all anyway.
Then, order of removal is important. In general, headgear is removed first (preferably by another person), then outer gloves, then fasteners released and gear removed, then inner gloves. All this is followed by hand washing (at a minimum). This makes sure that easier paths to infection get as little possible contact from anything that might have had contact with the pathogen.
The nurse screwed up by touching her face with her outer glove, and I suspect that disinfectant showers/UV were not done first.
Don't get me wrong, I think BitTorrent is very cool technological achievement. But transferring data between semi-random hosts around the globe and opening hundreds of TCP connections per computer while doing it, is like the ultimate way to clog the pipes.
BitTorrent uses UDP when done correctly, and pretty much becomes the absolute best way to get data to many computers very quickly.
A torrent with few seeders isn't very efficient, but one with many hundreds of well-configured peers is hard to beat on overall transfer speed.
Although Dan Brown's books do require extreme suspension of disbelief, I've never heard any claims that the author actually experienced any of the same kind of events in his life, which is what people are having a problem with in this case.
Unfortunately, the number of ridiculous hoops you need to go through to let an unsigned Java applet run an arbitrary network I/O makes it much less useful.
They now have a Flash version as well, so it's easier. But the numbers appear really low, claiming that my network buffer limits download to 140Mbps, yet I have often downloaded actual files from the Internet at faster than that.
OTOH, all the Ookla-powered sites claim I get over 70% of my 1Gbit network card speed, which I also find hard to believe, despite having a 20Gbps connection to our ISP (with literally thousands of users, one of which is a server I maintain that downloads at over 2000Mbps 24/7 backing up a remote site).
Saturating my download is simple, it's my upload that's hard.
You could also try for total composite speeds by using a bunch of torrents. My seedbox regularly runs 40Mbps total upload speed on the 50+ torrents, even though no one torrent is running that fast. I actually throttle the max upload to 50Mbps to allow 30Mbps free bandwidth to FTP the files to my home.
I'm relying on the very vocal detractors of systemd to be sufficiently competent to install the init system of their choice because they are adopting the position of saying they are smarter and better than the developers of systemd.
No, we're adopting the position of "if it ain't broke, don't fix it" and "provide your supposedly better solution as an alternative and let it win on its merits".
Soft keys but with great tactile feel, and completely programmable so you could easily swap the CTRL and CAPS LOCK keys. It was IBM's take on a silent keyboard but will all their (then) quality thrown in.
Like most keyboards, it has the function keys at the top, and the inverted-T arrow keys. These are the two things I absolutely hate, and why I have Northgate keyboards, like the Omnikey Ultra and the Omnikey Plus. Unfortunately, there were multiple versions of both of these keyboards, and some have the inverted-T.
But, the programmable keys are nice, and the Model M needs that as it suffers from the Caps Lock key being next to "A". I use Caps Lock so infrequently that I programmed it to be the Windows key (which my Northgates don't have).
I haven't seen a lot of pressure yet. With RHEL6 the urge to switch came on fast. Esp with regards to virtualization and what not. I haven't gotten the sense that RHEL7 has a whole lot of "must have now" tech in it, as opposed to the amount of systemd fear it has.
We're a mixed shop, running CentOS 6 and Ubuntu LTS (12 and upgrading to 14). We will not be upgrading to any version of either distro that has systemd until at least a year after release, and maybe even later, despite the fact that upgrades to LTS 14 happened within a month of release for many of our systems.
Yup - it's not SystemD's fault for solving this problem first.
It's systemd's problem for solving it in a non-portable way.
If I could download a replacement "VTd" package, install it, and get the new features, that'd be great. Unfortunately, I have to replace the Linux kernel (so that the VTs there don't complete with systemd's version) plus replace config files for a bunch of other services that systemd replaces, plus verify that my startup scripts work with systemd, plus.... Get the point?
This is not just a systemd problem, this is a problem for any init system that wants to support multi-seat, and sane switching between VTs:
I have no idea what you mean by "multi-seat", but the existing SysV init system has no problem with simultaneous logins from any number of users.
If you mean hooking up multiple keyboards and monitors to the same system at the same time, why would you ever want to do that when you can just log in to the system over the network (text or GUI, your preference).
Slashdot Mirror
One might as well complain about all the basic utilities under the GNU project umbrella.
I can use ls without having to use info, but I can't use systemd-networkd without using systemd. Conversely, there is no logging system other than systemd-journald that works with systemd.
In other words, each individual program that makes up the "systemd brand" must all be installed and running or else none of them work. This is completely different from the current init system, which doesn't care which system logger (for example) you use, and doesn't even require you to use one at all.
So, even though the "systemd brand" is many separate applications, the net result is no different from one monolithic application with many shared libraries.
If I didn't know otherwise, I'd say you were tinkering with my Fedora system.
I decided switching to a different distribution would be easier than getting systemd to do what I want.
(2) Most importantly, I can hack that bash script to do whatever I damn well please.
And, although you can hack the "sendmail.service" file shown in that link, your hack will be overwritten the next time sendmail is updated, since the file is in /usr/lib/systemd/system. Also, /usr could be a read-only filesystem.
Instead, to make your hack effective, you'll need to create a file somewhere under /etc/systemd (your guess is as good as mine...the documentation sucks) that will do what you want. Since there is also no documentation about what, exactly, must be in the file (do you need every entry that was in the original, or can you just override what you want to change?), you'll have to play around for a while to see what works.
The right thing to do is for the sendmail package maintainer to place an example user file in the right place, and comment out everything so it doesn't actually overwrite the default, but the comments will let a sysadmin know what to change if they need to. But this is yet another major problem with systemd...it has so damn many config files, that if every package maintainer did this, you'd have hundreds of files in the override directory, even though you only need a few for the changes you want to make.
So, the really right thing to do is to not keep config files of any sort in /usr/lib, but instead put them only in /etc, and then any changes the user makes there are applied as the should be. This is not the systemd way, though, as the systemd maintainers know much better how your system should be run than you do.
Considering this ship is designed to hold however many thousand cargo containers and the crew to support it, it needs a lot of empty space inside.
According to TFA, the crew while at sea is 15 people, which is nothing compared to the 5,000 or so on a fully-populated-for-war aircraft carrier.
When you add in the fact that a warship is supposed to be able to go for at least weeks at a time without any replenishment, needs a much stronger hull for its size, has a lot more electronics that need special cabling and conduits, etc., commercial vessels are actually quite easy to build in comparison.
The mongolian clusterfuck is the result of the byzantine cofiguration rules each vendor has for determining a blade's NIC or FC mapping with the blade center's (overpriced) internal switch bays.
Cisco's blades do all of this through software...you can add and delete NICs and fiber channel cards with a couple of mouse clicks on the Java applet that runs in the browser.
your cars, in your televisions, in your game systems, in your embedded devices...
Now your fanboyism is showing. None of those are significantly Linux-driven, which is mostly because of the GPL.
What are you talking about? Many of those devices do use the Linux kernel, which is all "Linux" really is. The userspace many people think of as "Linux" is really a collection of other software, much of it GNU.
For example, I have two media players, an eBook reader, a TV, a surveillance camera, and an A/V receiver that all use the Linux kernel. On the first three, I can get a shell prompt, and can cross-compile apps to run on them. None of the devices I'm talking about run Android, either.
I've read the agreements. The cable guy lied to you. They can't unbundle what they pay, but they can unbundle what you pay.
Apparently, you haven't read the agreements in detail. Not only does the cable company have to pay the content provider for the full package, the agreement also says that the cable company has to provide all those channels on some particular tier (like "basic", "expanded basic", etc.) so that the content provider gets to say "available in X% of homes" to keep advertising dollars higher.
So, yes, the cable company could unbundle the channels and only charge you for the channels you want, but you'd still get all the other channels from that content provider based on your "tier". The only way to avoid this would be to have a completely "a la carte" tier. If that happened, though, the content providers would have lower advertising dollars, and would thus have to charge more for their content, which would mean that a la carte would end up at similar prices to what you pay now, unless you truly do just watch one or two channels.
On the gripping hand, did anyone else read...
FTFY. Oh, and turn in your geek card.
Yeah, you say that now, but when we get more power, you can all but guarantee we'll use more power.
There's a limit to that before the Earth becomes Venus or Mercury.
Basically, the vast majority of the power on Earth right now came from the sun at some point in time. The exception is nuclear reactions (either in power plants or in the rocks in the ground). Right now, these exceptions are a very small part of the total, but cheap fusion would change that dramatically.
Eventually, all the power turns to heat, and too much of it is a very bad thing. Perhaps reducing CO2 (no <sub> tag support...really Slashdot?) emissions by cutting back dramatically on fossil fuel usage would help balance out and allow radiation of heat to space to be more efficient, but it's hard to say when everything is still speculation.
If you use a commonly used phrase, its the same as using a commonly used password.
As I said, if you "pick the words for your pass phrase from a small, well-known dictionary", you're going to be in trouble. A "commonly used phrase" would be the extreme version of using a limited pool of words.
Though, if it is enough words, and you use some kind of mutator (vary spacing, capitalization & punctuation, use l33t, etc.), even knowing the exact phrase would take a few thousand guesses to find which mutators were used in which position.
Commonly used passwords are vulnerable to dictionary attacks, that doesn't change when you use passphrases.
Yes, it does, unless you do all the following:
It's easy to create a phrase that is personal to you and won't appear in any Google search. But, even if it does, if you don't just use lowercase letters with the words run together, it will take a long time for the attacker to run through all the permutation tricks on a 40+ character phrase.
And here's a really good one...the part of your post that I quoted would make an excellent pass phrase, since it contains one word that isn't in the *nix words list. Something as simple as making a compound out of "pass phrase" is enough to cause an attacker pain if they use the wrong dictionary. And, when attackers start including every single "word" in their dictionary, it gets even closer to brute force. When you use "Tatooine" and "Mordor" in your pass phrase that doesn't in any other way reference "Star Wars" or LoTR, it's pretty secure: Tucson is hot, but it's no Mordor or Tatooine. Easy to remember, easy to type, but painful to crack.
It's also a potential DOS for the server if a bunch of people start submitting preposterously long "passwords" anywhere they have a password box.
Nobody's asking for sites to allow you to use your favorite novel as a password, but limiting to some insanely short value is not the right way to solve the problem.
Set a limit of 255 characters for the password, and you won't get any complaints about too short a limit while keeping the computing requirements for the hash creation reasonable.
this strain of ebola appears to have a 70% mortality rate.
Mortality rates for hemorrhagic fevers are often inversely proportional to the level of medical care available.
Keeping a patient cool and hydrated reduces mortality rate dramatically. Having antibiotics on hand to battle secondary infections is also a big plus. Even a supply of more powerful fever-reducing drugs than aspirin would be considered a luxury in many of the places where Ebola has a high mortality rate.
Uhh, she was in the process of undressing...
Then, correct procedures weren't followed.
For any truly infectious disease, proper procedure would have health care workers walk into a disinfectant shower (and possibly UV light) before removing protective clothes. Any disease that can survive that sort of thing is going to kill us all anyway.
Then, order of removal is important. In general, headgear is removed first (preferably by another person), then outer gloves, then fasteners released and gear removed, then inner gloves. All this is followed by hand washing (at a minimum). This makes sure that easier paths to infection get as little possible contact from anything that might have had contact with the pathogen.
The nurse screwed up by touching her face with her outer glove, and I suspect that disinfectant showers/UV were not done first.
Don't get me wrong, I think BitTorrent is very cool technological achievement. But transferring data between semi-random hosts around the globe and opening hundreds of TCP connections per computer while doing it, is like the ultimate way to clog the pipes.
BitTorrent uses UDP when done correctly, and pretty much becomes the absolute best way to get data to many computers very quickly.
A torrent with few seeders isn't very efficient, but one with many hundreds of well-configured peers is hard to beat on overall transfer speed.
Although Dan Brown's books do require extreme suspension of disbelief, I've never heard any claims that the author actually experienced any of the same kind of events in his life, which is what people are having a problem with in this case.
http://www.measurementlab.net/...
Unfortunately, the number of ridiculous hoops you need to go through to let an unsigned Java applet run an arbitrary network I/O makes it much less useful.
They now have a Flash version as well, so it's easier. But the numbers appear really low, claiming that my network buffer limits download to 140Mbps, yet I have often downloaded actual files from the Internet at faster than that.
OTOH, all the Ookla-powered sites claim I get over 70% of my 1Gbit network card speed, which I also find hard to believe, despite having a 20Gbps connection to our ISP (with literally thousands of users, one of which is a server I maintain that downloads at over 2000Mbps 24/7 backing up a remote site).
Saturating my download is simple, it's my upload that's hard.
You could also try for total composite speeds by using a bunch of torrents. My seedbox regularly runs 40Mbps total upload speed on the 50+ torrents, even though no one torrent is running that fast. I actually throttle the max upload to 50Mbps to allow 30Mbps free bandwidth to FTP the files to my home.
If your downlink out paces your disk writes, it's time to upgrade.
Or, if you have a good SSD and your download is faster than the disk, your network is fast enough...stop worrying.
I'm relying on the very vocal detractors of systemd to be sufficiently competent to install the init system of their choice because they are adopting the position of saying they are smarter and better than the developers of systemd.
No, we're adopting the position of "if it ain't broke, don't fix it" and "provide your supposedly better solution as an alternative and let it win on its merits".
Soft keys but with great tactile feel, and completely programmable so you could easily swap the CTRL and CAPS LOCK keys. It was IBM's take on a silent keyboard but will all their (then) quality thrown in.
Like most keyboards, it has the function keys at the top, and the inverted-T arrow keys. These are the two things I absolutely hate, and why I have Northgate keyboards, like the Omnikey Ultra and the Omnikey Plus. Unfortunately, there were multiple versions of both of these keyboards, and some have the inverted-T.
But, the programmable keys are nice, and the Model M needs that as it suffers from the Caps Lock key being next to "A". I use Caps Lock so infrequently that I programmed it to be the Windows key (which my Northgates don't have).
I think you may have misunderstood my post -- probably I wasn't sufficiently clear.
Too subtle, I think. Yes, I did misunderstand.
I haven't seen a lot of pressure yet. With RHEL6 the urge to switch came on fast. Esp with regards to virtualization and what not. I haven't gotten the sense that RHEL7 has a whole lot of "must have now" tech in it, as opposed to the amount of systemd fear it has.
We're a mixed shop, running CentOS 6 and Ubuntu LTS (12 and upgrading to 14). We will not be upgrading to any version of either distro that has systemd until at least a year after release, and maybe even later, despite the fact that upgrades to LTS 14 happened within a month of release for many of our systems.
Yup - it's not SystemD's fault for solving this problem first.
It's systemd's problem for solving it in a non-portable way.
If I could download a replacement "VTd" package, install it, and get the new features, that'd be great. Unfortunately, I have to replace the Linux kernel (so that the VTs there don't complete with systemd's version) plus replace config files for a bunch of other services that systemd replaces, plus verify that my startup scripts work with systemd, plus.... Get the point?
This is not just a systemd problem, this is a problem for any init system that wants to support multi-seat, and sane switching between VTs:
I have no idea what you mean by "multi-seat", but the existing SysV init system has no problem with simultaneous logins from any number of users.
If you mean hooking up multiple keyboards and monitors to the same system at the same time, why would you ever want to do that when you can just log in to the system over the network (text or GUI, your preference).