I think that there are around 69 systemd programs now:) but that is according with the main goal of the systemd project to create common plumbing for linux systems. Before systemd every distribution used different plumbings so if you managed Red Hat systems, Debian systems and so on you had to know all the different locations of say the networking settings, we who develop daemons had to create completely different init scripts for each and every distribution and so on. And I understand that this is received with quite some excitement by the people who run lots of virtualized servers / containers but that is not an area that I'm the least interested in so I have not looked at it closely.
So systemd are replacing all the really low level system daemons just like GNU replaced all the Unix tools way back. Whether to use them by default or not will still be up to the maintainers of the various distributions but so far they have liked what they have seen (upcoming Ubuntu will for example replace dnsmasq with systemd-dns) but there will always be room for distributions like Gentoo or Devuan for those that want to use other tools (and I don't see Ubuntu removing the dnsmasq package for example)
The "by the system" installed unit files are stored in/lib/systemd/system yes, then when you enable a service (and this is a very welcoming thing with systemd that you can enable and disable services without having to complete remove all files) systemd creates a link in/etc/systemd/system/ and this is also where you are supposed to put your own custom unit files and where you put your changed unit files (so a new changed file from the system will not overwrite your custom changes).
And they are all in cleartext so I don't really get what you meant by that part? "systemctl edit" is not the only or mandatory way to edit unit files, it's just there as a helper so that you can do all things systemd with the systemctl command. So this is more choice, not less.
I have programmed Unix for over 20+ years now and I cannot see how anything with systemd is anti Unix or anti Posix at all. All configs are clear text, it's still the old one program does one thing (which is why there are so many different systemd programs). The only thing that I can find (and the one thing that I'm not 100% agreeing with) is the binary logging by journald, I don't like it but I do understand why the systemd devs went in that direction and the people who have compared this with the Windows System Event logging have never ever had the misfortune to code for that piece of shit (I have). And it's note like there was never a binary format of anything throughout the Unix history.
Good question. We now for sure that they do not compile them with an out of the box version of Visual Studio since all their programs link with the C runtime provided with Windows (msvcrt.dll) and not with the compiler specific version that you have to install separately (vcredist) so we know that they "do something" with Visual Studio that we normal mortals do not.
Locating and editing the unit files are actually quite easy once you learn how to use systemd. If you write for example "systemctl status mediatomb.service" (and the name of the unit file can be tab completed so it's not all that if one does not remember the full name), then we get output like:
mediatomb.service - UPnP MediaServer
Loaded: loaded (/lib/systemd/system/mediatomb.service; enabled; vendor preset: enabled)
Active: active (running) since fre 2016-06-10 13:49:40 CEST; 6h ago
So after "Loaded:" we get the full path of the unit file so locating them are actually quite easy, then there is "systemctl edit mediatomb.service" for editing the unit file directly via the systemctl command.
That logind have made other programs creating a hard dependency on logind is of course unfortunate for people who wants to run other init's but that is simply due to logind providing support for things that the developers of those other programs wants (otherwise they would not have used them). And this is nothing new in Unixland, there are several programs that you cannot move from AIX to Solaris or from Solaris to HPUX and so on, and there are non systemd versions of logind that are being worked on so even this will probably (i.e if enough developers care about it) be a solved problem.
It does not take Klaus one hour to attempt. He spends the first 20 minutes discussing all the good things that systemd have introduced. Then he takes one minute to discuss why his specific use case (booting knoppix on dvds) rather would be without any init system at all. Then he for 20 minutes demonstrates three different ways how systemd could be removed from Ubuntu. So this is not an attempt, this is a demonstration, and the problems that he shows in demonstration #1 is because Ubuntu have put a hard dependency on systemd in their ubuntu-core package. The problem with demonstration #2 is that the version of Network Manager in Ubuntu uses the user sessions from logind and that the init script to remount root to read-writable have been removed by Ubuntu some time ago.
I would say that none of this was directly strange considering that Ubuntu have not used sysv for over 10 years now (6.10 was the first version with Upstart), so naturally there are no one that tests that everything works fine with sysv in Ubuntu any more. And that all the problems encountered where not from systemd itself but from the fact that systemd have been seen as providing such great functionality that developers of other software have begun to build in dependencies for it in their software. This is akin to complaining that glibc is bad because it provides things that say musl does not so we now have quite a few programs that only works with glibc.
Slashdot? You perhaps mean Slackware? Other than those every other distribution used sysv like Debian, Red Hat, SUSE, Ubuntu (prior to Upstart and Upstart is a POS in comparison with systemd) and so on. SysV was the primary init system on most Linux distributions until systemd came along. Yes there where some other smaller ones like OpenRC and Upstard but they never took off.
No there has never even been any "systemd or nothing" from the systemd people. You have no problem to apt-get remove all the systemd packages on Debian and replace it with sysv. HOWEVER developers of other software like Gnome saw what systemd provided, saw that it was good and thus begun to make use of the specific features that it provided (logind in this case). So what you now claim is that it's the systemd developer fault for creating something that other people wants to use...
And sysv init's failure path could lead to infinite loops on boot and shutdown (I have experienced both), I must say that so far I have found systemd much easier to find out why a daemon fails or not as compared with sysv.
Indeed, for example here in Sweden we have had several politicians that had to leave office due to them expressing anti-Semitic ideas. I have however not seen a single one being forced out due to expressing anti-Muslim ideas. And looking at television, news papers and discussion forums it seams that Muslims are a free for all target at the moment throughout Europe.
And even then (assuming that it has passed through parliament) it also would have to be put into the law by each member state. There are still some EU stated laws that as of yet (and probably never will) have not made it into the laws of my country.
The EU is a trade union so of course it's in their interest to lessen the burdens of trade and fake reviews are such a burden. That said there is nothing in TFA that implies that this is a proposal for a new law at all.
Of course the deals have to be renegotiated. All the trade deals have been towards the EU so if Britain leaves then they have to sign new trade deals since the existing ones are no longer valid for a Britain that is not part of the EU. With a Brexit you throw out all those free-trade agreements that exists within the EU and you have to go back to the old days where you had to negotiate such deals country by country.
Yes, but only on appointment. Once you are part of the commission the parliament cannot dismiss you, they can of course dismiss the entire commission but going that route to dismiss a single commissionaire sounds unlikely (and has not happened yet).
No that is not what is happening, daemon-reload simply instructs systemd that it should look at all the unit files and see if one of them have changed. Since Raspbian have chosen to use the systemd-dhcp client this client have a dependency on/etc/network/interfaces and the daemon-reload simpy makes systemd understand that the dhcp client is no longer needed and thus shuts it down.
This would have all been described to the parent AC if he had run "systemctl status networking" after he edited/etc/network/interfaces. Instead he assumed that this would all be extremely complex and started to go over the unit files and whatever.
I have never understood why the religious fundamentalists are so hell bent on implementing oppressive laws into government, don't they realize that at some point in time "the other side" will be in charge and then all those laws will be used against them, so it should be in their best interests to minimize the number of abusive laws.
If so then surely the logs would contain rows like "enable autopilot." and "autopilot disengaged due to strange error". Also it's not like the car is totalled into a 2x2m cube so testing can be done on it.
Such accelerations is not unheard of yes but have there even been a single case where it has been proven to be the fault of the car? And do not 99% of these cases always comes from the US? A country with a judicial system which encourages fraud like this.
That is only if you truly believe that the other party would bet $30000 on a 10% of winning. And to be honest I have heard of very very few such cases over here, now the loser pays is of course not the only difference towards to US system so other factors might make a play in that as well.
Slashdot Mirror
I think that there are around 69 systemd programs now :) but that is according with the main goal of the systemd project to create common plumbing for linux systems. Before systemd every distribution used different plumbings so if you managed Red Hat systems, Debian systems and so on you had to know all the different locations of say the networking settings, we who develop daemons had to create completely different init scripts for each and every distribution and so on. And I understand that this is received with quite some excitement by the people who run lots of virtualized servers / containers but that is not an area that I'm the least interested in so I have not looked at it closely.
So systemd are replacing all the really low level system daemons just like GNU replaced all the Unix tools way back. Whether to use them by default or not will still be up to the maintainers of the various distributions but so far they have liked what they have seen (upcoming Ubuntu will for example replace dnsmasq with systemd-dns) but there will always be room for distributions like Gentoo or Devuan for those that want to use other tools (and I don't see Ubuntu removing the dnsmasq package for example)
The "by the system" installed unit files are stored in /lib/systemd/system yes, then when you enable a service (and this is a very welcoming thing with systemd that you can enable and disable services without having to complete remove all files) systemd creates a link in /etc/systemd/system/ and this is also where you are supposed to put your own custom unit files and where you put your changed unit files (so a new changed file from the system will not overwrite your custom changes).
And they are all in cleartext so I don't really get what you meant by that part? "systemctl edit" is not the only or mandatory way to edit unit files, it's just there as a helper so that you can do all things systemd with the systemctl command. So this is more choice, not less.
I have programmed Unix for over 20+ years now and I cannot see how anything with systemd is anti Unix or anti Posix at all. All configs are clear text, it's still the old one program does one thing (which is why there are so many different systemd programs). The only thing that I can find (and the one thing that I'm not 100% agreeing with) is the binary logging by journald, I don't like it but I do understand why the systemd devs went in that direction and the people who have compared this with the Windows System Event logging have never ever had the misfortune to code for that piece of shit (I have). And it's note like there was never a binary format of anything throughout the Unix history.
Good question. We now for sure that they do not compile them with an out of the box version of Visual Studio since all their programs link with the C runtime provided with Windows (msvcrt.dll) and not with the compiler specific version that you have to install separately (vcredist) so we know that they "do something" with Visual Studio that we normal mortals do not.
mediatomb.service - UPnP MediaServer
Loaded: loaded (/lib/systemd/system/mediatomb.service; enabled; vendor preset: enabled)
Active: active (running) since fre 2016-06-10 13:49:40 CEST; 6h ago
So after "Loaded:" we get the full path of the unit file so locating them are actually quite easy, then there is "systemctl edit mediatomb.service" for editing the unit file directly via the systemctl command.
That logind have made other programs creating a hard dependency on logind is of course unfortunate for people who wants to run other init's but that is simply due to logind providing support for things that the developers of those other programs wants (otherwise they would not have used them). And this is nothing new in Unixland, there are several programs that you cannot move from AIX to Solaris or from Solaris to HPUX and so on, and there are non systemd versions of logind that are being worked on so even this will probably (i.e if enough developers care about it) be a solved problem.
It does not take Klaus one hour to attempt. He spends the first 20 minutes discussing all the good things that systemd have introduced. Then he takes one minute to discuss why his specific use case (booting knoppix on dvds) rather would be without any init system at all. Then he for 20 minutes demonstrates three different ways how systemd could be removed from Ubuntu. So this is not an attempt, this is a demonstration, and the problems that he shows in demonstration #1 is because Ubuntu have put a hard dependency on systemd in their ubuntu-core package. The problem with demonstration #2 is that the version of Network Manager in Ubuntu uses the user sessions from logind and that the init script to remount root to read-writable have been removed by Ubuntu some time ago.
I would say that none of this was directly strange considering that Ubuntu have not used sysv for over 10 years now (6.10 was the first version with Upstart), so naturally there are no one that tests that everything works fine with sysv in Ubuntu any more. And that all the problems encountered where not from systemd itself but from the fact that systemd have been seen as providing such great functionality that developers of other software have begun to build in dependencies for it in their software. This is akin to complaining that glibc is bad because it provides things that say musl does not so we now have quite a few programs that only works with glibc.
Well each to their own, as a writer of init scripts / unit files and manager of servers I atleast find systemd to be way way better than upstart.
Slashdot? You perhaps mean Slackware? Other than those every other distribution used sysv like Debian, Red Hat, SUSE, Ubuntu (prior to Upstart and Upstart is a POS in comparison with systemd) and so on. SysV was the primary init system on most Linux distributions until systemd came along. Yes there where some other smaller ones like OpenRC and Upstard but they never took off.
No there has never even been any "systemd or nothing" from the systemd people. You have no problem to apt-get remove all the systemd packages on Debian and replace it with sysv. HOWEVER developers of other software like Gnome saw what systemd provided, saw that it was good and thus begun to make use of the specific features that it provided (logind in this case). So what you now claim is that it's the systemd developer fault for creating something that other people wants to use...
And sysv init's failure path could lead to infinite loops on boot and shutdown (I have experienced both), I must say that so far I have found systemd much easier to find out why a daemon fails or not as compared with sysv.
Not to mention the problem with GUI guides when the user is using a completely different language on their desktop than you do.
How could the Ubuntu version schema be any more simple that what it already is? 16.10 as in released in 2016-10.
Indeed, for example here in Sweden we have had several politicians that had to leave office due to them expressing anti-Semitic ideas. I have however not seen a single one being forced out due to expressing anti-Muslim ideas. And looking at television, news papers and discussion forums it seams that Muslims are a free for all target at the moment throughout Europe.
And even then (assuming that it has passed through parliament) it also would have to be put into the law by each member state. There are still some EU stated laws that as of yet (and probably never will) have not made it into the laws of my country.
The EU is a trade union so of course it's in their interest to lessen the burdens of trade and fake reviews are such a burden. That said there is nothing in TFA that implies that this is a proposal for a new law at all.
Of course the deals have to be renegotiated. All the trade deals have been towards the EU so if Britain leaves then they have to sign new trade deals since the existing ones are no longer valid for a Britain that is not part of the EU. With a Brexit you throw out all those free-trade agreements that exists within the EU and you have to go back to the old days where you had to negotiate such deals country by country.
Yes, but only on appointment. Once you are part of the commission the parliament cannot dismiss you, they can of course dismiss the entire commission but going that route to dismiss a single commissionaire sounds unlikely (and has not happened yet).
No that is not what is happening, daemon-reload simply instructs systemd that it should look at all the unit files and see if one of them have changed. Since Raspbian have chosen to use the systemd-dhcp client this client have a dependency on /etc/network/interfaces and the daemon-reload simpy makes systemd understand that the dhcp client is no longer needed and thus shuts it down.
This would have all been described to the parent AC if he had run "systemctl status networking" after he edited /etc/network/interfaces. Instead he assumed that this would all be extremely complex and started to go over the unit files and whatever.
So you really had to do all that to come up with:
nano /etc/network/interfaces
sudo systemctl daemon-reload
Looks like the real problem is that the Raspbian guides are not updated with that crucial last step
Addicted to lube is that they are.
If anything it's the vocal anti systemd trolls that are causing this self-destruct that you are talking about.
I have never understood why the religious fundamentalists are so hell bent on implementing oppressive laws into government, don't they realize that at some point in time "the other side" will be in charge and then all those laws will be used against them, so it should be in their best interests to minimize the number of abusive laws.
The car should be available for 3d party examination?
If so then surely the logs would contain rows like "enable autopilot." and "autopilot disengaged due to strange error". Also it's not like the car is totalled into a 2x2m cube so testing can be done on it.
Such accelerations is not unheard of yes but have there even been a single case where it has been proven to be the fault of the car? And do not 99% of these cases always comes from the US? A country with a judicial system which encourages fraud like this.
Whoops :-)
That is only if you truly believe that the other party would bet $30000 on a 10% of winning. And to be honest I have heard of very very few such cases over here, now the loser pays is of course not the only difference towards to US system so other factors might make a play in that as well.