I have a reliable computer. It has been dead for 15 years with no signs of becoming bootable. Any software designed for this system can work perfectly, with 100% reliability, forever.
So the random bit flip you're not at high risk of, means that none happen? Derrrrrrrrrrr..... Oh, wait, my mistake, you're Anonymous Coward, it was my mistake to think you're more technical than a tree frog.
Fun fact: digital computers are made of analog parts, and each transistor has different performance. Random-seeming, unexpected bit flips can and do happen (dozens per day per host) just based on various minor manufacturing flaws and material inconsistencies. Rarely hurts anything important, I mean, unless you have a whole server farm, then you're having a small number of things go wonky all the time. If something only crashed one time, ever, you want it to have just restarted. You're not going to solve 100% of those in a large environment. The ones that call for resources and troubleshooting are the problems that happened 2 or more times, or the problems that line up with known causes, or that created a useful crash report in the logs.
Well, slashdot is a blog about news, not news, so the Law doesn't apply.
And in the news, headlines do ask real questions, and the reason the answer is always no is that if the answer is yes and it is news, there is no need for the question. Example: "Did City Council Pass New Law?" "City Council Passes New Law" When the answer is yes, the gimmick reduces the import of the story, while at the same time encouraging the reader to question if the reporter even knows the answer.
When the answer is no, that almost always means there is no story. It is the nature of news; it is about what did happen much more than what didn't happen. So if nothing happened, you can get some percent of people who are interested in the topic to read the story by phrasing it as a question. Many of those people would not read it if it said, "Nothing new to report; details follow." In this world of online advertising and click-metrics, the question-as-headline has increased in prominence, but the semantics haven't changed at all.
The Ask Slashdot I'd like to see is, "Can you say something bad about systemd that is both objective and true?" And then the answer would be no, and you'd have to read a bunch of subjective nonsense and lies to find that out.;)
Odd that of all possible choices, systemd developers chose almost always a way that is exactly the same or very similar to what Debian does. For example, I don't see stuff like/etc/sysconfig, while I do see things like/etc/hostname etc.
Everybody else has/etc/hostname too, sorry to pop your fanspiracy.
So basically, everybody with resources (such as skill+time, or money+desire) choose to migrate to systemd, and people without resources (no time or no technical skills, or no money) have to take the "scraps" (other people's work) and... they can't just keep using SysV init? Why, again?
No, they do have choices. Distros are real, they have real people working on them, and they choose package A or B based on their needs and philosophy. Keeping SysV init is an easy choice for a distro that doesn't care about the needs of sysadmins. Choosing systemd is also easy; it already exists and solves real problems.
Protip: if "your side" has a bunch of trolls name-calling the developer, double-check all technical claims. Don't just believe and repeat whatever FUD you hear that sounds technical, or came from a guy you're sure "knows something about... something."
They're all reminiscing about the old days where moving the audio from one output to another required restarting the app and restarting it with different command line options, and "per-application volume control" was the responsibility of the app, and if an app didn't have it, tough.
And integrating networking was impossible, you needed special server software that bypassed the sound system and did the networking by hand, and had its own mixers outside the sound ecosystem.
lol yeah, 15 years ago we were crying about how SysV sucks and somebody please replace it, and the replacements were all more broken than those shitty initscripts... until systemd came along and at a minimum, had a less broken design. OK, now that is a huge improvement; the suck is all related to implementation! That is suckiness that goes away, rapidly. YaY!
Nonsense, it is a load of crap. He even wants to throw away parallel startup. Why? "If all of the start-up processes are loaded sequentially, then order is really not a hard thing to sort out." Well, shucks Wilbur, this is like the guy that is against fuel injection because carburetors are easier to understand. Nevermind that fuel injection is better in every way, and that the technology is easy to work on for properly trained mechanics.
So, your brain can tell that you should start network file systems only after you've started the network, but how about the audio driver? Can you start named in one process while starting apache in another?
This is actually so much easier than you imagine... if understanding the dependencies is too hard for you, don't re-order them. You should not need to, it is not a normal part of system administration or software development. Let the good people who develop your distro worry about these hard parts. Trust that they hire or appoint people smart enough to understand their job, and get on with your life.
There is no new dependency "issue," there are just new capabilities.
If you'd rather understand what happens as your system starts up and not trust someone else's optimization, then systemd is not for you.
The bad news is, since a person with this concern doesn't understand how really any of the boot process and software works, distros are not for you and you should write your own because that is the only way you're going to understand what it does. The argument really falls on its face. If people understand what their computer is doing at that level, they can understand parallel startup. Sorry, they can.
We all still have the right to live under a bridge...
Here in the US every bridge is owned by a state or federal Transportation Department, and it is illegal to camp under them, or even to trespass under them in most cases. So no, you have no under-bridge rights.
With that many links they won't refute anything because nobody is going to click them. It is an obvious troll.
If there were just 3 or 4 links, then a person might think, "Oh, somebody researched it and found something contextual." If it is a giant list of links, it is more like, somebody did a search and pasted it. Which is just trolling, we all know how to do an internet search on our own. With that many links, I'll bet half of them repeat each other's language because they're repeats of the same source documents.
And the high percent of tumblr and youtube links makes it even more clear it is not serious information.
The vastly most likely answer is that it is pasted from some anti-feminist list.
I did check a couple of them and they were absurd crap, not anything relevant to this discussion, and not anything that a reasonable person would confuse with being relevant here.
It's the SJW ninnies that are trying to pretend that nerds are the perpetrators here when they are generally powerless and denigrated.
I find the idea that nerds would ever chase off women particularly amusing. Hell, most of us would KILL to have women around. If women are electing to not pursue the field, it's certainly not because they're unwelcome. On every team that I've ever been on with women, the guys went out of their way to be nice to them.
Actually, sorry to break it to you, but "OMG OMG OMG it's a female programmer she must be a nerd like us OMG OMG OMG AWESOME" is going to be very scary to most women. It is mal-adjusted and creepy. That it feels like a positive reaction to the guys who respond that way doesn't make it any less scary, it just makes it more difficult to correct. And any unusual behavior that is based on gender is going to be "unwelcome," it is not only intentionally unwelcoming behavior that is unwelcoming.
That there might be a co-worker who would "KILL" over having her around is... exceptionally unwelcome.
Actually you underscore the lack of socialization during programming. Your attempted counter example shows your social meter is differently calibrated than average people. You accept a very tiny bit of edge dialogue as a replacement for continued socialization all day in typical office jobs.
The loner might not be a hermit in the mountains, it doesn't change that the job is primarily solitary, even when coordinating with a large team.
Everyone I know that is either (1)uninterested in the systemd debate, or (2) actually in favor of systemd is (3) a completely incompetent hack in most areas that (4) have mostly bad ideas generally.
So rather than "Bullshit," your claim actually supports mine. The only ones who don't support it are the ones who don't care, or are idiots who already aren't listened to by their peers.
No, the "blowback" is just like #gamergate, a bunch of angry neckbeards who didn't even research the problem, and don't even know if their specific complaints are real things that happened, or hater propaganda they're repeating.
You'd almost think from the propaganda that systemd doesn't still allow the legacy SysV init scripts to run. But it turns out, they still work fine, and most of the distros with systemd haven't even ported many startup scripts, and even on a computer running systemd most of the startup is still done by the legacy initscripts.
I support systemd wholeheartedly, but I'm not going to port my old startup scripts. New software, sure, I'll learn the new stuff and integrate the way that is normal in the year that I'm writing the software. But backwards compatibility is a thing.
Your point seems to be, "new stuff requires new debugging methods on failure, avoid new stuff."
Like most sysadmins I generally agree. But also like most, I'd been waiting decades for SysV init to get replaced. It is worth learning something new every couple decades.
As an aside, most competent admins wouldn't be "resorting" to the kludge you describe, and will probably just stare and you and blink rapidly when told doing so is supposed to be a good thing.
No, a cromulent word has to be a made-up word that appears authoritative but isn't. (yet) "[B]oku" isn't cromulent at all, it is just a misspelling. I recommend you embiggen your vocabulary.
A reason for what, your lies and propaganda? I doubt it. The reason for the factoid you mention is that you made it up. Make up another one, it will have the same source, the same cause.
Slashdot Mirror
Yeah, make sure to write tests for every possible seemingly-random bit flip that could happen on any supported system configuration.
I have a reliable computer. It has been dead for 15 years with no signs of becoming bootable. Any software designed for this system can work perfectly, with 100% reliability, forever.
So the random bit flip you're not at high risk of, means that none happen? Derrrrrrrrrrr..... Oh, wait, my mistake, you're Anonymous Coward, it was my mistake to think you're more technical than a tree frog.
Fun fact: digital computers are made of analog parts, and each transistor has different performance. Random-seeming, unexpected bit flips can and do happen (dozens per day per host) just based on various minor manufacturing flaws and material inconsistencies. Rarely hurts anything important, I mean, unless you have a whole server farm, then you're having a small number of things go wonky all the time. If something only crashed one time, ever, you want it to have just restarted. You're not going to solve 100% of those in a large environment. The ones that call for resources and troubleshooting are the problems that happened 2 or more times, or the problems that line up with known causes, or that created a useful crash report in the logs.
Please try reading and fully understanding posts before responding.
If he comprehended your words, he wouldn't have had to post as a Coward. lol
Yes, he is. Coward.
Well, slashdot is a blog about news, not news, so the Law doesn't apply.
And in the news, headlines do ask real questions, and the reason the answer is always no is that if the answer is yes and it is news, there is no need for the question. Example: "Did City Council Pass New Law?" "City Council Passes New Law" When the answer is yes, the gimmick reduces the import of the story, while at the same time encouraging the reader to question if the reporter even knows the answer.
When the answer is no, that almost always means there is no story. It is the nature of news; it is about what did happen much more than what didn't happen. So if nothing happened, you can get some percent of people who are interested in the topic to read the story by phrasing it as a question. Many of those people would not read it if it said, "Nothing new to report; details follow." In this world of online advertising and click-metrics, the question-as-headline has increased in prominence, but the semantics haven't changed at all.
The Ask Slashdot I'd like to see is, "Can you say something bad about systemd that is both objective and true?" And then the answer would be no, and you'd have to read a bunch of subjective nonsense and lies to find that out. ;)
Odd that of all possible choices, systemd developers chose almost /etc/sysconfig, while I do see things /etc/hostname etc.
always a way that is exactly the same or very similar to what Debian does.
For example, I don't see stuff like
like
Everybody else has /etc/hostname too, sorry to pop your fanspiracy.
So basically, everybody with resources (such as skill+time, or money+desire) choose to migrate to systemd, and people without resources (no time or no technical skills, or no money) have to take the "scraps" (other people's work) and... they can't just keep using SysV init? Why, again?
No, they do have choices. Distros are real, they have real people working on them, and they choose package A or B based on their needs and philosophy. Keeping SysV init is an easy choice for a distro that doesn't care about the needs of sysadmins. Choosing systemd is also easy; it already exists and solves real problems.
Protip: if "your side" has a bunch of trolls name-calling the developer, double-check all technical claims. Don't just believe and repeat whatever FUD you hear that sounds technical, or came from a guy you're sure "knows something about... something."
They're all reminiscing about the old days where moving the audio from one output to another required restarting the app and restarting it with different command line options, and "per-application volume control" was the responsibility of the app, and if an app didn't have it, tough.
And integrating networking was impossible, you needed special server software that bypassed the sound system and did the networking by hand, and had its own mixers outside the sound ecosystem.
lol yeah, 15 years ago we were crying about how SysV sucks and somebody please replace it, and the replacements were all more broken than those shitty initscripts... until systemd came along and at a minimum, had a less broken design. OK, now that is a huge improvement; the suck is all related to implementation! That is suckiness that goes away, rapidly. YaY!
Nonsense, it is a load of crap. He even wants to throw away parallel startup. Why? "If all of the start-up processes are loaded sequentially, then order is really not a hard thing to sort out." Well, shucks Wilbur, this is like the guy that is against fuel injection because carburetors are easier to understand. Nevermind that fuel injection is better in every way, and that the technology is easy to work on for properly trained mechanics.
So, your brain can tell that you should start network file systems only after you've started the network, but how about the audio driver? Can you start named in one process while starting apache in another?
This is actually so much easier than you imagine... if understanding the dependencies is too hard for you, don't re-order them. You should not need to, it is not a normal part of system administration or software development. Let the good people who develop your distro worry about these hard parts. Trust that they hire or appoint people smart enough to understand their job, and get on with your life.
There is no new dependency "issue," there are just new capabilities.
If you'd rather understand what happens as your system starts up and not trust someone else's optimization, then systemd is not for you.
The bad news is, since a person with this concern doesn't understand how really any of the boot process and software works, distros are not for you and you should write your own because that is the only way you're going to understand what it does. The argument really falls on its face. If people understand what their computer is doing at that level, they can understand parallel startup. Sorry, they can.
We all still have the right to live under a bridge...
Here in the US every bridge is owned by a state or federal Transportation Department, and it is illegal to camp under them, or even to trespass under them in most cases. So no, you have no under-bridge rights.
With that many links they won't refute anything because nobody is going to click them. It is an obvious troll.
If there were just 3 or 4 links, then a person might think, "Oh, somebody researched it and found something contextual." If it is a giant list of links, it is more like, somebody did a search and pasted it. Which is just trolling, we all know how to do an internet search on our own. With that many links, I'll bet half of them repeat each other's language because they're repeats of the same source documents.
And the high percent of tumblr and youtube links makes it even more clear it is not serious information.
The vastly most likely answer is that it is pasted from some anti-feminist list.
I did check a couple of them and they were absurd crap, not anything relevant to this discussion, and not anything that a reasonable person would confuse with being relevant here.
Sorry greenhorn, looks like a troll to me.
Now get off my lawn!
It is creepy to be differently-nice to some people based on their gender, yes. WTF kind of place do you work at where nobody has ever mentioned it?
It is actually a bit of a "no-brainer."
It's the SJW ninnies that are trying to pretend that nerds are the perpetrators here when they are generally powerless and denigrated.
I find the idea that nerds would ever chase off women particularly amusing. Hell, most of us would KILL to have women around. If women are electing to not pursue the field, it's certainly not because they're unwelcome. On every team that I've ever been on with women, the guys went out of their way to be nice to them.
Actually, sorry to break it to you, but "OMG OMG OMG it's a female programmer she must be a nerd like us OMG OMG OMG AWESOME" is going to be very scary to most women. It is mal-adjusted and creepy. That it feels like a positive reaction to the guys who respond that way doesn't make it any less scary, it just makes it more difficult to correct. And any unusual behavior that is based on gender is going to be "unwelcome," it is not only intentionally unwelcoming behavior that is unwelcoming.
That there might be a co-worker who would "KILL" over having her around is... exceptionally unwelcome.
Actually you underscore the lack of socialization during programming. Your attempted counter example shows your social meter is differently calibrated than average people. You accept a very tiny bit of edge dialogue as a replacement for continued socialization all day in typical office jobs.
The loner might not be a hermit in the mountains, it doesn't change that the job is primarily solitary, even when coordinating with a large team.
Bullshit.
Everyone I know that is either (1)uninterested in the systemd debate, or (2) actually in favor of systemd is (3) a completely incompetent hack in most areas that (4) have mostly bad ideas generally.
So rather than "Bullshit," your claim actually supports mine. The only ones who don't support it are the ones who don't care, or are idiots who already aren't listened to by their peers.
No, the "blowback" is just like #gamergate, a bunch of angry neckbeards who didn't even research the problem, and don't even know if their specific complaints are real things that happened, or hater propaganda they're repeating.
You'd almost think from the propaganda that systemd doesn't still allow the legacy SysV init scripts to run. But it turns out, they still work fine, and most of the distros with systemd haven't even ported many startup scripts, and even on a computer running systemd most of the startup is still done by the legacy initscripts.
I support systemd wholeheartedly, but I'm not going to port my old startup scripts. New software, sure, I'll learn the new stuff and integrate the way that is normal in the year that I'm writing the software. But backwards compatibility is a thing.
The PaaS vendors are all excited about systemd. So that's just simply false. It is better for server.
Or, it is an important update and solves problems that are not related to the "desktop" or "server" roles of the box.
Your point seems to be, "new stuff requires new debugging methods on failure, avoid new stuff."
Like most sysadmins I generally agree. But also like most, I'd been waiting decades for SysV init to get replaced. It is worth learning something new every couple decades.
As an aside, most competent admins wouldn't be "resorting" to the kludge you describe, and will probably just stare and you and blink rapidly when told doing so is supposed to be a good thing.
Well, if that is all it is, then it is just a dishonest myth.
Of course, as a sysadmin I already knew that most sysadmins love systemd because it solves a whole basket of real problems.
No, a cromulent word has to be a made-up word that appears authoritative but isn't. (yet) "[B]oku" isn't cromulent at all, it is just a misspelling. I recommend you embiggen your vocabulary.
A reason for what, your lies and propaganda? I doubt it. The reason for the factoid you mention is that you made it up. Make up another one, it will have the same source, the same cause.
I'm still running the same timeseal binary for internet chess that I was running in the 90s. I'll bet the old *bsd builds still work, too.
It is over 15 years old.
I still sometimes run programs I wrote in 2001. I don't make any changes or upgrade anything, it all just still works.