I'm at a bar, I've had a couple drinks, but nothing excessive. It's not late and I can safely get myself home as I have done in the past, but there's a plausible chance I'd get busted for a DUI if I got stopped on the way home. I'm a little buzzed and 0.001% over is all it takes. I check my new iPhone app and lo and behold, there's a checkpoint on the only highway between the bar and my house. I don't want to spend the night in jail, so I take a cab instead.
That app would save me money and jail time, save my district a bunch of paperwork, and make the roads safer.
The other side of the argument is that people will know where the checkpoint is and try to drive around it. If anything, this being open should encourage better checkpoint planning. There are plenty of high traffic bottlenecks in every state, so that's a poor excuse. Worst case scenario is the appropriate side roads would need increased patrols.
Plan to stop drinking X hours before leaving, or wait X hours longer, then leave. Is it that fucking hard? If you're not capable of thinking ahead that much, and you don't have a sober friend along to drive you - wait, clearly you wouldn't due to aforementioned lack of planning - stay off my god damned roads, or I hope you get caught. Eat a plate of chicken wings and order some soda you creeps, there is no excuse.
Is this an episode of the Simpsons or some shit? "Hey, you look like you're trying to sober up buddy, get the fuck out outa here!"
Get real. You're lack of planning is not an _excuse_ to drive under the influence. Holy fucking ball sweat.
What doesn't work with LVM, unfortunately, is using snapshots that allow you to "roll forward" if the system updates work out all right and you want to accept the updates into your main volume:-/
I bet you couldn't find five sysadmins in the (one one) country who do this in production.
Brtf alone isn't going to bring to Linux:
create new boot environment 'foo' mount 'foo' to/mnt install patches/updates/new software to root at/mnt umount 'foo' activate 'foo' reboot
Then from the boot menu, pick 'previous_foo' if 'foo' is broken.
That's a lot of really different projects that would have to come together for Linux to pull it off, one being dramatically different distro to distro. We've already heard folks shout "layering violation" at basic parts of ZFS itself, what hope does all this have?
these are minors, they have no right, even if they had been told, to sign off on this spying scheme
I don't think they even lend out TEXT BOOKS without parental agreement of responsibility for any damages and ugly brown paper bag covers.
You can bet your ass the parents had to sign _something_. Were they told about this monitoring, who knows, neither of us read the article I'm guessing.
What if Apple gave you the following choice: iOS laptop or tablet starting at $600 Mac OS X laptop or workstation starting at $3500 Would you shell out the $3500 to get Mac OS X? The way I see it, that is the choice you will have in the near future: iOS for a "consumer" level computer, and Mac OS X for high end "professional" level computers.
Yah, and what if auto manufacturers offered a horse for $500, or a car for $1,000,000?
I don't see how anyone who is familiar with computers could find iOS's gestures "intuitive." I actually had to look up how to create folders after iOS 4.0 hit. Drag one application on top of another? How does that bear any resemblance to a) how things are already done on Windows/OS X/Gnome/KDE/etc. or b) common sense?
In Android, conversely, you long-press (stand-in for right-click) and bam, "New folder" is right there in the menu that comes up. Just like you're already used to.
"long press" and context sensitive menu is more intuitive than dragging one application on top of another to group them?
My issue with the registry is it's lack of comments and relatively non-intuitive naming scheme. Even gconf-editor in gnome which reminds me a lot of regedit has comments. When I want to configure something textually, I just go to my home directory in the file manager, and look around for a file that is named something similar to the program I want to configure excepting being preceded with a "dot", i.e., a dot file and that's it. Just edit that file. It will probably be liberally commented so it's really not that hard to figure out what you're doing. For system wide config, look in the/etc directory. Same deal just without the dots.
Making it more user friendly is sort of against the whole point. It's an interface designed for programatic manipulation, like XML for example.
Your problems should be addressed with online documentation in the application layer, not in the backend configuration store which should be clean and concise for programatic access. The behavior of a setting will depend on the version of application code you're running, and face it, the documentation you get is going to be targeted at developers wherever you actually find it, because these interfaces are not designed for end users.
If a program leaves end users to deal directly with configuration data, it is just broke. For every XML/registry complaint I hear, I can find one application that _fails_ at usability.
Re:Perl - the COBOL of scripting languages
on
Perl 5.14 Released
·
· Score: 1
Nothing that couldn't be done better as a side effect to a rewrite to more modern and alive language
That just made me bang my head against my desk.
How about 'redesign' (if needed) instead of 'rewrite in a newer language' (because)?
Fedora? Can't accept a HD formatted without partitions. Ubuntu can and under Linux it is perfectly valid. Why use MS-DOS partitions on a modern system?
So other x86 software knows the disk is used for [partition id].
You're saving yourself _nothing_ by using disks this way.
If not a MBR partition, use EFI/GPT at least. You should only be using raw disks if you have software that needs it. Otherwise it is stupid, just like your post, and moderation score.
The mouse is almost 50 years old. You'll have to be way more specific about what's actually wrong with the interfaces that exist, standard or otherwise.
Um... UNIX, POSIX.. then all the non-standards since then. I didn't think I'd have to explain that..
99% of all tasks I want to schedule with something like Cron need to run hourly or daily. It is a one-liner to add an hourly, daily, weekly, or monthly cron job. Oh, and there's a cron.d on my system.
Ok, on all the systems where task foo is enabled, change the minute to random(60) Puppet can do this, and their crontab handler is not a simple piece of code. I imagine you simply cannot do this or put way too much faith into search/replace.
Erm... if you say so. I mean, I can also put my entire cron configuration in version control -- hell, I can put my entire/etc under version control, and I can get meaningful diffs/patches, comments, etc. Can you say the same for Windows configuration?
Flat files are great for all that stuff - versioning, backups, etc, but that isn't configuration management. You can't just diff two sudoers files and figure out the effective differences. Nor httpd.conf, nor sendmail.cf, etc x1000 diff config syntaxes. Logically it is IS easier to diff exported registry settings, or Java property files.
Heh, tell me if the bonding settings are identical for any two Linux network interfaces. Pssst! Check both modprobe.conf and ifcfg-*! Do they actually match the running configuration?
Write a script to detect wether your sudoers file is vulnerable to this: http://sudo.miscellaneousmirror.org/sudo/alerts/runas_group_pw.html Or pick any other soduers vulnerability from the past year. HINT: everything on that example line can be expanded into god-only-knows-what, so you will have to reimplement a good chunk of the sudoers parser. But hey, you can do "echo >>" so you're all set right?
Yes, and you can install Kerberos on Linux.
Yah, and I won't hold my breath while you find ten Linux apps that support Negotiate authentication, much less ISV apps.
I find it really bizarre that you can say that in the same breath. Nondeterministic... so other APIs are deterministic? What do you mean by that?
In the sense that across a datacenter with multiple slightly different Linux distros and different patch levels, and hundreds of different BARELY parseable configuration syntaxes, you NEED a tool like Puppet to get anything done somewhat reliably, and those guys do a LOT of work laying down a sensible interface to a pile of random crap. You cannot manage a Linux datacenter with just "echo >>" and grep. No, you can, but BADLY.
Probably couldn't get it from that one command, but I could check the status before or after the fact:
I meant yum update, and did that actually do anything. The easiest way is to check mtime on/var/log/yum.log afterwards, and "then should I reboot or not" is some fun with "lsof | grep path inode" It works but checking mtime on a log file to tell if an action was performed or not is STUUUUPID. This is a "unix" system right? Why is the output unparseable? Why no detailed return codes? 0/1 is the bare minimum anything should have, that does not give any application specific information.
Oh, and... OS updates... Say, when can I expect an API to deliver my app via Windows Update, hmm? At least updates to it, if not the full installation?
http://www.microsoft.com/systemcenter/en/us/overview.aspx That's addon software akin to Puppet but apparently way more advanced. So discounting addons, Linux wins on software distribution and loses on programatic configuration. Um... I'll just let you decide which one has more value and leave it at that. *goes back to tooling with custom Puppet module to manage something done easier on Windows...
Which API is that? Amazon's... I'm a Puppet admin, I know what can be achieved with some hard work, but it is no thanks to Linux vendors/maintainers. My issue is that the abstraction that makes this happen has to be bolted onto a unix system and if users modify the underlying configuration files the results are hard to predict. I assume Amazon's cloud resources have the same problems if you go outside their API to change things.
On the return codes, I asked how do you know if something got installed or not, 0/1 does not tell you this. 0 also means there was nothing to update (success). At the very minimum I expect 0/1 for any CLI utility, but yum's output is not suitable for processing by other commands, so there should be a lot more return codes. We get crap for feedback, and this is a core OS utility, and a core function. This problem however seeps all throughout Linux.
I wish Linux was more like Solaris in a lot of ways. Return codes, relative stability of any interfaces, stuff like that should be in all man pages not just really really old ones:\ I'm not saying Solaris is perfect, hehe, just a lot more mature.
You have a good point with Exchange vs. your linux mail server, but this is going outside the OS discussion and Exchange does a whole lot more than a typical unix mail server. Exchange even has an API to do "quiesce yourself so I can do a backup" where about any given Linux solution has "just grab the flat files and we'll pretend this integrity thing isn't really an issue". That is pretty typical of any Linux program.. flat files are "good enough" without further consideration of locking (visudo, vipw, etc, barf yack gross)
I love working with unix man, and I dig the whole everything is a file thing, but it's time to move on. We can keep configuration settings in flat files (hopefully the same syntax this time?), and still improve a LOOOOOOOT. You can have stable interfaces, command line AND programmatic AAAAAAND still have flat file backends with a universal syntax. You can even have a system where applications are notified when your configuration changes and still have flat files.
Careful, SSH and haphazard interfaces do not play in Linux's favor.
Day 1) Walk in the door, optimistic about what can be done with this "Enterprise" platform.
You know this works exactly the same way s/Windows/Linux/ and vice-versa
I use Puppet for doing Linux configuration management, and it is an _awesome_ tool but shows just how much Linux is lacking in basic system instrumentation, configuration management, and datacenter level administration. That is an area Windows has a leg up.
Kerberos + formal APIs beats SSH + only standardized interfaces are 30 years old Think what it takes to programmatically manage crontab entries for a minute, compared to something with an API for scheduled tasks built in. It's not even a contest, Linux is HORRIFIC at programatic configuration, you cannot lie about this.
If SSH is your idea of remote automation heaven, then I hate to tell you, but you can install that on Windows. There's a good chance you wouldn't need to if its kerberized WMI interface does what you need, and you can't just install "instrumentation" into Linux. And these are standard interfaces I'm talking about.. http://www.dmtf.org/
Once you get to the datacenter level, doing things programmatically is KEY. This SSH + messy nondeterministic CLI bullcrap doesn't cut it. Sorry, but you home Linux admins that think Linux is the tits needs to wake up. Microsoft is reeling right now, but Linux vendors are getting complacent. As complacent as Microsoft was...
At least document fucking return codes, and do something useful with them if you can't write out something parseable! Sun got that much right for Chirst's sake, and Linux fanbois like to pretend their are working on a UNIX system. Want an example??
man yum, tell me how you would know if a package installed or not when I do "yum -y update foo" That is the bullcrap I'm talking about. OS updates... not exactly some obscure interface huh?
We can argue the theory until we are blue in the face, but we can't ignore all the good software out there and the reality of what separates winners from losers.
There is a clear trend, "Let me show you how you should be doing that." and general ease of use.
We wouldn't be having this discussion if Data Protection Software wasn't a PITA to use. You can blame the customer and we'll be back here a few years having the same argument over why people still aren't using it correctly, or at all. Or, someone will figure out how to bake Data Protection Pie and steal everyone's lunch./food
So you complain that the software doesn't have default cooking instructions programmed in that would just magically make cookies or cupcakes without you having to do all that extra work.
Yah actually, that is what I'm complaining about. I want a cupcake, I'll settle for the recipe, but just the ingredients? Thanks for nothing.
Example: LDAP, Kerberos, DNS vs. Active Directory
Sure, you _could_ use the above technologies to accomplish what AD does, with a ton of time, and still not get to the point where ISVs can even dream of integrating with it. There are an infinite number of ways to implement an authentication/delegation/identity/system management/configuration management/service advertising solution etc., and then there is Active Directory. The cupcake won.
The problem is that the users can't be bothered figuring out what they want, so the software is at fault.
See, that's your problem right there. You can't just ask the users what they want. Steve Jobs was right when he said this "You can't just ask customers what they want and then try to give that to them. By the time you get it built, they'll want something new." My favorite illustration of this mistake is the Homer Car http://simpsons.wikia.com/wiki/The_Homer That is the wrong way to design _anything_, including software.
If you can't figure out how to turn your pile of crap ingredients into a cupcake, then what chance do your customers have? That is what "box of ingredients" software tells me, the authors have no idea what to do with it.
This story is one of the main reasons why. Instead of doing the job the government was created to do (protect individual rights from thieves, murderers, etc), the politicians/bureaucrats are the ones doing the infringing on those rights.
I love America because Mozilla can do what they did, and are right by it.
I also enjoy exercising my free speech and calling you a fucking idiot.
"If it were possible to have no government, we would do so. It is only to protect our rights that we resort to any government at all." - Thomas Jefferson.
"only to protect our rights" in no way minimizes the role of government.
What rights are you quibbling about, right to an Internet domain name? Sorry, not buying it. Name one right being oppressed, and who grants it if applicable.
BSD license is more free, but does not preserve the freedoms.
If somebody builds on your work and doesn't release it back to you, you don't lose anything. Effectively, there's no difference between that and if they had never even touched your stuff... which they wont do if they didn't want to have to share their changes with a GPL project anyway.
So you're only retaining contributors that are OK with sharing anyway and you're excluding people who do not want to give their modifications away openly and for $free. My idea of freedom is not "here is a free widget, but you can't improve and sell it, you can only give it away" - WTF?
This is strong-arming people into open source, just like the unnecessary association of $free with open. This isn't preservation, protection, nothing like that, it is attempting to SPREAD an ideal that has lately been starting to freak me out, and is counter intuitive to a healthy economy. There simply is no market demand for these ideals. GNU and FSF resort to this asshattery to attack a (once healthy) software market, forcing reimbursement for software development into areas that are unfeasible for small software businesses all for the sake of ideals that have zeeeeeeero demand in the marketplace. "Look at me, you can get a quick start on your project, for FREEE, there's just this uh, one string attached... you must support my agenda, mwahahahah! (evil Bowser laugh)"
Look, nobody uses Ubuntu because it has source code available. They use it because it's $free. I know everyone here knows this... "well duh, it has to be $free or nobody would use it and open source wouldn't advance" Why doesn't creep out more people?
Did you RTFA? This isn't Donkey Kong Jr. we're talking about here. DLP software, while extremely sophisticated, isn't that hard to use - What's difficult is the requirement for a company to create business policies that define what data is critical and what isn't. If you turn the alerts up too high, end-users and IT security are bombarded by noise and warnings, making the system useless. If you turn the alerts down too low, then you run the risk of data leakage.
WOW, that's funny how it suddenly becomes a business problem when this software shows up! A sane person would reason, if the software invented this problem, the software should fix it!
Christ, we're supposed to be SOLVING problems with computers! This reminds me of enterprise backup implementations and shaking down non-IT organizations for data retention policies. Like it's their job to analyze the risks of [not] having snapshots of their data from arbitrary points in time other than YESTERDAY.
These both clearly map to the real world and are not entirely an invention of IT folks right??
"It's a mature market - please turn it on." John Vecchi
Well if it's mature already, maybe it just sucks? Two years to implement a system that is 100% overhead, no services rendered! Fuck, that, shit. You're doing it wrong.
When will it catch on with software publishers & independent developers, that no matter how narrow your niche, there are very few excuses for utterly ignoring ease of use.
Free? : No. Expensive? : No. Really Expensive! : What are you smoking? It's just hard work? : DUH, that's why you set out to make a tool for it right, it doesn't have to be a GD requirement.
Linux users should not have a problem with AV. Even if they are smart enough not to need it. Linux users already think with a security focused mind, as an effect using Linux in lieu of a AV client is laziness on our part (granted, we can recognise an infected machine, so we can afford a bit of laziness).
I'm sorry, but you have absolutely nothing to back this up with. A lot of geeks use Linux, and most geeks can avoid infection by being careful. That doesn't make automatic binary blacklist/filtering useless, it's just generally not worth it considering the risk. It also doesn't make Linux users more safe than anyone else unless you pretend _ONLY_ geeks will ever be Linux users.
How can you categorically claim Mac users are irresponsible, then spout this. Which computer safety school do new Linux users go to?
Lets play a game! Here, have a thousand new Linux users from the education sector. OK, I'll play bad guy, GO!
Click here for free math worksheet printer for Linux 1000 low cost science experiments.pdf This document may not open automatically under Linux, in that event you will have to change the association by running chmod a+x *, then double clicking the file. Free icon pack for Linux Microsoft Office for Linux Make your Linux computer look like your old Windows one.pl Porn
VMware's explanation of events is troubling to me. The company as a whole is responsible for any of its failures. Internally the company could blame an individual but to shareholders and other vested entities an individual employee's failure is not something they care about. A better PR response would be to say that "we" made an unscheduled change or simply an unscheduled change was made to our infrastructure that caused X.
I submit they are more so, since they have a falsely inflated sense of security.
I submit that Mac users are safer with their feelings of security because they will avoid scareware, a huge threat to platforms perceived to be less secure.
So, scareware out of the Mac side of the equation, and all else being equal, who is safer randomly downloading crap off the Internet?
Slashdot Mirror
I'm at a bar, I've had a couple drinks, but nothing excessive. It's not late and I can safely get myself home as I have done in the past, but there's a plausible chance I'd get busted for a DUI if I got stopped on the way home. I'm a little buzzed and 0.001% over is all it takes. I check my new iPhone app and lo and behold, there's a checkpoint on the only highway between the bar and my house. I don't want to spend the night in jail, so I take a cab instead.
That app would save me money and jail time, save my district a bunch of paperwork, and make the roads safer.
The other side of the argument is that people will know where the checkpoint is and try to drive around it. If anything, this being open should encourage better checkpoint planning. There are plenty of high traffic bottlenecks in every state, so that's a poor excuse. Worst case scenario is the appropriate side roads would need increased patrols.
Plan to stop drinking X hours before leaving, or wait X hours longer, then leave. Is it that fucking hard? If you're not capable of thinking ahead that much, and you don't have a sober friend along to drive you - wait, clearly you wouldn't due to aforementioned lack of planning - stay off my god damned roads, or I hope you get caught. Eat a plate of chicken wings and order some soda you creeps, there is no excuse.
Is this an episode of the Simpsons or some shit? "Hey, you look like you're trying to sober up buddy, get the fuck out outa here!"
Get real. You're lack of planning is not an _excuse_ to drive under the influence. Holy fucking ball sweat.
Sure, you can sort of do that with LVM now.
What doesn't work with LVM, unfortunately, is using snapshots that allow you to "roll forward" if the system updates work out all right and you want to accept the updates into your main volume :-/
I bet you couldn't find five sysadmins in the (one one) country who do this in production.
Brtf alone isn't going to bring to Linux:
create new boot environment 'foo' /mnt /mnt
mount 'foo' to
install patches/updates/new software to root at
umount 'foo'
activate 'foo'
reboot
Then from the boot menu, pick 'previous_foo' if 'foo' is broken.
That's a lot of really different projects that would have to come together for Linux to pull it off, one being dramatically different distro to distro. We've already heard folks shout "layering violation" at basic parts of ZFS itself, what hope does all this have?
Please do AOL and Facebook next!
hate @ Internet
these are minors, they have no right, even if they had been told, to sign off on this spying scheme
I don't think they even lend out TEXT BOOKS without parental agreement of responsibility for any damages and ugly brown paper bag covers.
You can bet your ass the parents had to sign _something_. Were they told about this monitoring, who knows, neither of us read the article I'm guessing.
I'm good at that.
What if Apple gave you the following choice:
iOS laptop or tablet starting at $600
Mac OS X laptop or workstation starting at $3500
Would you shell out the $3500 to get Mac OS X? The way I see it, that is the choice you will have in the near future: iOS for a "consumer" level computer, and Mac OS X for high end "professional" level computers.
Yah, and what if auto manufacturers offered a horse for $500, or a car for $1,000,000?
What if I had a pony?
Jailbreaking is not really a security problem. Firstly, because "jailbreaking" just means allowing unsigned code to run.
Why don't you re-read that and tell me where your logic flaw is.
So, it's a security problem, except when your OS completely lacks code or driver signing, then its a feature. I see whats going on here.
I don't see how anyone who is familiar with computers could find iOS's gestures "intuitive." I actually had to look up how to create folders after iOS 4.0 hit. Drag one application on top of another? How does that bear any resemblance to a) how things are already done on Windows/OS X/Gnome/KDE/etc. or b) common sense?
In Android, conversely, you long-press (stand-in for right-click) and bam, "New folder" is right there in the menu that comes up. Just like you're already used to.
"long press" and context sensitive menu is more intuitive than dragging one application on top of another to group them?
What?
My issue with the registry is it's lack of comments and relatively non-intuitive naming scheme. Even gconf-editor in gnome which reminds me a lot of regedit has comments. When I want to configure something textually, I just go to my home directory in the file manager, and look around for a file that is named something similar to the program I want to configure excepting being preceded with a "dot", i.e., a dot file and that's it. Just edit that file. It will probably be liberally commented so it's really not that hard to figure out what you're doing. For system wide config, look in the /etc directory. Same deal just without the dots.
Making it more user friendly is sort of against the whole point. It's an interface designed for programatic manipulation, like XML for example.
Your problems should be addressed with online documentation in the application layer, not in the backend configuration store which should be clean and concise for programatic access. The behavior of a setting will depend on the version of application code you're running, and face it, the documentation you get is going to be targeted at developers wherever you actually find it, because these interfaces are not designed for end users.
If a program leaves end users to deal directly with configuration data, it is just broke. For every XML/registry complaint I hear, I can find one application that _fails_ at usability.
Nothing that couldn't be done better as a side effect to a rewrite to more modern and alive language
That just made me bang my head against my desk.
How about 'redesign' (if needed)
instead of 'rewrite in a newer language' (because)?
"Disinformation Wants to be FREE!"
Information, Anonymity, Anarchy, please meet Veracity, Authenticity, and Perception.
Have fun!
Fedora? Can't accept a HD formatted without partitions. Ubuntu can and under Linux it is perfectly valid. Why use MS-DOS partitions on a modern system?
So other x86 software knows the disk is used for [partition id].
You're saving yourself _nothing_ by using disks this way.
If not a MBR partition, use EFI/GPT at least. You should only be using raw disks if you have software that needs it. Otherwise it is stupid, just like your post, and moderation score.
The mouse is almost 50 years old. You'll have to be way more specific about what's actually wrong with the interfaces that exist, standard or otherwise.
Um... UNIX, POSIX.. then all the non-standards since then. I didn't think I'd have to explain that..
99% of all tasks I want to schedule with something like Cron need to run hourly or daily. It is a one-liner to add an hourly, daily, weekly, or monthly cron job. Oh, and there's a cron.d on my system.
Ok, on all the systems where task foo is enabled, change the minute to random(60)
Puppet can do this, and their crontab handler is not a simple piece of code. I imagine you simply cannot do this or put way too much faith into search/replace.
Erm... if you say so. I mean, I can also put my entire cron configuration in version control -- hell, I can put my entire /etc under version control, and I can get meaningful diffs/patches, comments, etc. Can you say the same for Windows configuration?
Flat files are great for all that stuff - versioning, backups, etc, but that isn't configuration management. You can't just diff two sudoers files and figure out the effective differences. Nor httpd.conf, nor sendmail.cf, etc x1000 diff config syntaxes. Logically it is IS easier to diff exported registry settings, or Java property files.
Heh, tell me if the bonding settings are identical for any two Linux network interfaces. Pssst! Check both modprobe.conf and ifcfg-*! Do they actually match the running configuration?
Write a script to detect wether your sudoers file is vulnerable to this: http://sudo.miscellaneousmirror.org/sudo/alerts/runas_group_pw.html
Or pick any other soduers vulnerability from the past year. HINT: everything on that example line can be expanded into god-only-knows-what, so you will have to reimplement a good chunk of the sudoers parser.
But hey, you can do "echo >>" so you're all set right?
Yes, and you can install Kerberos on Linux.
Yah, and I won't hold my breath while you find ten Linux apps that support Negotiate authentication, much less ISV apps.
I find it really bizarre that you can say that in the same breath. Nondeterministic... so other APIs are deterministic? What do you mean by that?
In the sense that across a datacenter with multiple slightly different Linux distros and different patch levels, and hundreds of different BARELY parseable configuration syntaxes, you NEED a tool like Puppet to get anything done somewhat reliably, and those guys do a LOT of work laying down a sensible interface to a pile of random crap. You cannot manage a Linux datacenter with just "echo >>" and grep. No, you can, but BADLY.
Probably couldn't get it from that one command, but I could check the status before or after the fact:
I meant yum update, and did that actually do anything. The easiest way is to check mtime on /var/log/yum.log afterwards, and "then should I reboot or not" is some fun with "lsof | grep path inode"
It works but checking mtime on a log file to tell if an action was performed or not is STUUUUPID.
This is a "unix" system right? Why is the output unparseable? Why no detailed return codes?
0/1 is the bare minimum anything should have, that does not give any application specific information.
Oh, and... OS updates... Say, when can I expect an API to deliver my app via Windows Update, hmm? At least updates to it, if not the full installation?
http://www.microsoft.com/systemcenter/en/us/overview.aspx
That's addon software akin to Puppet but apparently way more advanced. So discounting addons, Linux wins on software distribution and loses on programatic configuration. Um... I'll just let you decide which one has more value and leave it at that. *goes back to tooling with custom Puppet module to manage something done easier on Windows...
Really? How's Amazon doing it?
Which API is that? Amazon's...
I'm a Puppet admin, I know what can be achieved with some hard work, but it is no thanks to Linux vendors/maintainers.
My issue is that the abstraction that makes this happen has to be bolted onto a unix system and if users modify the underlying configuration files the results are hard to predict. I assume Amazon's cloud resources have the same problems if you go outside their API to change things.
On the return codes, I asked how do you know if something got installed or not, 0/1 does not tell you this.
0 also means there was nothing to update (success). At the very minimum I expect 0/1 for any CLI utility, but yum's output is not suitable for processing by other commands, so there should be a lot more return codes. We get crap for feedback, and this is a core OS utility, and a core function. This problem however seeps all throughout Linux.
I wish Linux was more like Solaris in a lot of ways. Return codes, relative stability of any interfaces, stuff like that should be in all man pages not just really really old ones :\ I'm not saying Solaris is perfect, hehe, just a lot more mature.
You have a good point with Exchange vs. your linux mail server, but this is going outside the OS discussion and Exchange does a whole lot more than a typical unix mail server. Exchange even has an API to do "quiesce yourself so I can do a backup" where about any given Linux solution has "just grab the flat files and we'll pretend this integrity thing isn't really an issue". That is pretty typical of any Linux program.. flat files are "good enough" without further consideration of locking (visudo, vipw, etc, barf yack gross)
I love working with unix man, and I dig the whole everything is a file thing, but it's time to move on. We can keep configuration settings in flat files (hopefully the same syntax this time?), and still improve a LOOOOOOOT. You can have stable interfaces, command line AND programmatic AAAAAAND still have flat file backends with a universal syntax. You can even have a system where applications are notified when your configuration changes and still have flat files.
Careful, SSH and haphazard interfaces do not play in Linux's favor.
Day 1) Walk in the door, optimistic about what can be done with this "Enterprise" platform.
You know this works exactly the same way s/Windows/Linux/ and vice-versa
I use Puppet for doing Linux configuration management, and it is an _awesome_ tool but shows just how much Linux is lacking in basic system instrumentation, configuration management, and datacenter level administration. That is an area Windows has a leg up.
Kerberos + formal APIs beats SSH + only standardized interfaces are 30 years old
Think what it takes to programmatically manage crontab entries for a minute, compared to something with an API for scheduled tasks built in.
It's not even a contest, Linux is HORRIFIC at programatic configuration, you cannot lie about this.
If SSH is your idea of remote automation heaven, then I hate to tell you, but you can install that on Windows. There's a good chance you wouldn't need to if its kerberized WMI interface does what you need, and you can't just install "instrumentation" into Linux. And these are standard interfaces I'm talking about .. http://www.dmtf.org/
Once you get to the datacenter level, doing things programmatically is KEY. This SSH + messy nondeterministic CLI bullcrap doesn't cut it.
Sorry, but you home Linux admins that think Linux is the tits needs to wake up. Microsoft is reeling right now, but Linux vendors are getting complacent. As complacent as Microsoft was...
At least document fucking return codes, and do something useful with them if you can't write out something parseable! Sun got that much right for Chirst's sake, and Linux fanbois like to pretend their are working on a UNIX system. Want an example??
man yum, tell me how you would know if a package installed or not when I do "yum -y update foo"
That is the bullcrap I'm talking about. OS updates... not exactly some obscure interface huh?
But how many actual exploits have ever appeared for apache?
Dude... Sony.
and lots http://lmgtfy.com/?q=apache+exploit
We can argue the theory until we are blue in the face, but we can't ignore all the good software out there and the reality of what separates winners from losers.
There is a clear trend, "Let me show you how you should be doing that." and general ease of use.
We wouldn't be having this discussion if Data Protection Software wasn't a PITA to use. You can blame the customer and we'll be back here a few years having the same argument over why people still aren't using it correctly, or at all. Or, someone will figure out how to bake Data Protection Pie and steal everyone's lunch. /food
So you complain that the software doesn't have default cooking instructions programmed in that would just magically make cookies or cupcakes without you having to do all that extra work.
Yah actually, that is what I'm complaining about. I want a cupcake, I'll settle for the recipe, but just the ingredients? Thanks for nothing.
Example: LDAP, Kerberos, DNS vs.
Active Directory
Sure, you _could_ use the above technologies to accomplish what AD does, with a ton of time, and still not get to the point where ISVs can even dream of integrating with it. There are an infinite number of ways to implement an authentication/delegation/identity/system management/configuration management/service advertising solution etc., and then there is Active Directory. The cupcake won.
The problem is that the users can't be bothered figuring out what they want, so the software is at fault.
See, that's your problem right there. You can't just ask the users what they want.
Steve Jobs was right when he said this "You can't just ask customers what they want and then try to give that to them. By the time you get it built, they'll want something new."
My favorite illustration of this mistake is the Homer Car http://simpsons.wikia.com/wiki/The_Homer
That is the wrong way to design _anything_, including software.
If you can't figure out how to turn your pile of crap ingredients into a cupcake, then what chance do your customers have? That is what "box of ingredients" software tells me, the authors have no idea what to do with it.
This story is one of the main reasons why. Instead of doing the job the government was created to do (protect individual rights from thieves, murderers, etc), the politicians/bureaucrats are the ones doing the infringing on those rights.
I love America because Mozilla can do what they did, and are right by it.
I also enjoy exercising my free speech and calling you a fucking idiot.
"If it were possible to have no government, we would do so. It is only to protect our rights that we resort to any government at all." - Thomas Jefferson.
"only to protect our rights" in no way minimizes the role of government.
What rights are you quibbling about, right to an Internet domain name? Sorry, not buying it. Name one right being oppressed, and who grants it if applicable.
I don't understand this.
BSD license is more free, but does not preserve the freedoms.
If somebody builds on your work and doesn't release it back to you, you don't lose anything. Effectively, there's no difference between that and if they had never even touched your stuff... which they wont do if they didn't want to have to share their changes with a GPL project anyway.
So you're only retaining contributors that are OK with sharing anyway and you're excluding people who do not want to give their modifications away openly and for $free. My idea of freedom is not "here is a free widget, but you can't improve and sell it, you can only give it away" - WTF?
This is strong-arming people into open source, just like the unnecessary association of $free with open. This isn't preservation, protection, nothing like that, it is attempting to SPREAD an ideal that has lately been starting to freak me out, and is counter intuitive to a healthy economy. There simply is no market demand for these ideals. GNU and FSF resort to this asshattery to attack a (once healthy) software market, forcing reimbursement for software development into areas that are unfeasible for small software businesses all for the sake of ideals that have zeeeeeeero demand in the marketplace. "Look at me, you can get a quick start on your project, for FREEE, there's just this uh, one string attached... you must support my agenda, mwahahahah! (evil Bowser laugh)"
Look, nobody uses Ubuntu because it has source code available. They use it because it's $free. I know everyone here knows this... "well duh, it has to be $free or nobody would use it and open source wouldn't advance"
Why doesn't creep out more people?
whatever happened to PCI Compliance
"Will you be compromised in the next twelve months?" is not part of a PCI audit.
Besides, PCI-DSS is 99.9% common sense - codified. It's not a magic barrier.
Did you RTFA? This isn't Donkey Kong Jr. we're talking about here. DLP software, while extremely sophisticated, isn't that hard to use - What's difficult is the requirement for a company to create business policies that define what data is critical and what isn't. If you turn the alerts up too high, end-users and IT security are bombarded by noise and warnings, making the system useless. If you turn the alerts down too low, then you run the risk of data leakage.
WOW, that's funny how it suddenly becomes a business problem when this software shows up! A sane person would reason, if the software invented this problem, the software should fix it!
Christ, we're supposed to be SOLVING problems with computers!
This reminds me of enterprise backup implementations and shaking down non-IT organizations for data retention policies. Like it's their job to analyze the risks of [not] having snapshots of their data from arbitrary points in time other than YESTERDAY.
These both clearly map to the real world and are not entirely an invention of IT folks right??
"can take two years to fully implement, he said."
"It's a mature market - please turn it on." John Vecchi
Well if it's mature already, maybe it just sucks?
Two years to implement a system that is 100% overhead, no services rendered! Fuck, that, shit. You're doing it wrong.
When will it catch on with software publishers & independent developers, that no matter how narrow your niche, there are very few excuses for utterly ignoring ease of use.
Free? : No.
Expensive? : No.
Really Expensive! : What are you smoking?
It's just hard work? : DUH, that's why you set out to make a tool for it right, it doesn't have to be a GD requirement.
Linux users should not have a problem with AV. Even if they are smart enough not to need it. Linux users already think with a security focused mind, as an effect using Linux in lieu of a AV client is laziness on our part (granted, we can recognise an infected machine, so we can afford a bit of laziness).
I'm sorry, but you have absolutely nothing to back this up with. A lot of geeks use Linux, and most geeks can avoid infection by being careful. That doesn't make automatic binary blacklist/filtering useless, it's just generally not worth it considering the risk. It also doesn't make Linux users more safe than anyone else unless you pretend _ONLY_ geeks will ever be Linux users.
How can you categorically claim Mac users are irresponsible, then spout this. Which computer safety school do new Linux users go to?
Lets play a game! Here, have a thousand new Linux users from the education sector. OK, I'll play bad guy, GO!
Click here for free math worksheet printer for Linux
1000 low cost science experiments.pdf
This document may not open automatically under Linux, in that event you will have to change the association by running chmod a+x *, then double clicking the file.
Free icon pack for Linux
Microsoft Office for Linux
Make your Linux computer look like your old Windows one.pl
Porn
VMware's explanation of events is troubling to me. The company as a whole is responsible for any of its failures. Internally the company could blame an individual but to shareholders and other vested entities an individual employee's failure is not something they care about. A better PR response would be to say that "we" made an unscheduled change or simply an unscheduled change was made to our infrastructure that caused X.
"Transparency is bad" +4 Insightful
What the... ?
I submit they are more so, since they have a falsely inflated sense of security.
I submit that Mac users are safer with their feelings of security because they will avoid scareware, a huge threat to platforms perceived to be less secure.
So, scareware out of the Mac side of the equation, and all else being equal, who is safer randomly downloading crap off the Internet?
I know, I know!!1