It's still a work in progress, and as such it pretty much has to get better.
The biggest annoyance I find is that the documentation is still in the "written for the developer" stage, which means everything else is much harder than it should be. The second problem is that because it parallelizes startup and daemons have no way to report back to systemd that they are fully started and ready to accept requests, it's possible that a service will finish start up before one that it requires is ready.
Luckily, if the service still has a file in/etc/rc.d/init.d, you can bypass systemd by setting a variable.
When the malfunctioning GPU card was replaced, the sudden restarts were gone.
I ran Windows 2000 for nearly a year on a motherboard that Intel recalled because of a bad memory controller. Their several week quoted turnaround for replacement was too long considering that I only saw about one BSOD per month.
LVD SCSI drives with 80-pin connectors, IDE CD-ROM drives, Pentium III motherboards, power supplies with only 20-pin motherboard power and 4-pin Molex connectors, AGP video cards.
I have all of these in my "soon to be thrown out" box, and could build a computer out of them, but I have no idea if it would run any modern OS.
In my "replacement parts" box, I have 10Gb Ethernet cards, Opteron processors, DDR3 RAM, etc. This is more what your "random parts" box consists of...modern hardware that you don't happen to be using.
Home and End should be line sensitive, not history sensitive. Every time I type on a Mac CLI and have to use the left arrow fifty times to edit a command, I want to scream.
In Linux, this is controlled by/etc/inputrc, where you can map the actual keystroke to the action you want it to perform, and bash will follow the settings.
This should work in BSD, as the Fedora 15/etc/inputrc has some entries for the "freebsd console".
Oh, you're recompressing it? Enjoy the additional quality loss.
These days, the codecs are good enough that if you use the same one at the same bit rate, it will result in almost no loss of quality, as the first compression removed all the "hard to compress" parts.
Yes, Google has destroyed an important feature of their search engine (using +term to mean that term absolutely has to appear on the page to count as a result) just to keep from businesses from asking for human-readable URLs in Google+.
That said, the RIAA should just stop whining and learn to use "free" to make money.
Companies like Netflix and Apple have shown that the RIAA and MPAA companies can still charge for their content and make a boatload of money.
The problem is that every idea that makes a ton of money selling content has come from someone other than the big media companies, who instead spend much of their income fighting against it.
Just like an Xbox, PS3, or Wii is a "play machine", and generally used the same way...to consume content that somebody else has created.
There's nothing wrong with that, and tablets will continue to exist, but there will never be a true laptop replacement in tablet form, because to become a laptop replacement (hundreds of gigabytes of storage, easy and fast input of information, enough RAM to run serious programs, etc.), it will lose the tablet form factor. There will likely be more devices like the Asus Transformer that can be a tablet or a laptop/notebook, but no true full-time tablets that can replace a real portable computer.
It doesn't take a genius to figure out that this site facilitates the distribution of materials against the rights holders wishes.
So does Google.
TPB is nothing but a more specialized version of Google. They don't host any of the content, and don't run a BitTorrent tracker. When someone downloads the latest "Harry Potter" the only thing involving TPB is searching for it. Google does exactly the same thing, but also happens to index a lot of file types other than torrents.
Eventually, somebody is going to build a full Internet search engine that has a "filter out all the non-torrent stuff" setting that is configurable like Google's "Safe Searrch". Getting such a general-purpose search engine shut down would be next to impossible.
$yum search web browser
dillo.i686 : Very small and fast GUI web browser
epiphany.i686 : Web browser for GNOME
firefox.i686 : Mozilla Firefox Web browser
seamonkey.i686 : Web browser, e-mail, news, IRC client, HTML editor
arora.i686 : A cross platform web browser
elinks.i686 : A text-mode Web browser
Not only that, but frequently (very frequently unfortunately) a candidate will start working on a solution that just will not work.
And you know their solution won't work because...oh, yeah, you're smarter than they are. I suspect the real issue is that you have not communicated all the requirements correctly, but since you are very familiar with the problem, you know more than you have told them, and that biases your view on their solution.
If they come up with a solution that works, I will purposely say "ok, now what if you need to do this...." which will be something that I know will require them to change their current design to accommodate.
And, unless the problem is still trivial, the only correct answer is "how about we meet tomorrow at 2pm, which will give me time to see if it can be done at all, and give you time to think about the change you requested and better define it so that we don't make the same mistake of not having all the requirements up front". Seriously, if you change your mind a few minutes after seeing something that meets all your requirements, you really do need to sit back and think about what the requirements really are before asking someone to waste money (somebody's...you, pretending to be the client, the company, etc.).
Not only that, but "needing help" is not the same as "working collaboratively". If they can't work collaboratively, that's a bad sign.
In this sort of a test, either they need more input from you, or they don't. You are simulating the client, and "collaboration" is the wrong word to use for getting requirements from the client, implementing them, repeat and rinse as necessary. The company I work for is small enough that programmers are likely heavily involved in requirements gathering, but for larger companies, that might be something only system analysts do. Depending on the job opening, your test might seem completely out of left field to a programmer who has never done formal analysis tasks.
To test "collaboration", you need to assign a current employee to be the candidates "team" and let them solve the problem jointly. Either that, or you can be the collaborator, but then don't get to be the client, too, which would mean no "surprises" to spring on them. Unless, of course, you want to simulate being a bad manager who doesn't pass on client requests right away and then blames the programmer for when deadlines aren't met.
Unchanging, no. Clearly written...generally not at first.
But, before we start anything, we have something written down and signed off by the client, so both sides know what is expected. If the client wants changes, they follow the same process, and then we get to change things like cost and due date (if necessary).
My point was that a "whiteboard test" where requirements are changing every minute isn't related to the real world in any way. And, if an interviewer said my solution wasn't acceptable because they forgot to mention "important requirement 27", then my response would be "you'll have to live with this code for right now, until we can talk about how that new requirement is going to impact the cost and delivery time".
Although this is mainly a personal project, I can't really see the point in purchasing closed source software when the open source alternatives support everything I need and more.
If you just need the hypervisor and not the enterprise features (live migration [including automated], high availability, resource management, virtual distributed networking, etc.), then ESXi is free (but not open source).
I admittedly am pretty new to this game but what advantages can I get from VMware when Xen runs a VPS totally seamlessly?
I'm not sure how many of the features are part of the "enterprise" package, but support for "thin" virtual disks is available to even the free hypervisor, and this can save a lot of disk space. I still like the VMware management interface better, and even the limited resource management without vSphere seems to be better and more intuitive. Also, the virtual hardware seems like a better idea and gives better performance than trying to virtualize real hardware. For example, the vmxnet3 NIC is much better suited for a VM than a virtual Intel e1000.
Since it's free, download it and try it out and see what you think.
I can't speak for vpshere 5, their newest version which isn't very old yet, but my experience with vsphere 4 is all but good. While making the client crossplatform using java is a strong business tactic, I just hate it. I've yet to find an enterprise java application that doesn't feel heavy, cumbersome, slow and horribly outdated.
I think you are a bit confused. The 4.x vSphere Client is a.NET application, not Java.
I haven't used vSphere 5 yet, so perhaps that has a Java client?
port=ftp/data
echo "The $port port is `grep $port/etc/services | awk '{print $2}'`"
That would lose lots of points for me, what with the "port" being the wrong name (ftp-data, although that's just a minor issue), the assumption that the place to find the info is a file in/etc (which it doesn't have to be, and this is more important, especially in larger companies), and not knowing that "getent entry databasename" has already done all the heavy lifting that deals with queries like this (e.g., following the settings in/etc/nsswitch.conf).
You'd really call it OK with that many mistakes in two lines?
I haven't coded a linked list in C in 10 years either, but I could still do it in my sleep.
I'm sure I could still program most of the algorithms and data structures I learned 20 years ago, but I (and many other people here) have learned we aren't as good as the people who do nothing but code that stuff for libraries.
Sounds like you know less than you think you do.
Sounds like you don't know why it's not important to be able to do things that other people already done for you, and far better than you ever could. I can't tell you how many programs I've seen that perform poorly because somebody decided to write their own version of a library routine. The only important thing to know about a linked list is how it performs compared to other data structures, and how to use that information to determine the best data structure to use for a particular application.
There is another problem: programmers tend to focus only on programming, and forget to take an interest in the domain where they code.
My data structures prof wanted to show the worst case when using a binary tree to store input so he used the Unix "words" file as input for one of the test cases.
Knowing that "English words" was the domain for input data, I reversed every input string before storing it.
Can he remain polite when frustrated? Can he admit that his first approach to a problem was wrong?
Why do you assume this will happen? Is it because you don't give them all the real requirements until after he takes a stab at the answer? That's a common trick that interviewers use to make themselves feel smarter than the candidate. What do you do when someone asks you to give them a real, written requirements list before they start, as they would get if this was a real work situation?
Can he work collaboratively at all, or does he have to solve problems entirely by himself.
If your "test" is simple enough to do on a whiteboard, many good programmers won't need any help after reading the written requirements.
Because knowing there's a big difference between strcat, strncat, strcpy, strncpy, wcscat, and wcsncat is important. One small "typo" on the whiteboad might mean the logic is completely wrong, but an IDE with online help and function/argument completion would allow you to make sure to use the right one in a real work situation.
Or, what if you use a method/function on an object that doesn't support that method? Will the tester know it wouldn't work, or would they just assume that any referenced function is implemented on a whiteboard in the library? Either one could lead to an incorrect evaluation.
Do you see that sandboxing makes harder to jailbreak, which is something that Apple desires, and would be of great importance to them if they do start to limit OS-X app installs to the app store.
I think part of the reason that Apple would like all OS-X software to flow through the app store is that even 10% of a $400 app with a million sales is a lot more money than 30% of a $2 app with a few million sales.
Slashdot Mirror
Actually, if you know who sells D&D books right now, you are playing recent versions of D&D, which means you are NOT a geek.
I know who sells the current books due to /. articles about how Wizards of the Coast screwed up D&D with the latest revisions.
We shall see what systemd brings us I suppose.
It's still a work in progress, and as such it pretty much has to get better.
The biggest annoyance I find is that the documentation is still in the "written for the developer" stage, which means everything else is much harder than it should be. The second problem is that because it parallelizes startup and daemons have no way to report back to systemd that they are fully started and ready to accept requests, it's possible that a service will finish start up before one that it requires is ready.
Luckily, if the service still has a file in /etc/rc.d/init.d, you can bypass systemd by setting a variable.
When the malfunctioning GPU card was replaced, the sudden restarts were gone.
I ran Windows 2000 for nearly a year on a motherboard that Intel recalled because of a bad memory controller. Their several week quoted turnaround for replacement was too long considering that I only saw about one BSOD per month.
... and the most important. What do you do when your hard drive fails?
Replace it and let the RAID array rebuild, just like on any OS.
What constitutes random parts then?
LVD SCSI drives with 80-pin connectors, IDE CD-ROM drives, Pentium III motherboards, power supplies with only 20-pin motherboard power and 4-pin Molex connectors, AGP video cards.
I have all of these in my "soon to be thrown out" box, and could build a computer out of them, but I have no idea if it would run any modern OS.
In my "replacement parts" box, I have 10Gb Ethernet cards, Opteron processors, DDR3 RAM, etc. This is more what your "random parts" box consists of...modern hardware that you don't happen to be using.
Home and End should be line sensitive, not history sensitive. Every time I type on a Mac CLI and have to use the left arrow fifty times to edit a command, I want to scream.
In Linux, this is controlled by /etc/inputrc, where you can map the actual keystroke to the action you want it to perform, and bash will follow the settings.
This should work in BSD, as the Fedora 15 /etc/inputrc has some entries for the "freebsd console".
Oh, you're recompressing it? Enjoy the additional quality loss.
These days, the codecs are good enough that if you use the same one at the same bit rate, it will result in almost no loss of quality, as the first compression removed all the "hard to compress" parts.
Go to google.com, type +pepsi
Get it yet?
Yes, Google has destroyed an important feature of their search engine (using +term to mean that term absolutely has to appear on the page to count as a result) just to keep from businesses from asking for human-readable URLs in Google+.
That said, the RIAA should just stop whining and learn to use "free" to make money.
Companies like Netflix and Apple have shown that the RIAA and MPAA companies can still charge for their content and make a boatload of money.
The problem is that every idea that makes a ton of money selling content has come from someone other than the big media companies, who instead spend much of their income fighting against it.
No, that would be a mellibinnium.
My spell check says that you really meant to say "molybdenum".
The iPad is their play machine.
Just like an Xbox, PS3, or Wii is a "play machine", and generally used the same way...to consume content that somebody else has created.
There's nothing wrong with that, and tablets will continue to exist, but there will never be a true laptop replacement in tablet form, because to become a laptop replacement (hundreds of gigabytes of storage, easy and fast input of information, enough RAM to run serious programs, etc.), it will lose the tablet form factor. There will likely be more devices like the Asus Transformer that can be a tablet or a laptop/notebook, but no true full-time tablets that can replace a real portable computer.
It doesn't take a genius to figure out that this site facilitates the distribution of materials against the rights holders wishes.
So does Google.
TPB is nothing but a more specialized version of Google. They don't host any of the content, and don't run a BitTorrent tracker. When someone downloads the latest "Harry Potter" the only thing involving TPB is searching for it. Google does exactly the same thing, but also happens to index a lot of file types other than torrents.
Eventually, somebody is going to build a full Internet search engine that has a "filter out all the non-torrent stuff" setting that is configurable like Google's "Safe Searrch". Getting such a general-purpose search engine shut down would be next to impossible.
Same for yum:
$yum search web browser
dillo.i686 : Very small and fast GUI web browser
epiphany.i686 : Web browser for GNOME
firefox.i686 : Mozilla Firefox Web browser
seamonkey.i686 : Web browser, e-mail, news, IRC client, HTML editor
arora.i686 : A cross platform web browser
elinks.i686 : A text-mode Web browser
Not only that, but frequently (very frequently unfortunately) a candidate will start working on a solution that just will not work.
And you know their solution won't work because...oh, yeah, you're smarter than they are. I suspect the real issue is that you have not communicated all the requirements correctly, but since you are very familiar with the problem, you know more than you have told them, and that biases your view on their solution.
If they come up with a solution that works, I will purposely say "ok, now what if you need to do this...." which will be something that I know will require them to change their current design to accommodate.
And, unless the problem is still trivial, the only correct answer is "how about we meet tomorrow at 2pm, which will give me time to see if it can be done at all, and give you time to think about the change you requested and better define it so that we don't make the same mistake of not having all the requirements up front". Seriously, if you change your mind a few minutes after seeing something that meets all your requirements, you really do need to sit back and think about what the requirements really are before asking someone to waste money (somebody's...you, pretending to be the client, the company, etc.).
Not only that, but "needing help" is not the same as "working collaboratively". If they can't work collaboratively, that's a bad sign.
In this sort of a test, either they need more input from you, or they don't. You are simulating the client, and "collaboration" is the wrong word to use for getting requirements from the client, implementing them, repeat and rinse as necessary. The company I work for is small enough that programmers are likely heavily involved in requirements gathering, but for larger companies, that might be something only system analysts do. Depending on the job opening, your test might seem completely out of left field to a programmer who has never done formal analysis tasks.
To test "collaboration", you need to assign a current employee to be the candidates "team" and let them solve the problem jointly. Either that, or you can be the collaborator, but then don't get to be the client, too, which would mean no "surprises" to spring on them. Unless, of course, you want to simulate being a bad manager who doesn't pass on client requests right away and then blames the programmer for when deadlines aren't met.
Unchanging, no. Clearly written...generally not at first.
But, before we start anything, we have something written down and signed off by the client, so both sides know what is expected. If the client wants changes, they follow the same process, and then we get to change things like cost and due date (if necessary).
My point was that a "whiteboard test" where requirements are changing every minute isn't related to the real world in any way. And, if an interviewer said my solution wasn't acceptable because they forgot to mention "important requirement 27", then my response would be "you'll have to live with this code for right now, until we can talk about how that new requirement is going to impact the cost and delivery time".
Although this is mainly a personal project, I can't really see the point in purchasing closed source software when the open source alternatives support everything I need and more.
If you just need the hypervisor and not the enterprise features (live migration [including automated], high availability, resource management, virtual distributed networking, etc.), then ESXi is free (but not open source).
I admittedly am pretty new to this game but what advantages can I get from VMware when Xen runs a VPS totally seamlessly?
I'm not sure how many of the features are part of the "enterprise" package, but support for "thin" virtual disks is available to even the free hypervisor, and this can save a lot of disk space. I still like the VMware management interface better, and even the limited resource management without vSphere seems to be better and more intuitive. Also, the virtual hardware seems like a better idea and gives better performance than trying to virtualize real hardware. For example, the vmxnet3 NIC is much better suited for a VM than a virtual Intel e1000.
Since it's free, download it and try it out and see what you think.
I can't speak for vpshere 5, their newest version which isn't very old yet, but my experience with vsphere 4 is all but good. While making the client crossplatform using java is a strong business tactic, I just hate it. I've yet to find an enterprise java application that doesn't feel heavy, cumbersome, slow and horribly outdated.
I think you are a bit confused. The 4.x vSphere Client is a .NET application, not Java.
I haven't used vSphere 5 yet, so perhaps that has a Java client?
So, it actually does navigation.
Based on what people say about the Apple Maps app, no, it doesn't.
port=ftp/data /etc/services | awk '{print $2}'`"
echo "The $port port is `grep $port
That would lose lots of points for me, what with the "port" being the wrong name (ftp-data, although that's just a minor issue), the assumption that the place to find the info is a file in /etc (which it doesn't have to be, and this is more important, especially in larger companies), and not knowing that "getent entry databasename" has already done all the heavy lifting that deals with queries like this (e.g., following the settings in /etc/nsswitch.conf).
You'd really call it OK with that many mistakes in two lines?
I haven't coded a linked list in C in 10 years either, but I could still do it in my sleep.
I'm sure I could still program most of the algorithms and data structures I learned 20 years ago, but I (and many other people here) have learned we aren't as good as the people who do nothing but code that stuff for libraries.
Sounds like you know less than you think you do.
Sounds like you don't know why it's not important to be able to do things that other people already done for you, and far better than you ever could. I can't tell you how many programs I've seen that perform poorly because somebody decided to write their own version of a library routine. The only important thing to know about a linked list is how it performs compared to other data structures, and how to use that information to determine the best data structure to use for a particular application.
There is another problem: programmers tend to focus only on programming, and forget to take an interest in the domain where they code.
My data structures prof wanted to show the worst case when using a binary tree to store input so he used the Unix "words" file as input for one of the test cases.
Knowing that "English words" was the domain for input data, I reversed every input string before storing it.
Can he remain polite when frustrated? Can he admit that his first approach to a problem was wrong?
Why do you assume this will happen? Is it because you don't give them all the real requirements until after he takes a stab at the answer? That's a common trick that interviewers use to make themselves feel smarter than the candidate. What do you do when someone asks you to give them a real, written requirements list before they start, as they would get if this was a real work situation?
Can he work collaboratively at all, or does he have to solve problems entirely by himself.
If your "test" is simple enough to do on a whiteboard, many good programmers won't need any help after reading the written requirements.
Because knowing there's a big difference between strcat, strncat, strcpy, strncpy, wcscat, and wcsncat is important. One small "typo" on the whiteboad might mean the logic is completely wrong, but an IDE with online help and function/argument completion would allow you to make sure to use the right one in a real work situation.
Or, what if you use a method/function on an object that doesn't support that method? Will the tester know it wouldn't work, or would they just assume that any referenced function is implemented on a whiteboard in the library? Either one could lead to an incorrect evaluation.
Do you see that sandboxing makes harder to jailbreak, which is something that Apple desires, and would be of great importance to them if they do start to limit OS-X app installs to the app store.
I think part of the reason that Apple would like all OS-X software to flow through the app store is that even 10% of a $400 app with a million sales is a lot more money than 30% of a $2 app with a few million sales.
But, the advantage is that you can say anything you want and claim "I was talking on my phone".