2. We switch to different packaging. Instead of making disks larger we cram more of them into the same space similar to CPU cores - same MTBF per disk but lots of them presented out by one physical interface.
Um... isn't that what RAID does ? What you describe would just move the the the RAID controller inside the drive enclosure rather than on the PCI bus.
Unless you were thinking that we create a 10Tbyte disk from 10 x 1Tbyte discs, so if one fails you would only have to replicate 1Tbyte of data rather than the whole 10Tbyte. In which case, the RAID controller would have to be able to 'see' inside the 10Tbyte virtual disc to know which of the internal discs had failed and what needed replicating. So a 10Tbyte 'virtual' disc created by LVM gluing 10 x 1Tbyte RAID 1 arrays together to make them look like one large 10Tbyte disc ?.... all in one little box that could overheat, driven by one power supply that could fail or spike damaging the LVM or RAID controller chip corrupting the whole lot.
If you are using these websites to watch television programmes that are not being shown on TV at the same time. This is often described as a "catch up" service.
To view video clips on the internet, as long as what you are viewing is not being shown on TV at the same time as you are viewing it.
We don't own a TV.
We do watch BBC programs from the iPlayer site, but only after they are broadcast (using standard Flash plugin NOT the DRM AdobeAir version).
which means that technically we don't have to pay for a TV license. However we decided we would because we like what they produce and are happy to contribute something.
IF the BBC start to add DRM to everything, we will probably reconsider.
Cloud computing is just proprietary computing by another name. It can still be useful, but the control lies with the cloud owner rather than the user.
If you see cloud compute as a 'software service' system like Google Mail or Google Docs, then perhaps yes. We use cloud compute in the 'generic virtual machine provider' sense.
We use a couple of cloud compute providers to host our web servers. We pay a monthly fee and get the root password to a standard Linux virtual machine, what we install on it is up to us. If the physical hardware that is hosting our server fails, we just install it somewhere else. All of the install process is automated, and we can transfer our server and data from one provider to another in a couple of hours*.
We can alter the machine resources (cpu, memory disc) whenever we like. We review the costs on a regular basis and can move to another provider if think we can get a better deal**.
* Installing from backup on to a new machine takes about 10min, the biggest delay is waiting for cached DNS records to catch up. ** We currently use two providers, one in the US and one in the UK and currency exchange rates can make a significant difference.
Why a USB connector ? That causes the same problem as making SSD cards use the SATA interface - the serial interface becomes slower than the things it is connected to.
What I would like to see is a set of sockets on the motherboard, mapped into the main memory address space (not PCI), a physical switch on the board to make them read only and software in the BIOS to make them look like a bootable disk.
Four sockets with 16 or 32G in each would give you enough space to store the entire OS. I don't know how Windows would handle it, but in a Unix or Linux based system it would be fairly easy to mount the devices as read only partitions and map them into the filesystem. This would be ideal for a server system, mapping the entire OS into the main memory address space and making it read only.
In fact all the BIOS would need to do is make the first 100M visible as a boot partition, and leave the OS to handle the rest.
First, let me say thank you for responding to this thread. It is useful to have a honest and informative debate about this.
I often start my day by reading the SlashDot front page. I pick a couple of summaries that look interesting and normally open both the SlashDot page and the article it links to in separate tabs. However, if the SlashDot summary mentions multiple pages, login required, excessive advertising etc. I only open the SlashDot page and read the comments to see if it is worth reading TFA later. In this case Kdawson's warning meant I only opened the SlashDot page. I got enough information from the comments to decide that it probably wasn't worth opening TFA right now. I am interested in SSD devices, but the number of negative comments about the site readability put me off.
As I read more of the comments, your post as the site owner and the subsequent discussion actually became more interesting than the article itself. So, I did visit your site - not to find out more about SSD devices, but to see what all the fuss was about. The immediate impression I got is that this site isn't aimed at me, it is aimed a kids putting together games machines. I've visited many similar sites and my normal reaction would be to just skip it and move on to somewhere else. This time I stayed on the site and tried to read the article because I was curious to find out how intrusive the advertisements actually were.
The information you are presenting is detailed and technical but I found it very difficult to concentrate on the text because the flashing animated advertisements kept jumping around in my peripheral vision. I looked at a few pages but it was too tiring/irritating to actually bother with reading all of the pages. For me it isn't the number of advertisements, their position or content. The problem was the flashing animated advertisements vying for my attention meant I couldn't actually concentrate on the article I came to read. The insistent animated advertisements actually worked against themselves. In trying to read article text I had to actively concentrate on excluding them - so I have no idea what they were actually trying to sell. My own experience is that if I do actually register what company or product the advertisement is for, I am left with a negative impression of them - I remember the irritation factor more than the benefits the advertisement was trying to portray.
Last but not least - I opened your site in a separate tab, and then flipped back to the SlashDot page to compare it with what people were saying. At that point I noticed that the animations were taking up so much cpu that scrolling the SlashDot page was slow and clunky. The only way to restore my web browser to a usable state was to close the tab with your site. This is important, because when reading a story or article with links to other sites I normally open all of the associated pages in tabs and flip between them to compare and understand what they are saying. The animated advertisements meant I visited your site once, read a couple of pages and then closed the tab.
Based on my experience, unless it was providing really important information that no one else provided, then I wouldn't visit it again. My apologies, I hope you don't take this personally, it was meant as an unbiased report of user experience.
the original hypervisor was CP/CMS, developed at IBM in the 1960s
The major UNIX vendors, including Sun Microsystems, HP, IBM, and SGI, have been selling virtualized hardware since before 2000
As of late 2006, Solaris, Linux (Ubuntu and Gentoo), and FreeBSD have been ported to run on top of Hypervisor
Microsoft doesn't appear in the list until 2008, so they can hardly claim to be 'one of the most innovative companies' in this particular field.
.. and now they're trying to put out drivers for Linux under a license that guarantees they will be open and free
IF this is the case, then I agree, this move should be welcomed.
.. At every University they presented studies comparing open and closed source software.
Are you basing your statements on what the University studies said, or do you have personal experience of working with both open and closed source software yourself ?
Closed source wins on almost every level; fewer errors, quicker fixes, better performance.
As always your mileage may vary, and if this is what you have found then then this is good.
This hasn't been my experience having working in IT developing both open source and closed source software.
Many people find that closed source software does exactly what they need, which is fine by me.
Best tool for the job and all that.
Open source is a great ideal but just because someone COULD go through and edit and contribute code doesn't mean someone WILL go through and fix the errors
Having worked in (closed source) commercial software for a number of years, I know from experience that the same applies there too.
Wherever you are, priorities are subject to a cost benefit analysis.
Just becuse we COULD fix a bug, doesn't mean someone WILL go through and fix the errors, until an important customer notices it.
It could also be argued that Microsoft were able to contribute because the Linux kernel is open source.
How many people have the opportunity to see and contribute to the code for the Windows kernel ?
.. very few people are going to sit day after day and produce quality code and products if they don't get paid for it..
Indeed, and some of the best open source software is produced by people explicitly employed to do just that. By companies like IBM, Sun, Redhat, Oracle, Canonical... and many many more.
I am currently working on a project funded by the UK government to produce open source software for the science community to use.
Microsoft has contributed more to computing than any single company..
Um, not convinced on that.. I suspect that companies like IBM might be a better contender for that title.
Not only do they have a longer history and a wider range, they have also contributed some their patents to the community.
.. but everyone on this site just loves to hate them
... virii like aids or the swine flu or aids stop damaging the body and clean up after them... it would be beneficial to the virii too as their host would live longer and they would be in symbiotic harmony...
Sounds like endosymbiosis, which is how we ended up with mitochondria.
Apparently humans need the help of immunodepressive viruses during pregnancy, which I didn't know until I looked it up.
Thanks for prompting me - learn something new every day.
Would be really neat to have this in our lifetime.
Therin lies a problem... the human lifetime as compared to the rate IT technology is changing.
Whatever you got implanted today would be out of date within a year.
You either have to continually upgrade (which would probably involve more surgery, because the latest implants would not be backwards compatible with yours),
or you get stuck with the equivalent of a 800x600 display tattooed into your arm when everyone else has
1920x1200 HD and surround sound.
In the age of iPhone and Android, nothing says 'old fart' like having a clunky old phone that doesn't have a camera or net access,
but at least you can upgrade to a new model without involving surgery.
Yep - tried that. You also need to enable and start the network service too.
service network start
chkconfig network on
even then some parts of DHCP don't seem to work as reliably as they used to.
BUT - you can only do this *after* you have completed the install.
What is needed is a simple option on the install sequence that asks "is this a server".
If so, don't install all the bells and whistles that are designed for a laptop, just enable the standard network services.
As it is, you need to hand tweak each machine after the install. Not good for someone with a room full of machines.
I have yet to see a rational explanation of why anyone would want NetworkManager installed on a server.
I've been a RedHat and Fedora user for a similar length of time.
However I stopped updating at Fedora 8.
Fedora 9, 10 and 11 all use NetworkManager to configure the network interface, which breaks so many things that have been stable and working for a long time.
I can understand the benefits that NetworkManager brings for a laptop that connects and disconnects from lots of different networks, but can anyone explain what the benefits of NetworkManager are for a desktop or server that is wired into a fixed network ?
Something I would really like to see is a Fedora install package designed specifically for a desktop or server system, which leaves out many of the bells and whistles like NetworkManager or BlueTooth and restores the install option that said 'configure network from DHCP'.
In a word, NetworkManager.
But that is an over simplification.
I agree that a distros like Fedora with new versions of everything you expect bugs, and you have to accept
some problems in return for all the latest features.
However, the recent changes involving NetworkManager have broken many things 'just worked' for a standard server install.
What makes it a problem is that there isn't a simple way to work around all of the side effects, and some of the responses to the bug reports seem to treat problems with large scale server installations as a low priority.
NetworkManager is aimed primarily at laptops and mobile systems which regularly connect and disconnect from different networks - and when it all works, I look forward to installing it on my own laptop.
But for a server or desktop system that is wired to a fixed network, it causes a lot of extra problems and offers no real benefit.
But - as I said, you expect some problems when a new component is added.
What is missing is an option at the start of the install sequence that asks :
Is this a laptop or mobile system ? (which installs all the bells and whistles, including Networkmanager, BlueTooth etc.)
Is this a desktop or server system ? (which installs simple base system for you to extend)
It is worth noting that Ubuntu provide a separate download for a server install.
The whole thing fits on a single 700M CD, and runs happily in a VM with 64M of memory.
I think Fedora still works great as a viable preview of the future.
I agree, and it has also been a good indicator of what to expect in future versions of RedHat EL.
Have you tried installing Fedora 10 without NetworkManager or BlueTooth ?
On a laptop these make sense, but who in their right mind would want a room full of BlueTooth enabled servers ?
... people suggesting that a professional sysadmin cannot cope with the trivial differences between Linux distros...
Yep - point taken, which is why I am learning how to make my scripts distro neutral.
But, without a simple server install for Fedora, I think RedHat will be loosing out.
People like me install Fedora on their home desktop machine. I understand rpm/yum and all of the Red Hat specific configuration file locations.
Both cases are common.
The desktop distributions (Ubuntu/Fedora) are used as entry points for learning how the system tools work.
Developers and system admins become familiar with a particular distribution on their own desktop machine, and
use the corresponding server distribution (Ubuntu LTS/RedHat EL) when they come to install a new server.
Like you, I use Fedora on my desktop and RedHat EL on my servers because
that is what I'm familiar with.
However, I am beginning to think it might be time to change to Ubuntu.
The recent versions of Fedora (9, 10, 11) seem to be aimed at the home user and laptop market,
and are becoming less relevant for a development system.
I used to keep at least one machine updated with the latest Fedora release,
so that I could keep track of the new features and learn how to use them.
The last few Fedora releases (9, 10 and 11-alpha) have not been usable as servers, so all my machines have stayed at Fedora 8.
I am about the last one in our project team who does still use Fedora/RedHat, almost everyone else uses Ubuntu.
In which case they will prefer Ubuntu LTS rather than RedHat EL
when they come to install a live server, because that is what they are familiar with.
As a result I have started to migrate many of my configuration
scripts to use apt-get on Ubuntu rather than yum on Fedora.
So yes, there are still some of us who install Fedora on our desktop machines, and use
RedHat EL on our servers. But there may be fewer of us in the future.
The British government have not set out to control the populace, that will just be a purely unintentional side-effect.
They don't seem to realize what this many mean 5 or 10 years from now. The current government might not be planning to (mis)use these powers, but a future one might.
Another terrorist attack could get a fanatical nutter elected into government, and we are handing them a ready made police state. All the tools for complete control of the population installed and ready for (mis)use, all they would need to do is find an appropriate justification... and once you start (mis)using it, it is very hard to stop.
... you should keep in mind that reporter probably has no idea what the difference between downloading and uploading are, does not know the difference between theft and copyright infringement..
Pamela Jones, commonly known as PJ, is the creator and editor of Groklaw, an award-winning website that covers legal news of interest to the free and open-source software community. Jones is an Open Source advocate who previously trained and worked as a paralegal.
I think she probably does know the difference between theft and copyright infringement.
Neutronic functions make possible for the first time the ability to analyze regions of mathematics commonly thought to be undefined, such as the point where one is divided by zero.
.. At the middle of the concentrator another optic made of glass receives the incoming light, amplifies it and directs it toward a small solar cell.
Amplifying light with a glass 'optic' would be quite original.
Concentrating it yes, but amplifying it?
I would be very impressed if they have actually achieved it.
As an admin and dev, are you really in danger of downloading malware...
Under normal circumstances, no. On a bad day, when I'm concentrating on something else... perhaps.
As more and more vulnerabilities (and better ways of disguising malware) are found, it increases the chances that one day they will get lucky.
Most of the time I'm concentrating on solving my own problems and getting my own code to work, not on watching for potential malware attacks.
We have to be careful all of the time, the bad guys only need to get lucky once.
I must admit that if something did get access and used my account to do something like add a small cron job which ran when I wasn't at the machine, I doubt if I'd notice it immediately.
I agree with the sentiment in the original article, that we should not get complacent and never be taken in by the increasingly common but equally wrong idea that Linux is somehow invulnerable and 'bullet-proof'.
Loss of personal data (as in deletion) isn't the problem. Access to personal data (including identity) is the problem.
As someone else has already pointed out, first thing to look for would be the ~/.mozilla or ~/.thunderbird directories.
Which gives the malware all kinds of useful data, including address book, and saved passwords for websites.
If the malware was more technical and wanted to gain access to a server rather than just a desktop machine, then ~/.ssh would be a good place to start.
Given that the paper and metal tokens are just that - tokens that represent a promise to pay (something - now no longer certain) at a later date.
What exactly is 'real money' ?
By the time you see the warning, you have already used their computer system.
They could provide details of the alternative.
The warning pop-up could have a 'If you don't agree to these conditions, then follow this link for details on how to apply by post' option.
As it is, it basically says "We don't care if you visit or not. These are our terms, accept them or bugger off"
I will remember that next time I am thinking about attending a conference in the USA.
Slashdot Mirror
Um ... isn't that what RAID does ?
What you describe would just move the the the RAID controller inside the drive enclosure rather than on the PCI bus.
Unless you were thinking that we create a 10Tbyte disk from 10 x 1Tbyte discs, so if one fails you would only have to replicate 1Tbyte of data rather than the whole 10Tbyte. .... all in one little box that could overheat, driven by one power supply that could fail or spike damaging the LVM or RAID controller chip corrupting the whole lot.
In which case, the RAID controller would have to be able to 'see' inside the 10Tbyte virtual disc to know which of the internal discs had failed and what needed replicating.
So a 10Tbyte 'virtual' disc created by LVM gluing 10 x 1Tbyte RAID 1 arrays together to make them look like one large 10Tbyte disc ?
Also from the TV licensing site:
We don't own a TV. We do watch BBC programs from the iPlayer site, but only after they are broadcast (using standard Flash plugin NOT the DRM AdobeAir version). which means that technically we don't have to pay for a TV license. However we decided we would because we like what they produce and are happy to contribute something. IF the BBC start to add DRM to everything, we will probably reconsider.
If you see cloud compute as a 'software service' system like Google Mail or Google Docs, then perhaps yes.
We use cloud compute in the 'generic virtual machine provider' sense.
We use a couple of cloud compute providers to host our web servers. We pay a monthly fee and get the root password to a standard Linux virtual machine, what we install on it is up to us. If the physical hardware that is hosting our server fails, we just install it somewhere else. All of the install process is automated, and we can transfer our server and data from one provider to another in a couple of hours*.
We can alter the machine resources (cpu, memory disc) whenever we like.
We review the costs on a regular basis and can move to another provider if think we can get a better deal**.
* Installing from backup on to a new machine takes about 10min, the biggest delay is waiting for cached DNS records to catch up.
** We currently use two providers, one in the US and one in the UK and currency exchange rates can make a significant difference.
Why a USB connector ? That causes the same problem as making SSD cards use the SATA interface - the serial interface becomes slower than the things it is connected to.
What I would like to see is a set of sockets on the motherboard, mapped into the main memory address space (not PCI), a physical switch on the board to make them read only and software in the BIOS to make them look like a bootable disk.
Four sockets with 16 or 32G in each would give you enough space to store the entire OS. I don't know how Windows would handle it, but in a Unix or Linux based system it would be fairly easy to mount the devices as read only partitions and map them into the filesystem. This would be ideal for a server system, mapping the entire OS into the main memory address space and making it read only.
In fact all the BIOS would need to do is make the first 100M visible as a boot partition, and leave the OS to handle the rest.
First, let me say thank you for responding to this thread. It is useful to have a honest and informative debate about this.
I often start my day by reading the SlashDot front page. I pick a couple of summaries that look interesting and normally open both the SlashDot page and the article it links to in separate tabs.
However, if the SlashDot summary mentions multiple pages, login required, excessive advertising etc. I only open the SlashDot page and read the comments to see if it is worth reading TFA later. In this case Kdawson's warning meant I only opened the SlashDot page.
I got enough information from the comments to decide that it probably wasn't worth opening TFA right now.
I am interested in SSD devices, but the number of negative comments about the site readability put me off.
As I read more of the comments, your post as the site owner and the subsequent discussion actually became more interesting than the article itself. So, I did visit your site - not to find out more about SSD devices, but to see what all the fuss was about.
The immediate impression I got is that this site isn't aimed at me, it is aimed a kids putting together games machines. I've visited many similar sites and my normal reaction would be to just skip it and move on to somewhere else.
This time I stayed on the site and tried to read the article because I was curious to find out how intrusive the advertisements actually were.
The information you are presenting is detailed and technical but I found it very difficult to concentrate on the text because the flashing animated advertisements kept jumping around in my peripheral vision.
I looked at a few pages but it was too tiring/irritating to actually bother with reading all of the pages.
For me it isn't the number of advertisements, their position or content. The problem was the flashing animated advertisements vying for my attention meant I couldn't actually concentrate on the article I came to read.
The insistent animated advertisements actually worked against themselves. In trying to read article text I had to actively concentrate on excluding them - so I have no idea what they were actually trying to sell. My own experience is that if I do actually register what company or product the advertisement is for, I am left with a negative impression of them - I remember the irritation factor more than the benefits the advertisement was trying to portray.
Last but not least - I opened your site in a separate tab, and then flipped back to the SlashDot page to compare it with what people were saying. At that point I noticed that the animations were taking up so much cpu that scrolling the SlashDot page was slow and clunky. The only way to restore my web browser to a usable state was to close the tab with your site.
This is important, because when reading a story or article with links to other sites I normally open all of the associated pages in tabs and flip between them to compare and understand what they are saying.
The animated advertisements meant I visited your site once, read a couple of pages and then closed the tab.
Based on my experience, unless it was providing really important information that no one else provided, then I wouldn't visit it again.
My apologies, I hope you don't take this personally, it was meant as an unbiased report of user experience.
Not everyone is biased, your comment got moderated insightful didn't it ?
I agree BUT, based on their history, it may be wise to look at what the motives for this move may be.
For things like virtualization and Hypervisor:
Microsoft doesn't appear in the list until 2008, so they can hardly claim to be 'one of the most innovative companies' in this particular field.
IF this is the case, then I agree, this move should be welcomed.
Are you basing your statements on what the University studies said, or do you have personal experience of working with both open and closed source software yourself ?
As always your mileage may vary, and if this is what you have found then then this is good. This hasn't been my experience having working in IT developing both open source and closed source software. Many people find that closed source software does exactly what they need, which is fine by me. Best tool for the job and all that.
Having worked in (closed source) commercial software for a number of years, I know from experience that the same applies there too. Wherever you are, priorities are subject to a cost benefit analysis.
Just becuse we COULD fix a bug, doesn't mean someone WILL go through and fix the errors, until an important customer notices it.
It could also be argued that Microsoft were able to contribute because the Linux kernel is open source. How many people have the opportunity to see and contribute to the code for the Windows kernel ?
Indeed, and some of the best open source software is produced by people explicitly employed to do just that. By companies like IBM, Sun, Redhat, Oracle, Canonical ... and many many more.
I am currently working on a project funded by the UK government to produce open source software for the science community to use.
Um, not convinced on that .. I suspect that companies like IBM might be a better contender for that title.
Not only do they have a longer history and a wider range, they have also contributed some their patents to the community.
That is just plain scary.
Sounds like endosymbiosis, which is how we ended up with mitochondria.
Apparently humans need the help of immunodepressive viruses during pregnancy, which I didn't know until I looked it up. Thanks for prompting me - learn something new every day.
Therin lies a problem ... the human lifetime as compared to the rate IT technology is changing.
Whatever you got implanted today would be out of date within a year.
You either have to continually upgrade (which would probably involve more surgery, because the latest implants would not be backwards compatible with yours),
or you get stuck with the equivalent of a 800x600 display tattooed into your arm when everyone else has
1920x1200 HD and surround sound.
In the age of iPhone and Android, nothing says 'old fart' like having a clunky old phone that doesn't have a camera or net access,
but at least you can upgrade to a new model without involving surgery.
Yep - tried that. You also need to enable and start the network service too.
even then some parts of DHCP don't seem to work as reliably as they used to.
BUT - you can only do this *after* you have completed the install. What is needed is a simple option on the install sequence that asks "is this a server". If so, don't install all the bells and whistles that are designed for a laptop, just enable the standard network services. As it is, you need to hand tweak each machine after the install. Not good for someone with a room full of machines.
I have yet to see a rational explanation of why anyone would want NetworkManager installed on a server.
I've been a RedHat and Fedora user for a similar length of time. However I stopped updating at Fedora 8. Fedora 9, 10 and 11 all use NetworkManager to configure the network interface, which breaks so many things that have been stable and working for a long time.
I can understand the benefits that NetworkManager brings for a laptop that connects and disconnects from lots of different networks, but can anyone explain what the benefits of NetworkManager are for a desktop or server that is wired into a fixed network ?
Something I would really like to see is a Fedora install package designed specifically for a desktop or server system, which leaves out many of the bells and whistles like NetworkManager or BlueTooth and restores the install option that said 'configure network from DHCP'.
In a word, NetworkManager. But that is an over simplification.
I agree that a distros like Fedora with new versions of everything you expect bugs, and you have to accept some problems in return for all the latest features. However, the recent changes involving NetworkManager have broken many things 'just worked' for a standard server install. What makes it a problem is that there isn't a simple way to work around all of the side effects, and some of the responses to the bug reports seem to treat problems with large scale server installations as a low priority.
NetworkManager is aimed primarily at laptops and mobile systems which regularly connect and disconnect from different networks - and when it all works, I look forward to installing it on my own laptop. But for a server or desktop system that is wired to a fixed network, it causes a lot of extra problems and offers no real benefit.
But - as I said, you expect some problems when a new component is added. What is missing is an option at the start of the install sequence that asks :
It is worth noting that Ubuntu provide a separate download for a server install. The whole thing fits on a single 700M CD, and runs happily in a VM with 64M of memory.
I agree, and it has also been a good indicator of what to expect in future versions of RedHat EL.
Have you tried installing Fedora 10 without NetworkManager or BlueTooth ? On a laptop these make sense, but who in their right mind would want a room full of BlueTooth enabled servers ?
Yep - point taken, which is why I am learning how to make my scripts distro neutral. But, without a simple server install for Fedora, I think RedHat will be loosing out.
Both cases are common. The desktop distributions (Ubuntu/Fedora) are used as entry points for learning how the system tools work. Developers and system admins become familiar with a particular distribution on their own desktop machine, and use the corresponding server distribution (Ubuntu LTS/RedHat EL) when they come to install a new server.
Like you, I use Fedora on my desktop and RedHat EL on my servers because that is what I'm familiar with. However, I am beginning to think it might be time to change to Ubuntu. The recent versions of Fedora (9, 10, 11) seem to be aimed at the home user and laptop market, and are becoming less relevant for a development system. I used to keep at least one machine updated with the latest Fedora release, so that I could keep track of the new features and learn how to use them. The last few Fedora releases (9, 10 and 11-alpha) have not been usable as servers, so all my machines have stayed at Fedora 8.
I am about the last one in our project team who does still use Fedora/RedHat, almost everyone else uses Ubuntu. In which case they will prefer Ubuntu LTS rather than RedHat EL when they come to install a live server, because that is what they are familiar with. As a result I have started to migrate many of my configuration scripts to use apt-get on Ubuntu rather than yum on Fedora.
So yes, there are still some of us who install Fedora on our desktop machines, and use RedHat EL on our servers. But there may be fewer of us in the future.
They don't seem to realize what this many mean 5 or 10 years from now. The current government might not be planning to (mis)use these powers, but a future one might.
Another terrorist attack could get a fanatical nutter elected into government, and we are handing them a ready made police state. All the tools for complete control of the population installed and ready for (mis)use, all they would need to do is find an appropriate justification ... and once you start (mis)using it, it is very hard to stop.
In this instance I was quoting PJ from Groklaw.
I think she probably does know the difference between theft and copyright infringement.
Then what is this Sony v. Tenenbaum about then ?
They should have used a Neutronic function.
Amplifying light with a glass 'optic' would be quite original. Concentrating it yes, but amplifying it?
I would be very impressed if they have actually achieved it.
Under normal circumstances, no. On a bad day, when I'm concentrating on something else ... perhaps.
As more and more vulnerabilities (and better ways of disguising malware) are found, it increases the chances that one day they will get lucky. Most of the time I'm concentrating on solving my own problems and getting my own code to work, not on watching for potential malware attacks. We have to be careful all of the time, the bad guys only need to get lucky once. I must admit that if something did get access and used my account to do something like add a small cron job which ran when I wasn't at the machine, I doubt if I'd notice it immediately.
I agree with the sentiment in the original article, that we should not get complacent and never be taken in by the increasingly common but equally wrong idea that Linux is somehow invulnerable and 'bullet-proof'.
Loss of personal data (as in deletion) isn't the problem. Access to personal data (including identity) is the problem.
As someone else has already pointed out, first thing to look for would be the ~/.mozilla or ~/.thunderbird directories. Which gives the malware all kinds of useful data, including address book, and saved passwords for websites.
If the malware was more technical and wanted to gain access to a server rather than just a desktop machine, then ~/.ssh would be a good place to start.
I am both a developer and the sys admin for several machines, so how do I solve the problem on my desktop ?
Given that the paper and metal tokens are just that - tokens that represent a promise to pay (something - now no longer certain) at a later date.
What exactly is 'real money' ?
So pretty much the same as Google Analytics then.
No, it isn't finished. Because they still have several years worth of records, with details of every transaction you made during that time.
By the time you see the warning, you have already used their computer system.
They could provide details of the alternative. The warning pop-up could have a 'If you don't agree to these conditions, then follow this link for details on how to apply by post' option.
As it is, it basically says "We don't care if you visit or not. These are our terms, accept them or bugger off" I will remember that next time I am thinking about attending a conference in the USA.