Older versions of PHP still get patched for security holes, i`m sure you can run a patched version of 5.1... I run a patched version of 4.x, since some of my older apps don't work with 5
What i find the biggest problem of all... On the debian (or any other unix) servers, you only need to patch what you have installed, and you only have installed what you need, so the numbers of patches are relatively small and can usually be applied without rebooting. On windows however, not only must you patch the apps you actually use, but you need to patch things like ie and outlook express and directx etc, things that are designed for end users workstations and have absolutely no place being installed on a webserver. Add to that the fact that many of these patches require a reboot, and you have significant wasted time. Also, debian patches only fix the vulnerability, windows patches often introduce new features which must be tested before being deployed (or do the staff in the linked article blindly install patches without testing them first?)
If slashdot were done the way you describe, how would you understand the context when reading this reply? Would this post make any sense whatsoever without seeing it's context? Email is the same, because: Often in business people get CC:'d in right in the middle of a conversation (So you need to read the context) In mailing lists, it's often easiest to go to the last post and read the whole thing in one block than trying to find every post and read each one. And if you revisit an old email (many people keep their mails for years) it's much easier to read the whole thing in one block than read it backwards or find every related mail. And as for wading through crap, in mail you know the reply will be at the bottom, just go straight down, or configure your mailer to start after the indented text (your mailer can do that, right?)
Email replies should always start at the bottom, unless your writing in a language that reads from the bottom upwards (do any exist? theres a few left to right languages but i`m not sure about bottom to top)... Trying to follow a conversation where the reply comes BEFORE the original question is a horrendous thing to do, this is known as top posting... And consider the following short example:
A: Because it messes up the order in which people normally read text. Q: Why is top-posting such a bad thing? A: Top-posting. Q: What is the most annoying thing on usenet and in e-mail?
Ever noticed how on slashdot, the story is at the top, and the comments descend from there, with child posts being below their parents etc... Now imagine the whole thing upside down.
People who are responsible for maintaining computer systems _SHOULD_ know how they work, would you use a mechanic who knew nothing about how your car worked?
Why can't they provide 3d? At the very least, in the form of closed binary X11/OpenGL drivers like nvidia do... Then you could run ports of games like quake easily.
There is a difference between fair compensation for work, and ridiculous levels of profit. Nothing wrong with getting fair compensation for doing some work, and continuing to receive compensation for continuing to work... But it's very wrong for someone to continue getting ridiculous amounts of money for some work they did many years ago and have simply sat on their backside ever since.
IBM are selling their chips to non-console markets too. The console market provides the economy of scale, they can produce millions of chips for the console markets and sell them for a minimal profit, and then produce thousands of chips for more lucrative markets (the chips still cost the same to produce) and sell them at a huge markup. Plus, by producing chips for console makers, these aforementioned console makers have paid for a huge chunk of the research and upfront tooling costs which will hugely benefit IBM when they go to produce chips for other markets.
That's a ridiculously stupid rule... And how is it even enforced? Open Source is everywhere... Loads of embedded devices use open source, and a lot of commercial products implement it in places (windows - bsd tcp stack etc).
They're selling xeon and opteron based systems now too... And their UltraSparc-T1 is doing well, and it's fairly unique out there right now. Besides, Sun make most of their money from support contracts, and being able to provide a complete software stack is good from that perspective, letting other people/companies do some of the development work is even better for Sun.
But there's nothing to stop you from selling your rear seats to a third party, whereas trying to sell OEM bundled software will cause you a lot of hassle. You can also buy cars which have no rear seats to start with.
Not to mention that: Solaris 8 dropped support for sun4c (and sun4 i guess, not sure when that was dropped) Solaris 1 (SunOS 5.0) dropped support for m68k sun3 machines etc... Hardware moves on, there's not much point continuing support for modern software running on old hardware that can't be purchased anymore... Those machines which are still around, will still run the old versions of solaris they've been running for years. And when it comes to machines still supported by opensolaris, you have the ultra-5 and ultra-2, which are dirt cheap and readily available on ebay for an absolute pittance, and will easily outperform any old sun4m boxes.
NeoOffice is still much slower than the X11 version of OpenOffice, i have both installed on my mac (old dual G4/450) and there is a quite noticeable difference.
Only if Sun also wanted to sell it... Perhaps Sun's price or conditions were too high, that microsoft considered it more viable to make their own replacement instead. There's also the old case of "Not Invented Here" syndrome.
Installing any OS can be hairy, unless the OS and hardware are tied closely together. That's why the vast majority of computer users never install their OS. Receiving Gentoo preinstalled on a computer would work out quite well for end users, since once installed its easy to maintain and install new apps... It's also possible to do rolling updates of it, instead of having to completely reinstall a new version every few years.
> And why aren't these instructions on the wiki? because giving someone instructions to put into a console is much more concise, simple, and much more difficult to screw up than guiding them through a GUI.
EXACTLY! When are people going to understand this, they're still afraid of the command line but don't realise that from a support perspective, a command line is much easier!
A guide on a website - you can cut+paste the commands directly into your command line, couldn't be simpler, if you encounter problems you can cut+paste the output back into google.
Over the phone - telling someone exactly what to type, and having them read the output is much easier than trying to explain to them where to click (any kind of settings could affect where the icons/buttons are located or what they look like) and then trying to understand their responses, since they have to interpret the graphical output and you'l get their interpretation of it, rather than a direct reading of textual output. And again i re-iterate, the graphical output could be completely different depending on configuration (fonts, resolution, skins etc), whereas reading text is pretty universal regardless of how big the text is, where it's located, what font it's using or what colour it is.
Here's the thing, support media are inherently word based (verbal communication is simply another method of representing words) and words are most easily represented on a computer as text.
I've seen lots of windows users fall apart reading graphical guides because: The fonts are different, they changed the fonts to one they feel looks better The resolution is different, things look smaller or are in a different place They're using a corporate XP system, where the fisher-price interface has been turned off and yet the graphical guide has it turned on Items in the start menu are in a different place, because they have different apps installed etc
The issue with nvidia drivers is one caused by nvidia having proprietary drivers that ubuntu can't distribute with the distro... The problem doesn't occur if your using any videocard that has open drivers, like the very common intel chipsets. If you don't like it, complain to nvidia! On an intel chipset for example, linux is far more likely to have drivers for it out of the box than current versions of windows, where you'l need to not only install such drivers, but often locate them from the manufacturers website *AND* identify which of the drivers you actually need. If you find that windows doesnt have a driver for your network card by default, you could be in for a world of trouble... I've had no alternative but to boot linux livecd's on peoples systems in order to download windows network drivers to a usb stick.
What, in the same way that all the applications running on linux have to pay linus a license fee too? By writing a java program you are not creating a derivative work of java itself, you are using java for it's intended purpose. It's like creating a piece of C code that is compiled using GCC.
Viruses dont need to know the public address, viruses typically want to: Connect OUTBOUND to other systems in order to infect them. Connect OUTBOUND to a server (eg irc) from which they can be controlled. Connect OUTBOUND to send spam. Nat will not stop any of this... Modern malware doesn't open a listening port for incoming connections, this has already been rendered useless by the fact most isp's supply dynamic ip addresses.
As for local firewall security spotting malware, if the malware runs with superuser privileges then no local firewall will do any good since the malware can simply tamper with it and it's logs. And as to sending an email, you still have to send the email somewhere, and the outbound address will appear in the headers regardless.
And trojans, most trojans establish outbound connections to a server from which they can be controlled, a listening port is far more obvious than an outbound connection because it can be detected remotely, whereas an outbound connection can be hidden from the user by modifying the OS to hide it.
In the broader scheme of things, many p2p protocols are unusable, or don't work properly, when using nat, legitimate protocols that users will actually want to use, and the alternative is either opening holes in the nat (which means multiple users cant use the same port behind the same nat gateway) or using outbound connections to a central server, which increases latency and causes significant bandwidth usage on that central server.
Viruses nowadays don't need to know their real ip's, or accept incoming connections... They just need to make outbound connections, so they can connect to the server from which theyre being controlled, and make outbound connections to infect other machines. NAT doesnt help here. As for a good firewall built in to every OS, this is true, no OS should ever be without some kind of built in packet filtering ability, but so long as regular users have superuser privileges on these systems, a host based firewall becomes worthless. Any virus which executes as a privileged user has the ability to disable whatever firewall is in place, and if every user is running the same firewall it becomes much easier since the virus now only has one program it needs to neuter.
Very few modern viruses need to determine their real IP's, nor do they need the ability to accept incoming connections. Maybe in the days of back orifice this would have been a problem, but today most malware makes an outbound connection to somewhere from which it can be controlled.
It's probably defaulting to IPv6 and therefore trying to do AAAA (ipv6 address) DNS lookups first (which fail) and then doing normal A (ipv4 address) lookups afterwards. Depending on the dns servers in use, and how they respond to AAAA requests (return an error quickly, ignore the request so it has to time out etc) this could make browsing a LOT slower.
Often you find that your new GPU requires a relatively fast processor to feed it data quick enough... Not to mention newer versions of AGP/PCIe etc... I bought an Athlon64 3200 a while back, and pretty soon thereafter upgraded my video card too, since the old one had poor to non-existent 64bit drivers. And when i bought that videocard originally, for a K6-2/400, it provided virtually no performance improvement over the previous card i had because the system couldn't keep up with it.
What happens when/if steam is shut down? Valve won't run it forever, at which point you'll no longer be able to download games from it, and may no longer be able to run the ones you already have. I prefer console games, that due to the lack of ability to patch them, have to be written properly in the first place, also since the hardware is always the same I too can take them anywhere only needing the appropriate console, instead of an x86 compatible computer which meets a huge array of other requirements.
Yes, i found that too... Emerge is far more transparent as to it's process for installing packages, and all the scripts are plain text files and easily read/modified, as is the package database.
Slashdot Mirror
Older versions of PHP still get patched for security holes, i`m sure you can run a patched version of 5.1...
I run a patched version of 4.x, since some of my older apps don't work with 5
What i find the biggest problem of all...
On the debian (or any other unix) servers, you only need to patch what you have installed, and you only have installed what you need, so the numbers of patches are relatively small and can usually be applied without rebooting.
On windows however, not only must you patch the apps you actually use, but you need to patch things like ie and outlook express and directx etc, things that are designed for end users workstations and have absolutely no place being installed on a webserver. Add to that the fact that many of these patches require a reboot, and you have significant wasted time.
Also, debian patches only fix the vulnerability, windows patches often introduce new features which must be tested before being deployed (or do the staff in the linked article blindly install patches without testing them first?)
If slashdot were done the way you describe, how would you understand the context when reading this reply?
Would this post make any sense whatsoever without seeing it's context?
Email is the same, because:
Often in business people get CC:'d in right in the middle of a conversation (So you need to read the context)
In mailing lists, it's often easiest to go to the last post and read the whole thing in one block than trying to find every post and read each one.
And if you revisit an old email (many people keep their mails for years) it's much easier to read the whole thing in one block than read it backwards or find every related mail.
And as for wading through crap, in mail you know the reply will be at the bottom, just go straight down, or configure your mailer to start after the indented text (your mailer can do that, right?)
Email replies should always start at the bottom, unless your writing in a language that reads from the bottom upwards (do any exist? theres a few left to right languages but i`m not sure about bottom to top)...
Trying to follow a conversation where the reply comes BEFORE the original question is a horrendous thing to do, this is known as top posting... And consider the following short example:
A: Because it messes up the order in which people normally read text.
Q: Why is top-posting such a bad thing?
A: Top-posting.
Q: What is the most annoying thing on usenet and in e-mail?
Ever noticed how on slashdot, the story is at the top, and the comments descend from there, with child posts being below their parents etc... Now imagine the whole thing upside down.
People who are responsible for maintaining computer systems _SHOULD_ know how they work, would you use a mechanic who knew nothing about how your car worked?
Why can't they provide 3d? At the very least, in the form of closed binary X11/OpenGL drivers like nvidia do...
Then you could run ports of games like quake easily.
There is a difference between fair compensation for work, and ridiculous levels of profit.
Nothing wrong with getting fair compensation for doing some work, and continuing to receive compensation for continuing to work... But it's very wrong for someone to continue getting ridiculous amounts of money for some work they did many years ago and have simply sat on their backside ever since.
IBM are selling their chips to non-console markets too.
The console market provides the economy of scale, they can produce millions of chips for the console markets and sell them for a minimal profit, and then produce thousands of chips for more lucrative markets (the chips still cost the same to produce) and sell them at a huge markup.
Plus, by producing chips for console makers, these aforementioned console makers have paid for a huge chunk of the research and upfront tooling costs which will hugely benefit IBM when they go to produce chips for other markets.
That's a ridiculously stupid rule...
And how is it even enforced? Open Source is everywhere... Loads of embedded devices use open source, and a lot of commercial products implement it in places (windows - bsd tcp stack etc).
They're selling xeon and opteron based systems now too...
And their UltraSparc-T1 is doing well, and it's fairly unique out there right now.
Besides, Sun make most of their money from support contracts, and being able to provide a complete software stack is good from that perspective, letting other people/companies do some of the development work is even better for Sun.
But there's nothing to stop you from selling your rear seats to a third party, whereas trying to sell OEM bundled software will cause you a lot of hassle.
You can also buy cars which have no rear seats to start with.
Not to mention that:
Solaris 8 dropped support for sun4c (and sun4 i guess, not sure when that was dropped)
Solaris 1 (SunOS 5.0) dropped support for m68k sun3 machines
etc...
Hardware moves on, there's not much point continuing support for modern software running on old hardware that can't be purchased anymore... Those machines which are still around, will still run the old versions of solaris they've been running for years.
And when it comes to machines still supported by opensolaris, you have the ultra-5 and ultra-2, which are dirt cheap and readily available on ebay for an absolute pittance, and will easily outperform any old sun4m boxes.
NeoOffice is still much slower than the X11 version of OpenOffice, i have both installed on my mac (old dual G4/450) and there is a quite noticeable difference.
Only if Sun also wanted to sell it...
Perhaps Sun's price or conditions were too high, that microsoft considered it more viable to make their own replacement instead.
There's also the old case of "Not Invented Here" syndrome.
Installing any OS can be hairy, unless the OS and hardware are tied closely together. That's why the vast majority of computer users never install their OS.
Receiving Gentoo preinstalled on a computer would work out quite well for end users, since once installed its easy to maintain and install new apps... It's also possible to do rolling updates of it, instead of having to completely reinstall a new version every few years.
> And why aren't these instructions on the wiki? because giving someone instructions to put into a console is much more concise, simple, and much more difficult to screw up than guiding them through a GUI.
EXACTLY!
When are people going to understand this, they're still afraid of the command line but don't realise that from a support perspective, a command line is much easier!
A guide on a website - you can cut+paste the commands directly into your command line, couldn't be simpler, if you encounter problems you can cut+paste the output back into google.
Over the phone - telling someone exactly what to type, and having them read the output is much easier than trying to explain to them where to click (any kind of settings could affect where the icons/buttons are located or what they look like) and then trying to understand their responses, since they have to interpret the graphical output and you'l get their interpretation of it, rather than a direct reading of textual output. And again i re-iterate, the graphical output could be completely different depending on configuration (fonts, resolution, skins etc), whereas reading text is pretty universal regardless of how big the text is, where it's located, what font it's using or what colour it is.
Here's the thing, support media are inherently word based (verbal communication is simply another method of representing words) and words are most easily represented on a computer as text.
I've seen lots of windows users fall apart reading graphical guides because:
The fonts are different, they changed the fonts to one they feel looks better
The resolution is different, things look smaller or are in a different place
They're using a corporate XP system, where the fisher-price interface has been turned off and yet the graphical guide has it turned on
Items in the start menu are in a different place, because they have different apps installed etc
The issue with nvidia drivers is one caused by nvidia having proprietary drivers that ubuntu can't distribute with the distro...
The problem doesn't occur if your using any videocard that has open drivers, like the very common intel chipsets. If you don't like it, complain to nvidia!
On an intel chipset for example, linux is far more likely to have drivers for it out of the box than current versions of windows, where you'l need to not only install such drivers, but often locate them from the manufacturers website *AND* identify which of the drivers you actually need.
If you find that windows doesnt have a driver for your network card by default, you could be in for a world of trouble... I've had no alternative but to boot linux livecd's on peoples systems in order to download windows network drivers to a usb stick.
What, in the same way that all the applications running on linux have to pay linus a license fee too?
By writing a java program you are not creating a derivative work of java itself, you are using java for it's intended purpose. It's like creating a piece of C code that is compiled using GCC.
Viruses dont need to know the public address, viruses typically want to:
Connect OUTBOUND to other systems in order to infect them.
Connect OUTBOUND to a server (eg irc) from which they can be controlled.
Connect OUTBOUND to send spam.
Nat will not stop any of this... Modern malware doesn't open a listening port for incoming connections, this has already been rendered useless by the fact most isp's supply dynamic ip addresses.
As for local firewall security spotting malware, if the malware runs with superuser privileges then no local firewall will do any good since the malware can simply tamper with it and it's logs. And as to sending an email, you still have to send the email somewhere, and the outbound address will appear in the headers regardless.
And trojans, most trojans establish outbound connections to a server from which they can be controlled, a listening port is far more obvious than an outbound connection because it can be detected remotely, whereas an outbound connection can be hidden from the user by modifying the OS to hide it.
In the broader scheme of things, many p2p protocols are unusable, or don't work properly, when using nat, legitimate protocols that users will actually want to use, and the alternative is either opening holes in the nat (which means multiple users cant use the same port behind the same nat gateway) or using outbound connections to a central server, which increases latency and causes significant bandwidth usage on that central server.
Viruses nowadays don't need to know their real ip's, or accept incoming connections...
They just need to make outbound connections, so they can connect to the server from which theyre being controlled, and make outbound connections to infect other machines. NAT doesnt help here.
As for a good firewall built in to every OS, this is true, no OS should ever be without some kind of built in packet filtering ability, but so long as regular users have superuser privileges on these systems, a host based firewall becomes worthless. Any virus which executes as a privileged user has the ability to disable whatever firewall is in place, and if every user is running the same firewall it becomes much easier since the virus now only has one program it needs to neuter.
Very few modern viruses need to determine their real IP's, nor do they need the ability to accept incoming connections.
Maybe in the days of back orifice this would have been a problem, but today most malware makes an outbound connection to somewhere from which it can be controlled.
It's probably defaulting to IPv6 and therefore trying to do AAAA (ipv6 address) DNS lookups first (which fail) and then doing normal A (ipv4 address) lookups afterwards. Depending on the dns servers in use, and how they respond to AAAA requests (return an error quickly, ignore the request so it has to time out etc) this could make browsing a LOT slower.
Often you find that your new GPU requires a relatively fast processor to feed it data quick enough...
Not to mention newer versions of AGP/PCIe etc...
I bought an Athlon64 3200 a while back, and pretty soon thereafter upgraded my video card too, since the old one had poor to non-existent 64bit drivers. And when i bought that videocard originally, for a K6-2/400, it provided virtually no performance improvement over the previous card i had because the system couldn't keep up with it.
What happens when/if steam is shut down? Valve won't run it forever, at which point you'll no longer be able to download games from it, and may no longer be able to run the ones you already have.
I prefer console games, that due to the lack of ability to patch them, have to be written properly in the first place, also since the hardware is always the same I too can take them anywhere only needing the appropriate console, instead of an x86 compatible computer which meets a huge array of other requirements.
Yes, i found that too... Emerge is far more transparent as to it's process for installing packages, and all the scripts are plain text files and easily read/modified, as is the package database.