Patches are just updated code, if there was no updates everyone could cry just as much and things would remain broken.
Technically everything is patched unless development stops, Linux has had hundreds or thousands of patches since it's 1.0 release, by the same logic it's past due to scrap Linux and start over too.
Really? For whom? Not for the average individual. Saying that a car has a single purpose is much like saying a computer has a single purpose.
For me. My car gets me from point A to point B and that's it, I know there's more going on behind the scenes than that but it's main purpose is to move me around.
Computers are used from everything from writing email to rendering entire movies to control car engines.
Why is there an issue with self-patching computers? For home users, this is a great idea.
I never had one, I think it's a good idea as long as it's able to be turned off by someone who wants to. The only problem I see is that MS' patches break shit periodically, having WU automatically install something that's going to break a 3rd party app is a very bad thing.
A computer has: chassis, CPU, motherboard, hard drive, RAM, video card, ethernet card, mouse, keyboard, and monitor. I'm not seeing consumer level complexity here.
I was referring to software complexity, not hardware. With computers it's software that's the problem, not the hardware (most of the time).
A computer has had decades to just fulfill those, and has failed. A computer rarely does what the user wants it to do, within the bounds of it's design.
That's because what user's want from computers changes every few years. 10 years ago noone knew about the Internet, now email is almost a necessity next to electricity and running water.
Cars don't "just die" -- there is always a sound reason for why it died. Windows (and sometimes Linux, and any other computers) just die. Period. No reason, they just crash.
Computers don't just die either, there's always a reason. Just because you can't find it doesn't mean it's not there.
Noone said you should. This is being released so that people can install it on test machines, put their software on it, see if/how it works and report bugs back so they can be fixed before the final release. I doubt RH would sell you a copy of the beta even if you asked them too.
Are you using ACLs on XFS or another filesystem? I believe the XFS ACL implementation does support a default ACL, atleast in 2.4. With all the integration in 2.6 I'm not sure if that option survived.
Cars do a lot less than computers, you can do just about anything with a computer while a car has a single specific use. If you had a computer designed explicitely to send email it would work fine like that, but because you want the computer to also play games, do your taxes and play DVDs you have a much more complicated beast.
Eventually computers will get to a point where 99% of things 'just work' but computers have had a lot less time to mature than cars have. Do you really believe people in the early/mid days of autmotive travel just jumped in cars and everything worked?
They aren't interested, and would rather just have someone who is interested fix their problems. There is nothing wrong with that, and it doesn't make them stupid.
That's perfectly fine, infact it's a good business for a lot of people. But the problem is that the people who aren't interested only call for help when something's broken, in the case of this worm a lot of people don't even know they have a problem. Paying someone periodically for general maintenance isn't something people want to do.
Maybe we need a computer equivalent of the service engine soon light aka idiot light.
They're software project names, not techo-babble. If you don't know what they are by their names you probably don't care enough to click through, but why waste time posting about it?
Because RPM works just as good as any package format, RH even ships up2date with the boxes and you can use up2date to install things and it resolves dependencies for you.
And I would never be so stupid as to dump RH for Gentoo on a work box. I use Debian on all my home machines but we use RH at work because our clients require software with support and packages that actually go through some form of QA process.
Install apt4rpm and setup a local package repository, no more dealing with dependencies unless you install something not in your repository and even then you could add that package and it's dependencies to your repository so that you only have to fight with it once.
not many people are going to use 2.6.0, and I'm sure V4 will make it into 2.6.1. The wait will be what, like a month?
The point is that unless reiser4 is 100% self contained and stable it shouldn't make it into 2.6 at all now because there's been a feature freeze and a new filesystem and anything else it adds to the kernel are features. Hans waited too long, again, and technically reiser4 shouldn't be included in the standard/Linus kernel until 2.7 now.
XFS wasn't allowed to be included in 2.4 officially because of those reasons but the SGI developers didn't cry about it, they just kept their patches up to date and waited for 2.5 to start so they could get included.
If the filesystem is actually stable and has real benefits over ext3, XFS, JFS, etc it'll probably be patched into distros like RedHat anyway so it's not a huge deal.
Thats because things are closest to how the individual developers intended them. No redhat/mandrake/debian branding, backports or modified packages.
So? The original author might not do their development and testing on Debian/RedHat/Mandrake so little tweaks are needed to make it run and integrate well. Distribution patches and branding are necessary to make the package feel like it's part of the distribution, otherwise Debian would just untar a bunch of precompiled binaries into/usr/local and let you sort it out.
Besides - apt is old. portage is a new and innovative system
Innovative? It's not like all of the BSDs have had working ports systems for years or anything.
APT works, I can't think of anything it doesn't do that I've needed it to do, why change it?
It might not be as stable as debian stable, but it is more stable than mandrake, redhat and debian unstable. (and more up to date as well)
I've been running unstable for a while now and nothing major has broken, infact I can't think of anything breaking since the last perl transition and even that wasn't that bad. apt-listbugs helps catch anything that might cause big problems, it's not rocket science.
Switching to RPM wouldn't be that big of a deal, except that the RPM database is binary and the dpkg one is plain-text and a lot of people really like plain-text.
Almost all of what makes Debian great can be just as easily done with RPM, apt and dselect (yes, I like dselect). The thing that seperates Debian from RedHat is the process not the package format, it would be just as easy to make RPMs that have great post-install scripts that make use of debconf to keep all the packages integrated.
But since few disk reads are large and contiguos it doesn't matter as much as people make it out to. The obvious examples where it does matter are audio and video editing, but then it's usually easier and faster to mkfs the tmp partion between each capture.
XFS *is* faster except for when you have a bunch of smaller files, true.
When I setup my IMAP server with Maildirs I tested a bunch of filesystems because I knew ext2 sucks for large directories. XFS and Reiserfs both had almost the exact same speeds for working with the large Maildirs.
Now I'm using XFS because I've had a better track record with it and ReiserFS offered no performance gains. The IMAP box is even running XFS on an Ultra2 and I havn't lost any data to it.
Slashdot Mirror
Patches are just updated code, if there was no updates everyone could cry just as much and things would remain broken.
Technically everything is patched unless development stops, Linux has had hundreds or thousands of patches since it's 1.0 release, by the same logic it's past due to scrap Linux and start over too.
For me. My car gets me from point A to point B and that's it, I know there's more going on behind the scenes than that but it's main purpose is to move me around.
Computers are used from everything from writing email to rendering entire movies to control car engines.
Why is there an issue with self-patching computers? For home users, this is a great idea.
I never had one, I think it's a good idea as long as it's able to be turned off by someone who wants to. The only problem I see is that MS' patches break shit periodically, having WU automatically install something that's going to break a 3rd party app is a very bad thing.
A computer has: chassis, CPU, motherboard, hard drive, RAM, video card, ethernet card, mouse, keyboard, and monitor. I'm not seeing consumer level complexity here.
I was referring to software complexity, not hardware. With computers it's software that's the problem, not the hardware (most of the time).
A computer has had decades to just fulfill those, and has failed. A computer rarely does what the user wants it to do, within the bounds of it's design.
That's because what user's want from computers changes every few years. 10 years ago noone knew about the Internet, now email is almost a necessity next to electricity and running water.
Cars don't "just die" -- there is always a sound reason for why it died. Windows (and sometimes Linux, and any other computers) just die. Period. No reason, they just crash.
Computers don't just die either, there's always a reason. Just because you can't find it doesn't mean it's not there.
Have you tried compiling a kernel with Intel's compiler?
A lot of things, most notably the Linux kernel, use gcc specific extensions.
Noone said you should. This is being released so that people can install it on test machines, put their software on it, see if/how it works and report bugs back so they can be fixed before the final release. I doubt RH would sell you a copy of the beta even if you asked them too.
Are you using ACLs on XFS or another filesystem? I believe the XFS ACL implementation does support a default ACL, atleast in 2.4. With all the integration in 2.6 I'm not sure if that option survived.
Eventually computers will get to a point where 99% of things 'just work' but computers have had a lot less time to mature than cars have. Do you really believe people in the early/mid days of autmotive travel just jumped in cars and everything worked?
They aren't interested, and would rather just have someone who is interested fix their problems. There is nothing wrong with that, and it doesn't make them stupid.
That's perfectly fine, infact it's a good business for a lot of people. But the problem is that the people who aren't interested only call for help when something's broken, in the case of this worm a lot of people don't even know they have a problem. Paying someone periodically for general maintenance isn't something people want to do.
Maybe we need a computer equivalent of the service engine soon light aka idiot light.
Too bad the target audience of this worm doesn't have an AD to help them.
If you know what Gnome and KDE are you should be interested enough to click the link instead of bitching about the use of a product's name in a story.
They're software project names, not techo-babble. If you don't know what they are by their names you probably don't care enough to click through, but why waste time posting about it?
Because RPM works just as good as any package format, RH even ships up2date with the boxes and you can use up2date to install things and it resolves dependencies for you.
And I would never be so stupid as to dump RH for Gentoo on a work box. I use Debian on all my home machines but we use RH at work because our clients require software with support and packages that actually go through some form of QA process.
Install apt4rpm and setup a local package repository, no more dealing with dependencies unless you install something not in your repository and even then you could add that package and it's dependencies to your repository so that you only have to fight with it once.
There are several console apt front ends that let you browse, search, etc available packages. dselect, aptitude, synaptic, etc.
Well looking at the patch it makes changes to a number of files outside of fs/reiser4:
b d/transaction.cf s.h. hm ap.c
fs/buffer.c
fs/filesystems.c
fs/inode.c
fs/j
fs/sysfs/inode.c
include/linux/
include/linux/init_task.h
include/linux/jbd
include/linux/sched.h
kernel/ksyms.c
mm/file
The point is that unless reiser4 is 100% self contained and stable it shouldn't make it into 2.6 at all now because there's been a feature freeze and a new filesystem and anything else it adds to the kernel are features. Hans waited too long, again, and technically reiser4 shouldn't be included in the standard/Linus kernel until 2.7 now.
XFS wasn't allowed to be included in 2.4 officially because of those reasons but the SGI developers didn't cry about it, they just kept their patches up to date and waited for 2.5 to start so they could get included.
If the filesystem is actually stable and has real benefits over ext3, XFS, JFS, etc it'll probably be patched into distros like RedHat anyway so it's not a huge deal.
Pretty much, but as you see he got his way last time so what's to say it won't work again?
Interestingly enough one of our offices has bathrooms with a lock on the door that requires a 3 digit code to get in.
Thats because things are closest to how the individual developers intended them. No redhat/mandrake/debian branding, backports or modified packages.
/usr/local and let you sort it out.
So? The original author might not do their development and testing on Debian/RedHat/Mandrake so little tweaks are needed to make it run and integrate well. Distribution patches and branding are necessary to make the package feel like it's part of the distribution, otherwise Debian would just untar a bunch of precompiled binaries into
Besides - apt is old. portage is a new and innovative system
Innovative? It's not like all of the BSDs have had working ports systems for years or anything.
APT works, I can't think of anything it doesn't do that I've needed it to do, why change it?
It might not be as stable as debian stable, but it is more stable than mandrake, redhat and debian unstable. (and more up to date as well)
I've been running unstable for a while now and nothing major has broken, infact I can't think of anything breaking since the last perl transition and even that wasn't that bad. apt-listbugs helps catch anything that might cause big problems, it's not rocket science.
Switching to RPM wouldn't be that big of a deal, except that the RPM database is binary and the dpkg one is plain-text and a lot of people really like plain-text.
Almost all of what makes Debian great can be just as easily done with RPM, apt and dselect (yes, I like dselect). The thing that seperates Debian from RedHat is the process not the package format, it would be just as easy to make RPMs that have great post-install scripts that make use of debconf to keep all the packages integrated.
Exactly.
The cost of the IFS development kit for making filesystem drivers for Win2K3 (the Win2K one is already deprecated it seems) is $899.
But since few disk reads are large and contiguos it doesn't matter as much as people make it out to. The obvious examples where it does matter are audio and video editing, but then it's usually easier and faster to mkfs the tmp partion between each capture.
You need thousands of files for it to start making a difference though.
When I setup my IMAP server with Maildirs I tested a bunch of filesystems because I knew ext2 sucks for large directories. XFS and Reiserfs both had almost the exact same speeds for working with the large Maildirs.
Now I'm using XFS because I've had a better track record with it and ReiserFS offered no performance gains. The IMAP box is even running XFS on an Ultra2 and I havn't lost any data to it.
Perhaps you over-tuned it? I've been using XFS for a a few years on my desktop and never had any hangs I could blame on XFS and the speed is fine.