The problem is that ext filesystems are designed to run on small IDE drives, which nobody has anymore. As filesystems get bigger and bigger, XFS performs better than the others. google for benchmarks!
Well, if you have a particular point to make, please include the link. The last thing anybody is going to do is to Google for generic benchmarks without even knowing what kind of workloads we are talking about, etc. Plus, that is not very relevant to my comment about Ext3.
The article makes vague implicit claims which are misleading and inaccurate. I have nothing against ReiserFS, however this article is mostly smoke.
It all but claims that ext3 is unreliable and slow. Ext3 is certainly not the fastest in absolute terms, but it is not slow. As for reliability, it is clearly the best tested and most reliable. It should also be noted that Ext3 supports higher levels of journaling than other journaling FS-es like XFS and JFS. So, Ext3 is fast, very reliable, excellently supported and under active development (as ext4) - no need to despair.
The article also implies that ext4 is an immature and incompatible replacement, which is also not true. In fact Ext4 is probably considered more stable than ReiserFS 4 by some. If it wasn't for the decision to change the name from ext3 to ext4, many people probably wouldn't know when they started using "ext4" after an upgrade. There are more details in this LWN article: http://lwn.net/Articles/187336/
Such arrogance is charming, but not very informative. Having developed system software for both Windows and Linux for many years, and having read a couple of books in the interim, thank you, I have learned (perhaps the hard way) that in practice the Unix security model is vastly superior to the Windows one. Perhaps surprisingly the Unix model actually has theoretical advantages too.
Do not underestimate the intrinsic value of simplicity for security. ACLs, for example, are actually very harmful unless you _really_ need them; in fact they are not needed nearly as often as Windows-only users think.
You can't literally as a LocalSystem, but you can operate a shell as it, which is more than close enough. IIRC, one of the ways to do it was to schedule a command prompt using the "at" service. Don't tell that to the fanboys though:-)
LOL. I will have whatever you are drinking, because it must be pretty strong. I would like to see the connection at your destination that will allow you to restore gigabytes of data "without a dent" in your productivity. You must not travel a lot. I would also very much like to sign up with your insurance company, though I suspect they may be filing for bankruptcy soon...:-)
In the real world, best case scenario is you are out a couple of thousand bucks buying a new laptop, then waste at least several days restoring your system and data, and keep hoping to may be able to recover some (half?) of the money later on from your insurance company.
Depending on what you've got on your HDD and how strong your principles and especially you wallet are, it may or may not be worth it.
Mod me down if you care, bit I really wish that PC Perspective had edited the transcript just a little (or more). This is just unacceptable. Carmack is speaking with half sentences, etc, which is OK if you are listening to him, but is really absurdly stupid if you have to read it in print.
Standards of written text have really gone down on the web.
You make good points. In my opinion the improved security fully justifies using Vista for a new home PC. I am trying to be objective - I use Linux myself professionally, but I am very glad to have Vista instead of XP on my wife's laptop.
However if you consider a 3GHz CPU with 1 GB RAM to be "an old box", then you have some serious perception problems...:-) An "old box" would be an Athlon 2000 with 512MB PC133 RAM and PATA66. XP runs just fine, thank you.
At the same time I have Vista Home Premium (dual booting with Debian) on a relatively powerful quad-core PC with 3GB RAM, 512MB NVIDIA 8XXX card, SATA, etc (the works), and while it is not slow, it is not snappy either ! I expect most things to be instantaneous on such hardware and they aren't. Sometimes I get the the waiting cursor even for trivial tasks like opening the control panel, with no other apps running ! (well, except Steam, the anti-virus and the other craplets that come with a pre-installed PC:-) That is a disgrace.
I know it is popular to bash Comcast these days, but I am a Comcast subscriber and I regularly get about 1 MB/s download speed, when downloading actual stuff, not benchmarks. One recent example was the Sun Java SDK. This is under Linux BTW, although I don't think it matters. So, I am pretty sure that they are not cheating, at least in my case.
Harrison Ford is Han Solo. There can never be a Han Solo played by a different actor. Or Indiana Jones for that matter. (For the same reason the "young Indiana Jones" movies had nothing to do with the "real" Indiana Jones.)
I've never understood how a programmer/direction manager/geek like Torvalds could raise so much interest over his opinion
It is simple really. Unlike 99% of the population and 99.9999999% of the people who waste their time on Slashdot (me included), Linus Torvalds has actually created something:-) Actions do speak louder than words and also help underscore them. It is very reasonable and practical to pay attention to the opinions of somebody who has proven himself as able to make good choices and deliver results, even if it is only in a narrow technical area.
There are a just few people who are globally known as exceptionally good software engineers - John Carmack would be another one, for example. It always pays to hear their opinions carefully.
I am in the same boat. I intend to keep my analog cable for as long as I can. I actually used to have Comcast Digital Cable several years ago and it was pretty bad: the quality of the digital channels was _visibly_ worse, there was a delay when flipping through channels, and I had to find place for yet another box and remote. Oh yes, it was two times more expensive too:-)
Hypothetically, I would consider switching only if I already had a HDTV set (say, if my current analog set broke down and HDTV sets were less expensive) and the digital subscription was the same price as analog.
You are completely right - you might as well have been reading my mind. The good news is, there are many of us who agree with you (even if we not very visible on public forums; probably because we have a lot of work to do:-). I think it is safe to disregard the negative replies - they are probably from people who haven't yet had the joy of distributing and supporting commercial (even open source!) software under Linux. Or even haven't had the joy of using commercial software on a distribution which isn't "officially supported" (understand: isn't Red Hat:-)
A related problem, for which I have been flamed countless times, is the idiotic notion that one has to upgrade his entire OS in order to obtain newer versions of software. This is simply and utterly unacceptable, unless you are a geek (luckily, I am one). There must be a standard, secure and reliable way to install 3rd party packages. The way _must not_ involve compiling the source:-)
The solution that is being pushed - making more frequent OS releases (see Ubuntu) - is simply absurd. People absolutely do not want to upgrade their OS every six months. This is complete idiocy. People however do want to be able to run the new versions of OpenOffice, Firefox, or whatever.
Sadly that is not true. Most people naively assume that wear leveling magically fixes everything, while in reality the situation is different.
First of all, due to the nature of NAND flash, you can only erase whole 128KB blocks and then write in 2KB chunks. You can simply not rewrite a 2KB page, unless you erase the entire 128KB block. So, it is not simply a matter of wear leveling, but of fragmentation and how to avoid it. This is much more serious. A naive algorithm means that you will be erasing 64 times more often, which means that the life drops from 10,000 rewrites/block to 160 rewrites/block. I am not aware of any ideal solution, unless battery-backed RAM is used for buffering.
Secondly, most (all?) flash controllers, do not implement "global" wear leveling - that is across the entire storage area. They separate the storage in zones and do it only inside each zone - this decreases the effectiveness.
This is not to say that there are no solutions. They may be expensive though. In any case, thinking that "wear leveling" is magic bullet is simply wrong.
We tried switching of Open Office in our company and we've had an endless stream of problems. Don't get me wrong - OpenOffice on its own is good enough and in some things better than MS Office. However before it is 100% compatible with MSOffice, it cannot be used as a replacement. And when I say 100%, I mean it.
Some of our problems included: - Inability to edit files created with MSOffice. The formatting would be lost, or some parts (e.g. pictures) were not editable at all. - Exporting files was similarly problematic (alas, 100% of our partners and customers require to receive MS Office documents). Sometimes files would look different in MSOffice. - MSOffice has better help system - Everybody already is familiar with MSOffice.
It turned out that everybody needed to have a copy of MSOffice anyway, to check if the files converted cleanly, etc. It was all pointless.
HA HA HA HA HA. Yeah, right - some garage shop with revenue of 10's of thousands. I suspect that much of the US population does not realize how badly broken the US judical system is. Any IP-related lawsuit in the US costs hundreds of thousands of dollars just to get the ball rolling. To win, you'd need to spend a couple of millions at the very least. If you are not a company with at least tens of millions yearly revenue (this is optimistic), you absolutely cannot afford to get involvbed in a lawsuit _at all_. I am not even mentioning countersuits, or damages in case you lose. These expenses are in case you win !! (And you are lucky if you can recover part of them eventually, because surprisingly often you cannot).
Of course it is much better if you are on the other side and have the money. It gives you a legal way to bankrupt any small business - just file an arbitrary lawsuit against them and be prepared to spend a million or two. Of course it would probably be easier to just buy them.
In the US money equals justice (at least as far as IP is concerned). If SCO had sued a smaller company instead of IBM, who'd you think would have won ? (Unless of course EFF and the likes came to the rescue).
By the way,.MSI is a (poorly) reversed engineered version of the InstallShield Installer.
For the record, this is completely false. I agree with your other points, though. Only the last sentence is misinformation which detracts from the valid points you are making.
This seems like some kind of a scam. What can one think after reading this:
9.) Will OS5 be based on any code from OS3 or OS4?
BM>> I would prefer not answering this at this time. As we are ready to release more information it will become a much clearer picture.
10.) Will OS5 have a full-blown Java VM?
BM>> Wait and see. We have some pretty interesting plans with regard to Java.
11.) Will OS5 be a 64-bit OS, i.e. run on 64-bit CPUs, with a
64-bit address space? What about memory protection?
BM>> I would prefer not answering this at this time. As we are ready to release more information it will become a much clearer picture.
13.) In the Amiga.Org answers, you said that OS5 didn't use Tao
technology, yet you also stated that OS5 would run on multiple CPU
architectures. How do you support different CPU families without the Tao technology? Fat binaries?
BM>> You support them in the method and way that we are going to support them.
It is obvious that either this guy has no idea at all of what is going on, or that he is lying and there is no development at all, the latter being much more likely. I read the other interview linked from the article and it was full of the same nonsense - definitely not anything that I'd expect from a serious business let alone its CEO. It is completely ridiculous.
Although I respect what Amiga was in the past (although I never personally used it), my advice to the Amiga fans and hobbyists is to forget about this "company". Amiga is dead.
Well, that is something. Personally, it always annoys me that the "Preferences" in Firefox are under "Tools" in Windows and under "Edit" in Linux. Perhaps it is a part of the aforementioned patches and I agree that it is a pointless change.
On a different subject - are there issues with Debian using the SeaMonkey name ?
Debian has HUGE patches that are of questionable quality
According to who ? I have seen this claim repeated several times, in most cases almost certainly by people who either have never looked at the patches, or are not qualified to. I am not saying that you are either of these, but since you repeated the claim, are you ? Why are the patches questionable ?
I will stick with a distro that is more in touch with reality, thanks.
May be not everyone's reality is the same as yours:-) If you want to have fun upgrading all you servers and installations with the "latest stable software" (according to you) every six months, it is your prerogative, but I certainly wouldn't let anyone with that attitude anywhere near my servers:-)
That said, I agree with some of your points. I wish Debian would periodically add new versions of packages to the stable distribution - not to replace the old ones! For example when OpenOffice 2 (or Firefox 1.5) came out, they should have added them to stable without removing the old packages.
Slashdot Mirror
No, I don't have it backwards. See for example: http://en.wikipedia.org/wiki/Ext3#Advantages .
Well, if you have a particular point to make, please include the link. The last thing anybody is going to do is to Google for generic benchmarks without even knowing what kind of workloads we are talking about, etc. Plus, that is not very relevant to my comment about Ext3.
The article makes vague implicit claims which are misleading and inaccurate. I have nothing against ReiserFS, however this article is mostly smoke.
It all but claims that ext3 is unreliable and slow. Ext3 is certainly not the fastest in absolute terms, but it is not slow. As for reliability, it is clearly the best tested and most reliable. It should also be noted that Ext3 supports higher levels of journaling than other journaling FS-es like XFS and JFS. So, Ext3 is fast, very reliable, excellently supported and under active development (as ext4) - no need to despair.
The article also implies that ext4 is an immature and incompatible replacement, which is also not true. In fact Ext4 is probably considered more stable than ReiserFS 4 by some. If it wasn't for the decision to change the name from ext3 to ext4, many people probably wouldn't know when they started using "ext4" after an upgrade.
There are more details in this LWN article: http://lwn.net/Articles/187336/
Such arrogance is charming, but not very informative. Having developed system software for both Windows and Linux for many years, and having read a couple of books in the interim, thank you, I have learned (perhaps the hard way) that in practice the Unix security model is vastly superior to the Windows one. Perhaps surprisingly the Unix model actually has theoretical advantages too.
Do not underestimate the intrinsic value of simplicity for security. ACLs, for example, are actually very harmful unless you _really_ need them; in fact they are not needed nearly as often as Windows-only users think.
You can't literally as a LocalSystem, but you can operate a shell as it, which is more than close enough. IIRC, one of the ways to do it was to schedule a command prompt using the "at" service. Don't tell that to the fanboys though :-)
LOL. I will have whatever you are drinking, because it must be pretty strong. I would like to see the connection at your destination that will allow you to restore gigabytes of data "without a dent" in your productivity. You must not travel a lot. I would also very much like to sign up with your insurance company, though I suspect they may be filing for bankruptcy soon ... :-)
In the real world, best case scenario is you are out a couple of thousand bucks buying a new laptop, then waste at least several days restoring your system and data, and keep hoping to may be able to recover some (half?) of the money later on from your insurance company.
Depending on what you've got on your HDD and how strong your principles and especially you wallet are, it may or may not be worth it.
Mod me down if you care, bit I really wish that PC Perspective had edited the transcript just a little (or more). This is just unacceptable. Carmack is speaking with half sentences, etc, which is OK if you are listening to him, but is really absurdly stupid if you have to read it in print.
Standards of written text have really gone down on the web.
You make good points. In my opinion the improved security fully justifies using Vista for a new home PC. I am trying to be objective - I use Linux myself professionally, but I am very glad to have Vista instead of XP on my wife's laptop.
... :-) An "old box" would be an Athlon 2000 with 512MB PC133 RAM and PATA66. XP runs just fine, thank you.
:-) That is a disgrace.
However if you consider a 3GHz CPU with 1 GB RAM to be "an old box", then you have some serious perception problems
At the same time I have Vista Home Premium (dual booting with Debian) on a relatively powerful quad-core PC with 3GB RAM, 512MB NVIDIA 8XXX card, SATA, etc (the works), and while it is not slow, it is not snappy either ! I expect most things to be instantaneous on such hardware and they aren't. Sometimes I get the the waiting cursor even for trivial tasks like opening the control panel, with no other apps running ! (well, except Steam, the anti-virus and the other craplets that come with a pre-installed PC
I know it is popular to bash Comcast these days, but I am a Comcast subscriber and I regularly get about 1 MB/s download speed, when downloading actual stuff, not benchmarks. One recent example was the Sun Java SDK. This is under Linux BTW, although I don't think it matters. So, I am pretty sure that they are not cheating, at least in my case.
Man, thank you !!! I haven't laughed that hard in a long time :-)
Harrison Ford is Han Solo. There can never be a Han Solo played by a different actor. Or Indiana Jones for that matter. (For the same reason the "young Indiana Jones" movies had nothing to do with the "real" Indiana Jones.)
Not in this case. Android cannot be ported to the Neo 1973 because sadly it is not open source (yet).
Good points!
It is simple really. Unlike 99% of the population and 99.9999999% of the people who waste their time on Slashdot (me included), Linus Torvalds has actually created something :-) Actions do speak louder than words and also help underscore them. It is very reasonable and practical to pay attention to the opinions of somebody who has proven himself as able to make good choices and deliver results, even if it is only in a narrow technical area.
There are a just few people who are globally known as exceptionally good software engineers - John Carmack would be another one, for example. It always pays to hear their opinions carefully.
I am in the same boat. I intend to keep my analog cable for as long as I can. I actually used to have Comcast Digital Cable several years ago and it was pretty bad: the quality of the digital channels was _visibly_ worse, there was a delay when flipping through channels, and I had to find place for yet another box and remote. Oh yes, it was two times more expensive too :-)
Hypothetically, I would consider switching only if I already had a HDTV set (say, if my current analog set broke down and HDTV sets were less expensive) and the digital subscription was the same price as analog.
You are completely right - you might as well have been reading my mind. The good news is, there are many of us who agree with you (even if we not very visible on public forums; probably because we have a lot of work to do :-). I think it is safe to disregard the negative replies - they are probably from people who haven't yet had the joy of distributing and supporting commercial (even open source!) software under Linux. Or even haven't had the joy of using commercial software on a distribution which isn't "officially supported" (understand: isn't Red Hat :-)
:-)
:-)
A related problem, for which I have been flamed countless times, is the idiotic notion that one has to upgrade his entire OS in order to obtain newer versions of software. This is simply and utterly unacceptable, unless you are a geek (luckily, I am one). There must be a standard, secure and reliable way to install 3rd party packages. The way _must not_ involve compiling the source
The solution that is being pushed - making more frequent OS releases (see Ubuntu) - is simply absurd. People absolutely do not want to upgrade their OS every six months. This is complete idiocy. People however do want to be able to run the new versions of OpenOffice, Firefox, or whatever.
Good bye, positive karma
Sadly that is not true. Most people naively assume that wear leveling magically fixes everything, while in reality the situation is different.
First of all, due to the nature of NAND flash, you can only erase whole 128KB blocks and then write in 2KB chunks. You can simply not rewrite a 2KB page, unless you erase the entire 128KB block. So, it is not simply a matter of wear leveling, but of fragmentation and how to avoid it. This is much more serious. A naive algorithm means that you will be erasing 64 times more often, which means that the life drops from 10,000 rewrites/block to 160 rewrites/block. I am not aware of any ideal solution, unless battery-backed RAM is used for buffering.
Secondly, most (all?) flash controllers, do not implement "global" wear leveling - that is across the entire storage area. They separate the storage in zones and do it only inside each zone - this decreases the effectiveness.
This is not to say that there are no solutions. They may be expensive though. In any case, thinking that "wear leveling" is magic bullet is simply wrong.
We tried switching of Open Office in our company and we've had an endless stream of problems. Don't get me wrong - OpenOffice on its own is good enough and in some things better than MS Office. However before it is 100% compatible with MSOffice, it cannot be used as a replacement. And when I say 100%, I mean it.
Some of our problems included:
- Inability to edit files created with MSOffice. The formatting would be lost, or some parts (e.g. pictures) were not editable at all.
- Exporting files was similarly problematic (alas, 100% of our partners and customers require to receive MS Office documents). Sometimes files would look different in MSOffice.
- MSOffice has better help system
- Everybody already is familiar with MSOffice.
It turned out that everybody needed to have a copy of MSOffice anyway, to check if the files converted cleanly, etc. It was all pointless.
Great post(s) ! Seriously, you are reading my mind. Your post might as well have been written by me (although I am not as eloquent).
Same problem here with Firefox and JDK 1.5. Shrug.
HA HA HA HA HA. Yeah, right - some garage shop with revenue of 10's of thousands. I suspect that much of the US population does not realize how badly broken the US judical system is. Any IP-related lawsuit in the US costs hundreds of thousands of dollars just to get the ball rolling. To win, you'd need to spend a couple of millions at the very least. If you are not a company with at least tens of millions yearly revenue (this is optimistic), you absolutely cannot afford to get involvbed in a lawsuit _at all_. I am not even mentioning countersuits, or damages in case you lose. These expenses are in case you win !! (And you are lucky if you can recover part of them eventually, because surprisingly often you cannot).
Of course it is much better if you are on the other side and have the money. It gives you a legal way to bankrupt any small business - just file an arbitrary lawsuit against them and be prepared to spend a million or two. Of course it would probably be easier to just buy them.
In the US money equals justice (at least as far as IP is concerned). If SCO had sued a smaller company instead of IBM, who'd you think would have won ? (Unless of course EFF and the likes came to the rescue).
For the record, this is completely false. I agree with your other points, though. Only the last sentence is misinformation which detracts from the valid points you are making.
This seems like some kind of a scam. What can one think after reading this:
It is obvious that either this guy has no idea at all of what is going on, or that he is lying and there is no development at all, the latter being much more likely. I read the other interview linked from the article and it was full of the same nonsense - definitely not anything that I'd expect from a serious business let alone its CEO. It is completely ridiculous.
Although I respect what Amiga was in the past (although I never personally used it), my advice to the Amiga fans and hobbyists is to forget about this "company". Amiga is dead.
Well, that is something. Personally, it always annoys me that the "Preferences" in Firefox are under "Tools" in Windows and under "Edit" in Linux. Perhaps it is a part of the aforementioned patches and I agree that it is a pointless change.
On a different subject - are there issues with Debian using the SeaMonkey name ?
Debian has HUGE patches that are of questionable quality
According to who ? I have seen this claim repeated several times, in most cases almost certainly by people who either have never looked at the patches, or are not qualified to. I am not saying that you are either of these, but since you repeated the claim, are you ? Why are the patches questionable ?
I will stick with a distro that is more in touch with reality, thanks.
May be not everyone's reality is the same as yours :-) If you want to have fun upgrading all you servers and installations with the "latest stable software" (according to you) every six months, it is your prerogative, but I certainly wouldn't let anyone with that attitude anywhere near my servers :-)
That said, I agree with some of your points. I wish Debian would periodically add new versions of packages to the stable distribution - not to replace the old ones! For example when OpenOffice 2 (or Firefox 1.5) came out, they should have added them to stable without removing the old packages.