This is why you should always make your logical raid volumes 1% smaller (approximately)
There is abut 7% difference between 2^30 and 10^9. I have seen disks being exactly 80*10^9 bytes, I believe they were sold as 80GB disks. If you find an 80GB disk which is really 80GB, you will have to leave 7% unused, that is 5.5GB waste.
If a linux sysadmin finds a command line a bit too much, then you've probably got bigger problems on your hands
You missed the point. Installing updates from the command line is really no problem to me, I prefer to do it that way. But I need to know when updates are available, for that purpose I cannot imagine anything much better than the RHN applet.
The script I use to call 'rpm -Fvh' also uses 'rpm -K' first and only installs if that reports no errors. What bothers me is, that it is still possible by mistake to install a package with no signature. Better not make mistakes while logged in as root.
I use the applet to know when updates are available. The actual download and install I do from the command line. Basically it is just three steps:
wget to get the updates from their http server
rpm -K to verify signatures
rpm -Fvh to install
so I for one is able to use it just fine without having to pay anything. Besides I think you can actually use the trial version for an arbitrary time if you just fill in their survey once in a while. The amount of service you can get from Red Hat without paying is actually quite impressive.
A command line interface is not exactly what was requested. RHN-applet is a good example of a litle icon on the screen clearly indicating when updates are available. I have it on my screen all the time. You can also use it to install the updates, though personally I preffer to use the command line for that.
You can use a serial console or try out some version of the netconsole patch to get the messages on another computer. (Notice that netconsole over the internet is probably possible, but it is sent in clear and can be snooped or modified). I also recall reading about some patch to keep a new kernel ready in memory that could be booted with arguments telling it where to find the log from the old kernel, I even think it included a checksum to prevent booting the new kernel if it had been corrupted.
... when this hobby OS gets a real file-locking features. Instead of that "advisory locks" FUD that is poured onto us.
Linux does offer mandatory locking. It is not enabled by default, because it really does have major drawbacks. If you really want mandatory locking, you can choose on which filesystems and files to use it. The problem with mandatory locking is, that it is PITA and can in some cases even cause DoS attacks. Advisory locks is all you need. The purpose of locks is to prevent corruption in case of simultaneous updates, and inconsistent reads while data is being updated. As long as the programs reading and writing a file use locking and behaves correctly, it will work eventhough the locks are not mandatory. Mandatory locks does not prevent broken programs from corrupting the data, so what does it give you? First of all mandatory locks prevents you from reading the file while it is being updated, even if you really want to. Linux (and UNIX) does not prevent you from doing stupid things, because in doing so, it would also prevent you from doing clever things. Sure you can probably mess up some files by reading/writing while there is a mandatory lock. But you shouldn't be prevented, because often you know what you are doing, and really want to read the file. One of the most annoying things about Windows (the versions I have seen) is, that you cannot start reading a file before it has been completely written. I often want to start reading a file while I'm still not finished downloading it.
While I will agree that there is a possibility of patches taking a while to appear from closed-source software (and that it has happened, usually regarding Internet Explorer)
In other news there are 30 security holes in IE which are well known and documented, which Microsoft has not yet done anything about.
Since NVIDIA is the only high-end graphics company with quality Linux drivers, they're pretty much the only option in the market.
The quality is being questioned. NVIDIA drivers have been blamed for some kernel crashes. The lack of options is the major reasons we don't want to see those drivers go. But that is not necesarilly what counts for NVIDIA. How large a fraction of the NVIDIA cards sold are being used with Linux? That is important to NVIDIA.
Does anyone know how their drivers for Linux are licensed?
The usual closed source stuff is the best I would hope for. Surely I would prefer a GPL driver, or at least just specifications released, so somebody else could write it. But I was afraid they might have gone the other way, and just release a closed source driver for Windows, and no specs. Luckily that seems not to be the case (yet). Still your question is important, much more important than so much other stuff, but unfortunately I don't know the answer. No matter how great the hardware is, it is not worth much without a proper driver. So the answer to your question really should have been in the resume.
The mentioned code was 'main()', not just 'main('. When you start looking for substrings, you can obviously find them everywhere. It is not like I have copyright on the letter I though it appears in a message I wrote. Look for the correct sequence like I did, and you will see most occurences in documentation and comments within drivers I don't use anyway.
Luckily no important part of Linux uses that construct. It is mentioned a few times in the documentation and comments, but we can remove that without breaking anything. (Hint: Linux is a kernel, not a program.)
If you are to look on a CRT screen for a long time, you certainly want a high refresh rate. How much is required to be enough probably depends on who you are, but 75Hz is not enough for me. But I can hardly tell the difference between 85Hz and 100Hz. I think 100Hz is enough for most people.
When you have chosen a refresh rate, the optimal FPS is exactly the same number. Generating more FPS is waste because it is just gives worse quality. You would either be skiping frames, which harms animations. Or you would be showing parts of different frames at the same time, which gives visible horisontal lines, where the two parts doesn't match. And yes, you will spot those broken images even when only shown for 1/100th of a second.
But generating 100 FPS and showing 100 FPS is not enough, you have to ensure each frame is showed exactly once. It requires a litle help from the graphics hardware, but nothing that is hard to implement. Having a litle extra processing power is important, you must be able to produce ever frame fast enough. You don't want to miss a deadline because occationally one frame takes just a litle more CPU time to render.
That encryption is involved makes it the kind of rights-restricting scheme the DMCA makes illegal to re-engineer.
But this feature is not preventing violation of copyright. It is "protecting" documents not owned by Microsoft possibly against the copyright holders wish. Besides reverse engineering for interoperability is explicitly permitted by law in many countries.
Every/. and k5 page where i have a post should then come up /. has a lot of large generated html pages, and following links you can find every single comment viewed in a lot of different ways. Google simply doesn't crawl all of it. And it seems google did not implement features for inteligent crawling of slashdot.
Google should fix that but i assume it takes quite a bit of googlepower to implement
The link: feature is broken, and I think they should fix it, or at least stop
advertising a feature they cannot make work.
The problem with windows was never it's gui. (Well, not for most users at least.)
At least the GUI is not the major problem with Windows. The Windows GUI is not bad, but there are a few things I dislike about the Windows GUI. For example the virtual desktops available in most Linux GUIs should have been standard in Windows by now. In an earlier
comment I told about some of the reasons I like the average Linux GUI more than Windows.
The simple answer is that a normal CDRom is just not built for advanced testing.
I agree with that. I don't expect functionality equivalent to specialized hardware. I just want to do as much as possible with standard hardware. A standard CD-ROM drive will read some raw data from media and then do error-correction. The error-correction is probably done with a combination of specialized hardware and some firmware code. I just want the drive to skip this step and send me the raw data from the disc in whatever format it is. An identical error-correction step could then be performed in software, though it would probably be slower. But you get the chance to know how many bits was actually changed by the error-correction, and you can apply more advanced algorithms to do the error-correction if the normal one fails.
These guys used CD analyzer hardware (CDA-3000) to check the quality of the discs.
Why is it that standard CD-ROM hardware will not allow us to do this? Providing the raw data from the media including ECC would allow some amount of quality analysis in software, and it would also help in recovering parially damaged discs. And BTW keep in mind, that a CDROM is more reliable than a CDDA, because on CDDA only 25% of the disc is used for ECC while on a CDROM 35% is used for ECC.
Slashdot Mirror
This is why you should always make your logical raid volumes 1% smaller (approximately)
There is abut 7% difference between 2^30 and 10^9. I have seen disks being exactly 80*10^9 bytes, I believe they were sold as 80GB disks. If you find an 80GB disk which is really 80GB, you will have to leave 7% unused, that is 5.5GB waste.
If a linux sysadmin finds a command line a bit too much, then you've probably got bigger problems on your hands
You missed the point. Installing updates from the command line is really no problem to me, I prefer to do it that way. But I need to know when updates are available, for that purpose I cannot imagine anything much better than the RHN applet.
Actually that is more or less what I do. There are a few differences:
- Rather than checking all
.rpm files each time, I check only those I am about to install.
- The 2>&1 part is not necesarry.
- I match on a longer part of the string.
The script I use to call 'rpm -Fvh' also uses 'rpm -K' first and only installs if that reports no errors. What bothers me is, that it is still possible by mistake to install a package with no signature. Better not make mistakes while logged in as root.I don't know about 8.x series but RH9 also has a windows-like popup that appears on the gnome desktop that says if there are updates.
It has been shown by default since RH7.3. BTW 8.0 is the only release in the "8.x series".
- wget to get the updates from their http server
- rpm -K to verify signatures
- rpm -Fvh to install
so I for one is able to use it just fine without having to pay anything. Besides I think you can actually use the trial version for an arbitrary time if you just fill in their survey once in a while. The amount of service you can get from Red Hat without paying is actually quite impressive.apt-get upgrade
A command line interface is not exactly what was requested. RHN-applet is a good example of a litle icon on the screen clearly indicating when updates are available. I have it on my screen all the time. You can also use it to install the updates, though personally I preffer to use the command line for that.
You can use a serial console or try out some version of the netconsole patch to get the messages on another computer. (Notice that netconsole over the internet is probably possible, but it is sent in clear and can be snooped or modified). I also recall reading about some patch to keep a new kernel ready in memory that could be booted with arguments telling it where to find the log from the old kernel, I even think it included a checksum to prevent booting the new kernel if it had been corrupted.
Teach me how to lock a file from a root-level account on Linux.
. html
http://tldp.org/HOWTO/Secure-Programs-HOWTO/files
... when this hobby OS gets a real file-locking features. Instead of that "advisory locks" FUD that is poured onto us.
Linux does offer mandatory locking. It is not enabled by default, because it really does have major drawbacks. If you really want mandatory locking, you can choose on which filesystems and files to use it. The problem with mandatory locking is, that it is PITA and can in some cases even cause DoS attacks. Advisory locks is all you need. The purpose of locks is to prevent corruption in case of simultaneous updates, and inconsistent reads while data is being updated. As long as the programs reading and writing a file use locking and behaves correctly, it will work eventhough the locks are not mandatory. Mandatory locks does not prevent broken programs from corrupting the data, so what does it give you? First of all mandatory locks prevents you from reading the file while it is being updated, even if you really want to. Linux (and UNIX) does not prevent you from doing stupid things, because in doing so, it would also prevent you from doing clever things. Sure you can probably mess up some files by reading/writing while there is a mandatory lock. But you shouldn't be prevented, because often you know what you are doing, and really want to read the file. One of the most annoying things about Windows (the versions I have seen) is, that you cannot start reading a file before it has been completely written. I often want to start reading a file while I'm still not finished downloading it.
In random configurations, yes, there are issues with NVIDIA's stability. In closed configurations, there is not.
That's strange, it is the first time I hear about random configurations. Most people I know use pretty deterministic configurations.
While I will agree that there is a possibility of patches taking a while to appear from closed-source software (and that it has happened, usually regarding Internet Explorer)
In other news there are 30 security holes in IE which are well known and documented, which Microsoft has not yet done anything about.
Check your EULA. If you want more than 5 simultaneous connections to your NT/2K web server, you are NOT ALLOWED to use anything but IIS.
For how long time does IE keep a connection open after it was last used?
Since NVIDIA is the only high-end graphics company with quality Linux drivers, they're pretty much the only option in the market.
The quality is being questioned. NVIDIA drivers have been blamed for some kernel crashes. The lack of options is the major reasons we don't want to see those drivers go. But that is not necesarilly what counts for NVIDIA. How large a fraction of the NVIDIA cards sold are being used with Linux? That is important to NVIDIA.
Does anyone know how their drivers for Linux are licensed?
The usual closed source stuff is the best I would hope for. Surely I would prefer a GPL driver, or at least just specifications released, so somebody else could write it. But I was afraid they might have gone the other way, and just release a closed source driver for Windows, and no specs. Luckily that seems not to be the case (yet). Still your question is important, much more important than so much other stuff, but unfortunately I don't know the answer. No matter how great the hardware is, it is not worth much without a proper driver. So the answer to your question really should have been in the resume.
When will enough be, finally, enough?
Never. (Trivial answer, but true)
$ find linux-2.6.0-test5 -name '*.c' | xargs grep '^int main('
The mentioned code was 'main()', not just 'main('. When you start looking for substrings, you can obviously find them everywhere. It is not like I have copyright on the letter I though it appears in a message I wrote. Look for the correct sequence like I did, and you will see most occurences in documentation and comments within drivers I don't use anyway.
main() as it will infringe on SCO copyrights
Luckily no important part of Linux uses that construct. It is mentioned a few times in the documentation and comments, but we can remove that without breaking anything. (Hint: Linux is a kernel, not a program.)
Otherwise he is just a moron.
Al Gore or Darl McBride?
If you think KDE is *slow* wait 'til you run it over morse code....
How does that compare to KDE over RFC1149?
If you are to look on a CRT screen for a long time, you certainly want a high refresh rate. How much is required to be enough probably depends on who you are, but 75Hz is not enough for me. But I can hardly tell the difference between 85Hz and 100Hz. I think 100Hz is enough for most people.
When you have chosen a refresh rate, the optimal FPS is exactly the same number. Generating more FPS is waste because it is just gives worse quality. You would either be skiping frames, which harms animations. Or you would be showing parts of different frames at the same time, which gives visible horisontal lines, where the two parts doesn't match. And yes, you will spot those broken images even when only shown for 1/100th of a second.
But generating 100 FPS and showing 100 FPS is not enough, you have to ensure each frame is showed exactly once. It requires a litle help from the graphics hardware, but nothing that is hard to implement. Having a litle extra processing power is important, you must be able to produce ever frame fast enough. You don't want to miss a deadline because occationally one frame takes just a litle more CPU time to render.
That encryption is involved makes it the kind of rights-restricting scheme the DMCA makes illegal to re-engineer.
But this feature is not preventing violation of copyright. It is "protecting" documents not owned by Microsoft possibly against the copyright holders wish. Besides reverse engineering for interoperability is explicitly permitted by law in many countries.
Every /. and k5 page where i have a post should then come up
/. has a lot of large generated html pages, and following links you can find every single comment viewed in a lot of different ways. Google simply doesn't crawl all of it. And it seems google did not implement features for inteligent crawling of slashdot.
Google should fix that but i assume it takes quite a bit of googlepower to implement
The link: feature is broken, and I think they should fix it, or at least stop advertising a feature they cannot make work.
The problem with windows was never it's gui. (Well, not for most users at least.)
At least the GUI is not the major problem with Windows. The Windows GUI is not bad, but there are a few things I dislike about the Windows GUI. For example the virtual desktops available in most Linux GUIs should have been standard in Windows by now. In an earlier comment I told about some of the reasons I like the average Linux GUI more than Windows.
The simple answer is that a normal CDRom is just not built for advanced testing.
I agree with that. I don't expect functionality equivalent to specialized hardware. I just want to do as much as possible with standard hardware. A standard CD-ROM drive will read some raw data from media and then do error-correction. The error-correction is probably done with a combination of specialized hardware and some firmware code. I just want the drive to skip this step and send me the raw data from the disc in whatever format it is. An identical error-correction step could then be performed in software, though it would probably be slower. But you get the chance to know how many bits was actually changed by the error-correction, and you can apply more advanced algorithms to do the error-correction if the normal one fails.
These guys used CD analyzer hardware (CDA-3000) to check the quality of the discs.
Why is it that standard CD-ROM hardware will not allow us to do this? Providing the raw data from the media including ECC would allow some amount of quality analysis in software, and it would also help in recovering parially damaged discs. And BTW keep in mind, that a CDROM is more reliable than a CDDA, because on CDDA only 25% of the disc is used for ECC while on a CDROM 35% is used for ECC.