You clearly have an axe to grind which is fair enough (maybe you've lost valuble data to ext3 disk corruption?). However corruption can be down to more than the filesystem. If you don't actually know what caused the corruption in the first place it's hard to start doing comparisons - e.g. I've seen ext3 corruption several times over the past few years but it has always been down to disks just about to die completely or unknown corner case motherboard bugs.
I've been told by a kernel dev that reiserfs rarely causes corruption on perfect hardware but is far less tolerant of flakey hardware than ext3. What if those new checks are to try and protect people on failling hardware? Continuing to give right results when you're being given wrong data is hard...
What if the kernel has had a dodgy binary driver stuffed into it and it is corrupting bits of kernel memory thus leading to on disk corruption? The kernel filesystem can try and stop a complete and utter blowout but it can only do so many checks...
The initial state of the system is also important. The only times I've known Windows 2000 to have disk corruption there's always been a partition resizer involved at some point.
Finally 100% bug free software is hard to create and even harder to prove. What works for me could quite well fail for you. Neither Ext3, NTFS nor Reiserfs are magic and sometimes things go wrong. Recovery is almost always possible if you are prepared to forgo access to the data and pay a fiscal price.
(Just for the record didn't MS ship NTFS before Ext3 was released into the major distros? Technically MS got to the widespread atomic filesystem first in comparison to Linux...).
Taking a look at the SDL website news shows that version 1.2.8 of the main library was released on December 15th (under a month ago at the time of writing). There also appears to be active development on supporting libraries (such as SDL_mixer, SDL_ttf and SDL_image).
Sure Loki were a major supporter but development didn't stop after Loki disappeared...
I was going to give you a link to where the bug database was but it seems to have turned into a promotional page - Coverity Linux "bug database" page. Normally that link goes to a bugzilla like database showing you a list of bugs, what type of error was found and a highlighted code snippet. All very nifty.
However this hits upon something I don't like about this press release. I wonder whether the headline should be "Tool finds less bugs in Linux than rival OSes when previous bugs found by the tool are fixed in Linux". Assuming fixes for all problems that means only new and changed code gives a chance for these errors to occur (which is a subset of the entire code).
Were all codebases that were compared given coverity bug databases and a chance to fix the reported bugs too and over what time scale? If the Linux kernel code gets 6 monthly reports and say OpenBSD kernel only gets reports one every two years and you test towards the end of those two years is that fair test? What does it say other than our tool finds less bugs in code that is regularly checked by previous versions of our tool? Is that in itself a good justification?
Is this a fixed bug? I just tested the above in a firefox 1.0PR1 in a basic HTML file and nothing nearly as bad as not being able to type happened. In fact, seemingly nothing happened (other than the image never appeared to finish loading).
This is a bit offtopic but what the hey. You ask serveral questions in this post so I my answer may not answer all your points:
To the best of my knowledge Firefox will only offer to migrate IE bookmarks. It doesn't do cookies although if you are willing to mess about by hand I think you can get IE to export it's cookies to a file and then copy the cookies file to the firefox profile.
Will the migration be painless? Who knows? It's been generally painless for me on recent machines but that's no guarantee for you.
It is possible using certain languages to formally verify and basically prove that for a certain program a particular set of properties does or does not hold true. I am fairly certain it is not possible to do this with a non trivial program in C (unless you try and annotate it and/or restrict which operations you are allowed to do) though, so this point is a bit moot.
If you are a MandrakeClub member then there has been access to a well packaged version of firefox throughout the 10.0 (the good news is that there is a recent version of firefox in Mandrake 10.1 contribs now). Feodra also had a well packaged version for it over at Fedora Extras. The moral is that if you break away from your distro specific packaging, expect problems...
ACPI works well on my desktop (as does software suspend, only one problem so far) but there are reports of laptops suffering beacuse of a lack of nolapic support. It's also worth noting that many machines come with poorly programmed BIOSes and need updates or extra code to have working ACPI. Why not try a LiveCD distro to test with until you find out you have support? It may also help if you pick a big vendor and report the issue to them so that they can track it and have you test fixes...
I suspect you won't get round to reading this since it was posted days after your original post but well, you never know...
I installed RealPlayer 10 Gold about a month ago on my home Linux box. Real Player on Linux has never seemed as scary as the Windows version and I was still using v8 to listen to BBC streams. Let me digress and just thank you for bringing out a Linux version of your codecs and software so many years ago. It is appreciated and not quickly forgotten.
Anyhow, after poking at the insides of the RPM and noticing a few worrying scripts that made changes that could potentially not be undone by removing the RPM, I installed it.
Over the next few days I found the initial good impression becoming more and more favourable. To put it bluntly, it wasn't long before it was added as a button along side a small selection of other frequently used software. It was already doing well with me and was up for scoring more points by having open source roots.
Now I'm not going to claim that people like me are in any way pivotal to open source projects. I'm predominately a user but I do like to give back in small ways like filing bug reports and ocassionally participating in bugzilla database grooming. I like being able to trawl a bug database and find an issue I've been having is already known. That way I don't waste time reporting it again and it's nice seeing issues being worked out and not disappearing into a black hole.
I had a few niggles (plugin crashes taking down mozilla) and suggestions so I decided to look at Helix Community to see whether they were known issues. Like all good projects I found that there was a bugzilla (called tracker) which makes following problems so much easier. However when I went to try and look at bugs I got back a "You are not authorized" message. I thought perhaps it was because I didn't have an account. So I went and made an account and tried again after logging in. All attempts to view ANY bugs result in the "you are not authorized" error. All searches seemed to result in nothing found.
Confused, I decided to try the IRC channel and made a brief mention on whether anyone knew about plugin problems and the way tracker wouldn't return any bugs. A good forty-five minutes later no one had answered and I logged off. I know that products have security issues which result in locked bugs but are ALL the bugs in Helix Player of such vital importance that the random public can't peruse them?
It seems such a shame that a lot of effort has been put into creating a community with the useful bits are walled or stone cold to the casual user. Well at any rate community issues are not going to stop me using RealPlayer but it does disuade me from trying to give back.
...when you are on the receiving end of a torrent of spam / virus email.
When someone complains to an ISP and the ISP finds out that the complaint is substantiated, they really have little choice but to do some sort of cutting off to prevent the ongoing attacks on other systems. In an ideal world the ISP would just block up the ports that the were doing the sending but maybe this isn't feasible.
Senderbase provides monthly and dail counts on mail seen from particular IP addresses and thus is capable of spotting when something dramatically out of the (previously known) ordinary happens.
The parent post is correct, Quake 3 DOES work over the network providing the "server" supports GLX extensions. It won't be super fast but it's better than you might expect.
It will certainly be interesting to see what the effect of using OpenGL surfaces for most windows would make... I wonder what effect limited texture memory would make.
One reason could be that gzip can be made to produce rsync friendlier archives.
Often, changing even a small part of an archive will change it sufficently that rsync is forced to retransfer the entire lot. However, there is a patch for gzip (which I believe is in Fedora Core 2 and is scheduled for mainline gzip inclusion too) that produces rsync friendly archives at the cost of slightly larger archives.
I'm a bit confused as to whether you are refering to the original (see the UCB/LBL section) or the revised BSD license.
Under the revised BSD license (which is very similar to the X11 license and is what I am assuming is what you are refering to as the "MIT license") you need only mention copyrights in documentation.
Under the original BSD license you HAVE to mention the copyrights and contributors when the program is used or when the program is advertised.
So will GNOME eventually mandate that its projects not be allowed to ask for copyright assignment? As many others have pointed out, it makes relicencing a pain...
Of course, this is not the same as saying they are the widest used graphics card or that they will sell a large amount of standalone cards but still.
Most people don't buy external sound cards any more but once upon a time everyone did. So those cheapo AC97 based things are ALL over the place - OK Intel don't make them all but they did come up with the AC97 codec.
I've talked to people of various importance who feel that in a certain number of years the graphics card market will go a similar way to the sound card market. The impression I was given was that only people wanting high end quality/speed will go for an extra card but most others will be satisfied by onboard.
It is not really RPMs job to enforce security, it is up to the dependcy manager and both yum and urpmi can be set to do this since those are going to be the things installing stuff automatically.
In yum.conf add gpgcheck=1 and it will check that the GPG signature.
urpmi is set to check signatures by default and normally a repository will tell you what sigs its RPMs are signed with when you add it.
Follow this link about MIME type detection in Internet Explorer. It turns out that IE will sniff the data or filename if the extension is considered to be ambiguous (if it is returned as 'text/plain', 'application/octet-stream', an empty string, or null).
Phil Reginda has an explanation of why IE does this (basically to workaround Apache behaviour) and workaround if you get bitten by this.
I've set up a Brother 1440 on Fedora and it was as simple as installing CUPS ticking the approrpriate options in starting the fedora print tool, saying next a few times and ticking share this printer. It even spotted what printer it was on the parallel port. Works fine and even Windows machines can print to it using IPP (so no it doesn't turn up as samba share and magically send the right drivers to the remote machine but it does work). It wasn't always like that but reporting the hiccups I had when sharing printers led to a few bugs being fixed.
As for why CUPS is so huge, well that's because it tries to support lots of printers in one go. Due to the way Linux distros set it up, it's not modular in the sense that only the modules you need are installed. However it's the same for most distro's kernels and for the pain it would cause to implement your suggestion compared to the saving in disk space it really isn't worth it. It's hard to buy disk drives smaller than 20Gbytes and on such a drive CUPS is taking up.002% of the space. I'm sure you can think of 100 things you'd rather have fixed in your distro over 40M on CUPS.
Does CUPS work smoothly every time? Nope. Is it usually as smooth as installing a printer on Windows? Probably not. Will posting on Slashdot make it better? Doubt it. If you had a problem, the best thing to do is sit down and carefully make your way through the steps and report it back to your distro manufacturer. What distro were you using btw?
Some of these options may cause you to hit problems with buggy web servers. Some of them will up the load on a web server and thus aren't very polite. Others may cripple performance on a weak machine.
I do have to admit to be unsure as to how you can pipeline the first request though. Perhaps you send the first X requests all at once...
Akami appear to be having some DNS issues at the moment so it appears that apple.com, www.google.com, www.microsoft.com etc. are all down. If you have cached their ip addresses you should be able to reach them.
Slashdot Mirror
You clearly have an axe to grind which is fair enough (maybe you've lost valuble data to ext3 disk corruption?). However corruption can be down to more than the filesystem. If you don't actually know what caused the corruption in the first place it's hard to start doing comparisons - e.g. I've seen ext3 corruption several times over the past few years but it has always been down to disks just about to die completely or unknown corner case motherboard bugs.
I've been told by a kernel dev that reiserfs rarely causes corruption on perfect hardware but is far less tolerant of flakey hardware than ext3. What if those new checks are to try and protect people on failling hardware? Continuing to give right results when you're being given wrong data is hard...
What if the kernel has had a dodgy binary driver stuffed into it and it is corrupting bits of kernel memory thus leading to on disk corruption? The kernel filesystem can try and stop a complete and utter blowout but it can only do so many checks...
The initial state of the system is also important. The only times I've known Windows 2000 to have disk corruption there's always been a partition resizer involved at some point.
Finally 100% bug free software is hard to create and even harder to prove. What works for me could quite well fail for you. Neither Ext3, NTFS nor Reiserfs are magic and sometimes things go wrong. Recovery is almost always possible if you are prepared to forgo access to the data and pay a fiscal price.
(Just for the record didn't MS ship NTFS before Ext3 was released into the major distros? Technically MS got to the widespread atomic filesystem first in comparison to Linux...).
Taking a look at the SDL website news shows that version 1.2.8 of the main library was released on December 15th (under a month ago at the time of writing). There also appears to be active development on supporting libraries (such as SDL_mixer, SDL_ttf and SDL_image).
Sure Loki were a major supporter but development didn't stop after Loki disappeared...
I was going to give you a link to where the bug database was but it seems to have turned into a promotional page - Coverity Linux "bug database" page. Normally that link goes to a bugzilla like database showing you a list of bugs, what type of error was found and a highlighted code snippet. All very nifty.
Here's kernel hacker davej talking about coverity bugs. So they do let the kernel folks know about the bugs they have found.
However this hits upon something I don't like about this press release. I wonder whether the headline should be "Tool finds less bugs in Linux than rival OSes when previous bugs found by the tool are fixed in Linux". Assuming fixes for all problems that means only new and changed code gives a chance for these errors to occur (which is a subset of the entire code).
Were all codebases that were compared given coverity bug databases and a chance to fix the reported bugs too and over what time scale? If the Linux kernel code gets 6 monthly reports and say OpenBSD kernel only gets reports one every two years and you test towards the end of those two years is that fair test? What does it say other than our tool finds less bugs in code that is regularly checked by previous versions of our tool? Is that in itself a good justification?
Is this a fixed bug? I just tested the above in a firefox 1.0PR1 in a basic HTML file and nothing nearly as bad as not being able to type happened. In fact, seemingly nothing happened (other than the image never appeared to finish loading).
This is a bit offtopic but what the hey. You ask serveral questions in this post so I my answer may not answer all your points:
To the best of my knowledge Firefox will only offer to migrate IE bookmarks. It doesn't do cookies although if you are willing to mess about by hand I think you can get IE to export it's cookies to a file and then copy the cookies file to the firefox profile.
Will the migration be painless? Who knows? It's been generally painless for me on recent machines but that's no guarantee for you.
Two of the most famous spyware removal utilities are Adware and Spybot search and Destroy.
It is possible using certain languages to formally verify and basically prove that for a certain program a particular set of properties does or does not hold true. I am fairly certain it is not possible to do this with a non trivial program in C (unless you try and annotate it and/or restrict which operations you are allowed to do) though, so this point is a bit moot.
I'm guessing that your guess was right because I just followed your 3 steps, visted a bug and actually saw some results!
Cheers for following this up.
If you are a MandrakeClub member then there has been access to a well packaged version of firefox throughout the 10.0 (the good news is that there is a recent version of firefox in Mandrake 10.1 contribs now). Feodra also had a well packaged version for it over at Fedora Extras. The moral is that if you break away from your distro specific packaging, expect problems...
Here is a cooker mailing list post refering to the bug being resolved.
ACPI works well on my desktop (as does software suspend, only one problem so far) but there are reports of laptops suffering beacuse of a lack of nolapic support. It's also worth noting that many machines come with poorly programmed BIOSes and need updates or extra code to have working ACPI. Why not try a LiveCD distro to test with until you find out you have support? It may also help if you pick a big vendor and report the issue to them so that they can track it and have you test fixes...
Hi Rob,
I suspect you won't get round to reading this since it was posted days after your original post but well, you never know...
I installed RealPlayer 10 Gold about a month ago on my home Linux box. Real Player on Linux has never seemed as scary as the Windows version and I was still using v8 to listen to BBC streams. Let me digress and just thank you for bringing out a Linux version of your codecs and software so many years ago. It is appreciated and not quickly forgotten.
Anyhow, after poking at the insides of the RPM and noticing a few worrying scripts that made changes that could potentially not be undone by removing the RPM, I installed it.
Over the next few days I found the initial good impression becoming more and more favourable. To put it bluntly, it wasn't long before it was added as a button along side a small selection of other frequently used software. It was already doing well with me and was up for scoring more points by having open source roots.
Now I'm not going to claim that people like me are in any way pivotal to open source projects. I'm predominately a user but I do like to give back in small ways like filing bug reports and ocassionally participating in bugzilla database grooming. I like being able to trawl a bug database and find an issue I've been having is already known. That way I don't waste time reporting it again and it's nice seeing issues being worked out and not disappearing into a black hole.
I had a few niggles (plugin crashes taking down mozilla) and suggestions so I decided to look at Helix Community to see whether they were known issues. Like all good projects I found that there was a bugzilla (called tracker) which makes following problems so much easier. However when I went to try and look at bugs I got back a "You are not authorized" message. I thought perhaps it was because I didn't have an account. So I went and made an account and tried again after logging in. All attempts to view ANY bugs result in the "you are not authorized" error. All searches seemed to result in nothing found.
Confused, I decided to try the IRC channel and made a brief mention on whether anyone knew about plugin problems and the way tracker wouldn't return any bugs. A good forty-five minutes later no one had answered and I logged off. I know that products have security issues which result in locked bugs but are ALL the bugs in Helix Player of such vital importance that the random public can't peruse them?
It seems such a shame that a lot of effort has been put into creating a community with the useful bits are walled or stone cold to the casual user. Well at any rate community issues are not going to stop me using RealPlayer but it does disuade me from trying to give back.
...when you are on the receiving end of a torrent of spam / virus email.
When someone complains to an ISP and the ISP finds out that the complaint is substantiated, they really have little choice but to do some sort of cutting off to prevent the ongoing attacks on other systems. In an ideal world the ISP would just block up the ports that the were doing the sending but maybe this isn't feasible.
Senderbase provides monthly and dail counts on mail seen from particular IP addresses and thus is capable of spotting when something dramatically out of the (previously known) ordinary happens.
The parent post is correct, Quake 3 DOES work over the network providing the "server" supports GLX extensions. It won't be super fast but it's better than you might expect.
It will certainly be interesting to see what the effect of using OpenGL surfaces for most windows would make... I wonder what effect limited texture memory would make.
Towards the end of the pixmap bug report you can see that one of the developers provided a fix that appears to have solved the problem.
According to the Changelog the fix went in on 2004-08-27 (Fix for XV memory allocation) and is in the branch for 6.8.0.
One reason could be that gzip can be made to produce rsync friendlier archives.
Often, changing even a small part of an archive will change it sufficently that rsync is forced to retransfer the entire lot. However, there is a patch for gzip (which I believe is in Fedora Core 2 and is scheduled for mainline gzip inclusion too) that produces rsync friendly archives at the cost of slightly larger archives.
If it's reproducable (which it sounds like it is) and you can provide the document then please let the OOo folks know over on their issue tracker.
I'm a bit confused as to whether you are refering to the original (see the UCB/LBL section) or the revised BSD license.
Under the revised BSD license (which is very similar to the X11 license and is what I am assuming is what you are refering to as the "MIT license") you need only mention copyrights in documentation.
Under the original BSD license you HAVE to mention the copyrights and contributors when the program is used or when the program is advertised.
So will GNOME eventually mandate that its projects not be allowed to ask for copyright assignment? As many others have pointed out, it makes relicencing a pain...
Of course, this is not the same as saying they are the widest used graphics card or that they will sell a large amount of standalone cards but still.
Most people don't buy external sound cards any more but once upon a time everyone did. So those cheapo AC97 based things are ALL over the place - OK Intel don't make them all but they did come up with the AC97 codec.
Here's a Register story which mentions that Intel have 31.7% of the graphics card market.
I've talked to people of various importance who feel that in a certain number of years the graphics card market will go a similar way to the sound card market. The impression I was given was that only people wanting high end quality/speed will go for an extra card but most others will be satisfied by onboard.
It is not really RPMs job to enforce security, it is up to the dependcy manager and both yum and urpmi can be set to do this since those are going to be the things installing stuff automatically.
In yum.conf add gpgcheck=1 and it will check that the GPG signature.
urpmi is set to check signatures by default and normally a repository will tell you what sigs its RPMs are signed with when you add it.
Follow this link about MIME type detection in Internet Explorer. It turns out that IE will sniff the data or filename if the extension is considered to be ambiguous (if it is returned as 'text/plain', 'application/octet-stream', an empty string, or null).
Phil Reginda has an explanation of why IE does this (basically to workaround Apache behaviour) and workaround if you get bitten by this.
I've set up a Brother 1440 on Fedora and it was as simple as installing CUPS ticking the approrpriate options in starting the fedora print tool, saying next a few times and ticking share this printer. It even spotted what printer it was on the parallel port. Works fine and even Windows machines can print to it using IPP (so no it doesn't turn up as samba share and magically send the right drivers to the remote machine but it does work). It wasn't always like that but reporting the hiccups I had when sharing printers led to a few bugs being fixed.
.002% of the space. I'm sure you can think of 100 things you'd rather have fixed in your distro over 40M on CUPS.
As for why CUPS is so huge, well that's because it tries to support lots of printers in one go. Due to the way Linux distros set it up, it's not modular in the sense that only the modules you need are installed. However it's the same for most distro's kernels and for the pain it would cause to implement your suggestion compared to the saving in disk space it really isn't worth it. It's hard to buy disk drives smaller than 20Gbytes and on such a drive CUPS is taking up
Does CUPS work smoothly every time? Nope. Is it usually as smooth as installing a printer on Windows? Probably not. Will posting on Slashdot make it better? Doubt it. If you had a problem, the best thing to do is sit down and carefully make your way through the steps and report it back to your distro manufacturer. What distro were you using btw?
Some of these options may cause you to hit problems with buggy web servers. Some of them will up the load on a web server and thus aren't very polite. Others may cripple performance on a weak machine.
I do have to admit to be unsure as to how you can pipeline the first request though. Perhaps you send the first X requests all at once...
Akami appear to be having some DNS issues at the moment so it appears that apple.com, www.google.com, www.microsoft.com etc. are all down. If you have cached their ip addresses you should be able to reach them.