Not a Tivo, get them a ReplayTV 5000 (not the newer 5500) since it will automatically skip commercials, and you can share what you've recorded with friends over the net. It has a scene skip feature (commercial break) and a 30 second skip ahead, so it's even better than Tivo for skipping.
I just downloaded the latest gentoo for sparc64 and am trying to install it on a Sunblade 150. So far all I can say is it's a piece of crap. Everything works until I attempt to create any type of filesystem. fdisk works fine, but often mke2fs locks up tight. It also frequently locks up when it attempts to mount the CDROM when booting up the install CD.
All I can say is I'm not at all happy with this. I've installed and configured many Linux systems on x86 and have spent a lot of time compiling various packages for Solaris on sparc, but so far this is not looking good.
I thought Marvin in the original TV series was perfect. He's "Your plastic pal who's fun to be with" Sure, he looked like a cheap, crappy, prop, but THAT WAS THE IDEA! Marvin made this crashing sound every time he took a step and lumbered around, since all the diodes on his left side were always bothering him.
With that huge head, how's he going to stick it in a bucket of water and wash his head at Zaphod?
Disney just doesn't get it. They ruin just about everything they get their hands on. I imagine they'll try to make the Vogon ships look sleek and cool.
I have the original TV series on Laser Disc. I got it long before DVDs were available.
According to the review, Doom 3 does indeed run on it, but the performance is unplayable with the integrated graphics. However, the fact that it runs at all on the integrated graphics is good. On many integrated chipsets (i.e. Intel) half the games don't work at all.
At least where I am I can demand a paper ballot even though my <a href="http://sfgate.com/cgi-bin/article.cgi?f=/c/<nobr>a<wbr></wbr></nobr> <nobr> <wbr></wbr></nobr>/2004/09/08/BAGN68L64F1.DTL">county uses Deibold</a>. I urge everyone to ask for a paper ballot when voting.
Get out and vote and ask for paper over plastic
on
Pre-Election Discussion
·
· Score: 4, Interesting
I look forward to voting tomorrow. One thing I am thankful for is that I can demand a paper ballot instead of the standard electronic (Diebold) one. I urge everyone who votes tomorrow to ask for a paper ballot, even if they are not available just to make a statement.
Even if you live in a state that is clearly going to one candidate or another, your voice will help add to how strong your state's voice is.
Also don't forget all the local items, where your voice typically is much louder.
And finally, remember that you're not just voting for a candidate, but for all their support staff. For example, a vote for Bush is also a vote for Michael Powell, John Ashcroft, Dick Cheney, Carl Rove and all the other people that come along for the ride. Not to mention that the next president will likely select one or more supreme court justices.
After months of problems with DMA timeouts and lockups caused by using a Highpoint RAID controller and a Promise IDE controller I finally bit the bullet and bought a 3Ware Escalade controller. All the sudden, everything is completely stable.
Do yourself a favor and get a good hardware raid controller and make sure it has good Linux support. Promise sucks. They advertise Linux support on the box - they lie, only with specific 2.4 kernels. 3Ware has good driver support for Linux included with the Linux kernel source code.
I too upgraded to SuSE 9.1 and had a lot of problems, many due to hardware compatibility issues with the 2.6 kernel. Some of the issues were fixed by the online updates, which includes a newer kernel.
I have been a long-time SuSE user and just ordered the 9.2 upgrade, which I expect to be better.
In my long time experience with SuSE, some versions are just not stable.
8.0 was great, 8.1 sucked, 8.2 and 9.0 were good, 9.1 not as good.
I have a SuSE 8.2 server that has been up 451 days without a reboot. I use it for file, NFS, web, FTP, ssh, DNS, imap, smtp, spam filter, and as a remote applications server for running Mozilla.
I agree that C++ can be a good thing. A number of years ago I worked on an Asynchronous Transfer Mode (ATM) networking driver which included the full ATM stack and Lan Emulation. The whole driver, except for the lowest level base driver, was written entirely in C++. C++ kept the code simple and modular and cleaner than the equivalent C code. When I had to expand the LAN Emulation code to handle multiple ELANS, it took very little time, just changing a pointer to the LANE code to an array of pointers.
I had to debug this code at the assembler level in many cases on the platform the code was written (OS/2), and I did not see any significant overhead caused by the use of C++. Then again, this was using the Watcom 16-bit C++ compiler and not G++. (Note: most OS/2 device drivers were 16-bit, even though the OS kernel was mostly 32-bits).
Of course, the C++ data path code was written to be very flat without a lot of inheritance, no multiple inheritance, no templates, exception handling, etc., so as to be lean, fast and transparent. Other less performance critical code used more of the C++ features.
At the same time we had another driver we worked on for Windows NT and Solaris which was written entirely in C. The C++ driver was 100K lines of code. The C driver was over 350K lines of code. The C++ driver was much smaller (binary) and faster, not to mention, it was far more stable and had more features. By using inheritance, the LANE code could emulate either Ethernet or Token Ring or both on a per-elan basis, with the particular network personality being subclassed from the main LANE class.
Both drivers code bases were written to be quite portable as well.
Usually when handling multiple devices you keep passing a pointer to a data structure for the device for each function. In C++, you can think of that as the this pointer in a class.
There is no reason for C++ code to be slow if it is written with performance in mind.
My parents have a Bengal cat. One thing about Bengals is they are also hypoallergenic for most people. A friend of the family was deathly allergic to cats, yet this cat didn't bother him at all, even though the cat would rub all over his head whenever he sat on the couch and sleep on his bed when he stayed over.
Bengals are a lot less than $3500, but they can be quite a handfull in terms of temperament. Bengals are a cross between a housecat and an Asian leopard.
A few years ago I worked in a lab with one other engineer. No matter how many times we complained, it was always way too cold in there.
At this company we made networking adapters. What we did was to install a hook in the ceiling and connect a CAT 3 cable to a networking card. We taped a paper towel to the card and had a coffee mug full of heat facilitating fluid (water) which was applied to the paper towel. By pulling on the CAT3 wire, we could raise and lower the paper towel over the thermostat and hence warm it up in the lab. (CAT5 wire was too stiff).
By cooling the thermostat, we could warm up the lab.
For the first time ever, the heater actually came on. You could smell that odor of burnt dust when a furnace comes on after not having run for a long period of time.
-Aaron
Update TCP, don't add new protocol
on
Replacing TCP?
·
· Score: 3, Informative
While there are a number of issues with TCP, I think it would be much better in the long run to work on fixing TCP rather than replace it. That way all the existing apps can take advantage of the fixes.
One thing that bothers me is I see ISPs applying policing to their subscriber's bandwidth. Policing is quite unfriendly to TCP, unlike, say, shaping. With policing, a router decides either to pass, drop, or mark a packet based on if it exceeds certain bandwidth constraints. Shaping, on the other hand, will buffer packets and introduce additional latency, thus helping TCP find the sweet spot. Of course shaping will also drop, since nobody provides infinite buffer space.
TCP is relatively easy to extend. There are still some free flag bits and additional fields can be added to the TCP header if needed.
I could just see it, someone makes an auto-TV turn-off device that's sound activated. Everytime everyone cheers, the TV goes off or changes channels to the Home Shopping Network.
At my company we are using Timesys embedded Linux for its real-time options. For telecom, it is absolutely necessary to have real-time in many cases. I.e. if a packet is delayed, it will cause a glitch in the audio or somesuch. The major problem we have with Timesys is that their kernel is a bit out of date (2.4.18) and is missing some critical patches and that they have not applied their changes to the 2.6 kernel tree.
Other things are things like priority inheritance support, to prevent problems caused by priority inversion (which caused problems in the original Mars rover).
For voice support, if you don't mind crappy sound or are only handling one or two calls, you can get away without real-time, but for serious use, it is essential.
Maybe for control path processing it isn't essential, but as soon as it becomes part of the data path, real-time is essential.
Linus has pooh-poohed other important patches in the past as well. That doesn't mean it doesn't go into the kernel, it just means that one needs to demonstrate the advantages of it to Linus and show it doesn't cause other problems.
We are currently successfully using Timesys Linux in an embedded product and are having great success (despite problems with Timesys). Realtime support is essential for telecom and networking applications. It also is quite useful for multimedia, so that video appears smooth and not jumpy. Granted, realtime is often not very useful for a general purpose computer, but in the embedded world, it is extremely useful.
I'd like you to take a look at the kdebindings support. It includes bindings for Perl, Python, C, C#, Java, and more. If that's not enough, I don't know what is.
However, I have yet to find anything that really uses those bindings. I guess one could conceivably write Python, C, or Java KDE programs, but I personally am not aware of any of them. I will say that in my experience that the kdebindings package is huge, having compiled it on Solaris.
Speaking of Gnome, I gave up trying to compile recent Gnome versions for Solaris 8. Too many problems compiling since Sun's X doesn't support the render extension. KDE/QT compile just fine. If Sun is supporting Gnome, they're doing a rather poor job of it. We have been trying for ages to compile a recent version of Evolution for Solaris since they downgraded us to Exchange at work. Our ancient crusty Netscape Calendar/mail worked much better in many ways than Exchange (especially for Unix users), even though the calendar only ran with Netscape 4.6x.
While many agree that nobody is exactly sure what will happen with increased CO2 in the atmosphere, most models predict major problems. Should we wait until someone develops a perfect model, which might not be possible due to limited computing power, or do something about it now. If there's a 50% chance that widespread damage will occur due to global warming (i.e. rising oceans, more hurricanes/tornadoes, stronger weather, more drought, frozen Europe), then doesn't it make a lot of sense to try and limit the damage we cause?
Bush and others claim huge costs with complying with Kyoto. I don't buy it. Some companies (Du Pont, BP) are already complying and have found that they're saving money because they use less energy. Sure, hybrid vehicles cost a bit more to produce than regular vehicles, but I think the extra efficiency will more than make up for it over the life of the vehicle, especially with the rising cost of oil.
And using the fact that China doesn't have to comply is just an excuse. If the rest of the world follows Kyoto, it will help encourage China and those who don't to follow. Not only that, it will make the technology they need to comply cheaper.
I've seen a number of articles about other effects caused by the warming of the oceans. For example, while melting the polar ice cap won't in itself raise the oceans, it will raise the temperature further since ice reflects the sunlight back out into space whereas water absorbs it. A rise in the ocean temperature could cause massive amounts of methane to enter the atmosphere from all the methane hydrides at the ocean floor, and methane is a much bigger greenhouse gas than CO2.
Ignorance of the full global warming effects is no excuse for inaction when we have enough evidence that serious problems are likely.
Maybe we should also try and aid Indonesia with putting out their peat bog fires, which are releasing huge amounts of CO2 as a start.
As it is, today I was contemplating replacing my ancient inefficient refrigerator with a new efficient one. I think I'll go ahead and do it (it helps that Orchard Supply has a tax-free day today when sales tax is 8.75%).
I'm not sure I would call VxWorks a microkernel. In VxWorks, everything is in the same address space. There's basically no memory protection (unless you count the bastardized VxWorks AE). Think of VxWorks as the ultimate monolithic environment, where everything goes into the kernel.
Re:I wonder if it's true real-time
on
RT Linux Patches
·
· Score: 2, Informative
I hate to respond to my own post, but reading the article (what, someone rtfa) it seems that they did add priority inheritance to the kernel mutexes avoid priority inversion. For proper real-time support, hopefully they also made this available to user-space threads.
-Aaron
I wonder if it's true real-time
on
RT Linux Patches
·
· Score: 5, Interesting
Where I work we did a project using Timesys Linux which implements true real-time support and has some really cool scheduler options. For example, with Timesys, you can, for example, guarantee that a task will get a minimum of 15.7ms execution time every 31ms. It even allows you to set priorities for interrupts, such that an interrupt can be scheduled at a lower priority than a user thread. And finally, they added support for priority inheritance to avoid the problem of priority inversion, which occurs when a low priority thread has acquired a semaphore and a high priority thread blocks on it.
Not only can you reserve CPU bandwidth, but also network bandwidth. Of course it also has all the other standard features one would expect of a real-time OS.
Sadly, Timesys has not applied their patches yet to the 2.6 kernel at this time.
I most certainly remember Cyberpromo, Agis, and the rest of the early ilk. At the time I was rather inexperienced with the intricacies of email and thought I was being clever to create a filter to forward Cyberpromo's crap back to any email addresses I could find. Unfortunately (or maybe fortunately) it created a mail loop. The company I worked for had a T-1 line. It ended up filling up the T-1 line redirecting email back to Cyberpromo and knocked them off the net for a bit. They called my employer, threatening to sue. I reluctantly turned off my bounce filter.
I don't think any of today's spammers are anywhere near what Cyberpromo and some of the other early spammers were doing. I'd often get over 100 messages a day in my inbox from only a few spam outfits (usually customers of Agis). As far as I can tell, none of the new bottom feeders are anywhere near what Cyberpromo or the other early spammers did.
I remember when they tried to set up a do-not-spam database, which they used as a do-spam database.
I will feel no pity for Spammy if they lock him up and throw away the key. Hopefully he'll share his cell with a nice big inmate named Bubba and they'll feed him nothing but some mystery meat and play Weird Al's Spam song full blast in his cell 24 hours a day and only give earplugs to Bubba.
This article is case in point, it takes multiple large corporations to fight for our rights - because you're rights are something that cost money, lots and lots and lots of money. More money then the average man would have, that's for certain.
Huh? If anything, large companies are the ones fighting to eliminate our rights. For example, Disney and a bunch of companies turned around and sued ReplayTV because the Replay allowed one to share what they recorded and it had the capability to automatically skip commercials. Big companies sued over the early MP3 players (and lost) in an effort to prevent the spread of MP3 music. Disney and friends tried to force Replay to record everything the PVR users did (every button press, etc.) in addition to banning the sharing capability and commercial skip feature.
Big companies are only for freedom if it helps their bottom line. In many, if not most cases, they are against expanded freedoms. Who do you think is fighting for the broadcast flag? Large corporations. Who fought for adding the DAT copy protection? Large corporations. Who is fighting P2P networks? Large corporations.
Unfortunately, the large corporations are often writing the laws and winning because they do have the money behind them.
Bush has totally screwed things up, both before and after 9/11.
Before 9/11, his administration took its eye off of the terrorist threat. Unlike Clinton, who had daily briefings on the terrorist threat, Bush didn't have one until September. On 9/10, John Ashcroft wanted to reduce funding for terrorism investigation.
Like John Kerry said in the debate, we left much of the Afghanistan operation up to the warlords, who let many of Al Queda slip away, either through incompetence or through bribery.
Even today, the Taliban are returning to Afghanistan and much of the country is no better than it was before the invasion, if you don't count the capital Kabul.
Saddam Hussein was a bad man, but was not at all behind 9/11. By the time we invaded Iraq, the weapons inspectors were back and searching the country, in part due to the threat of force from President Bush. Saddam would not be able to manufacture anything while the inspectors were there. Also, major portions of Iraq were effectively outside the control of Saddam, with the no-fly zones and whatnot. The Kurds were effectively governing themselves.
In fact, I would say that there were many other countries that were far more of a threat than Iraq, like North Korea, who will sell anything to the highest bidder, or Pakastan, who did sell nuclear technology right under our nose.
Saddam was not a threat to the US, nor would he likely become one with the inspectors on the ground. Even so, we should have first finished up in Afghanistan before moving on to Iraq. Besides, what country was more responsible for 9/11? What about Saudi Arabia? Most of the hijackers came from Saudi Arabia as did much of their funding. Much of the fundamentalist intolerant Islam comes out of that country. They set up madrasses to teach their religion in poor countries like Afghanistan where they teach the Koran, but not much else. They don't even teach things like that the Earth is round (many don't know this) and to hate the United States and the West. The Saudi school books teach hatred, and it continued even after 9/11.
Invading Iraq was a major mistake. For all those innocent civilians who are hurt or killed, how many people are turned to hate the US? You kill a mother, a father, a child, brother, or sister, and now you have a whole family mad at you, not to mention all their friends and neighbors.
As bad as Saddam was, he kept Iraq under control and kept it out of the hands of religious fundamentalists, including Al Queda. He was a problem, but he was Iraq's problem, not ours. None of his neighbors was particularly worried about him after '91. None of his neighbors is in the coalition of the willing (Poland does not border Iraq).
After 9/11 I would much rather see a lot more troops in Afghanistan weeding out Al Queda and the Taliban. I also would want to see very strong action taken against Saudi Arabia who was behind much of the funding and their peddling of hate. We had a strong opportunity to unite the world closer together after 9/11 to stamp out terrorism and Bush totally squandered it. He rejected other countries offering troops and supplies to weed out the terrorists in Afghanistan and went after Iraq instead (who Bush had wanted to invade long before 9/11).
I was recently involved in setting up a corporation and rather than set it up in my own state, there were advantages, other than taxes, for setting it up in Nevada. You will notice, for example, that many companies are incorporated in Deleware or elsewhere, often for the better legal protection provided by that state's laws. For example, trying to sue a Nevada corporation may be more difficult than many other states.
In our case, taxes were not the intent at all. We still pay local state taxes as well, so the savings are not that significant. There is some tax savings since some of the taxes are paid to Nevada instead, but nothing significant.
Now what I do feel bad about is how some companies set up their offices offshore in places like the Caymen islands to avoid federal income taxes or other federal laws. If a US based company does this, then they should not get the benefits of being a US company. I also feel that the federal government should not be allowed to sign contracts with companies that do this. I.e. why should my tax dollars go to Haliburton when the company sets up offices (usually just a mail stop) in places like the Caymen islands or elsewhere to not only avoid paying US taxes, but to also circumvent US laws and do business in places like Iran.
Strangely enough both of my cats avoid my open computer cases and they both wear magnets. I installed a cat door that requires a magnet to be worn on the cat's collar for it to permit entry. The magnet works great for other things as well. My cats always seem to find those little springs and screws I lose in the carpet. Also, my cats seem to have found a never ending supply of bent nails. I have no idea where they pick them up from.
The magnets have worked well to keep out the other neighborhood cats and a couple of raccoons who tried to chase my cat inside.
It can also be comical to watch how the cat reacts when the magnet comes into contact with a magnetic object. I quickly learned to remove spoons from bowls before offering to let my cats pre-rinse them.
Note that Cat 1.0 is not always compatible with Cat 2.0. Often random behavior will occur. Entertainment factor = # of cats squared.
Slashdot Mirror
Not a Tivo, get them a ReplayTV 5000 (not the newer 5500) since it will automatically skip commercials, and you can share what you've recorded with friends over the net. It has a scene skip feature (commercial break) and a 30 second skip ahead, so it's even better than Tivo for skipping.
-Aaron
I just downloaded the latest gentoo for sparc64 and am trying to install it on a Sunblade 150. So far all I can say is it's a piece of crap. Everything works until I attempt to create any type of filesystem. fdisk works fine, but often mke2fs locks up tight. It also frequently locks up when it attempts to mount the CDROM when booting up the install CD.
All I can say is I'm not at all happy with this. I've installed and configured many Linux systems on x86 and have spent a lot of time compiling various packages for Solaris on sparc, but so far this is not looking good.
I thought Marvin in the original TV series was perfect. He's "Your plastic pal who's fun to be with" Sure, he looked like a cheap, crappy, prop, but THAT WAS THE IDEA! Marvin made this crashing sound every time he took a step and lumbered around, since all the diodes on his left side were always bothering him.
With that huge head, how's he going to stick it in a bucket of water and wash his head at Zaphod?
Disney just doesn't get it. They ruin just about everything they get their hands on. I imagine they'll try to make the Vogon ships look sleek and cool.
I have the original TV series on Laser Disc. I got it long before DVDs were available.
According to the review, Doom 3 does indeed run on it, but the performance is unplayable with the integrated graphics. However, the fact that it runs at all on the integrated graphics is good. On many integrated chipsets (i.e. Intel) half the games don't work at all.
At least where I am I can demand a paper ballot even though my <a href="http://sfgate.com/cgi-bin/article.cgi?f=/c/<nobr>a<wbr></wbr></nobr> <nobr> <wbr></wbr></nobr>/2004/09/08/BAGN68L64F1.DTL">county uses Deibold</a>. I urge everyone to ask for a paper ballot when voting.
I look forward to voting tomorrow. One thing I am thankful for is that I can demand a paper ballot instead of the standard electronic (Diebold) one. I urge everyone who votes tomorrow to ask for a paper ballot, even if they are not available just to make a statement.
Even if you live in a state that is clearly going to one candidate or another, your voice will help add to how strong your state's voice is.
Also don't forget all the local items, where your voice typically is much louder.
And finally, remember that you're not just voting for a candidate, but for all their support staff. For example, a vote for Bush is also a vote for Michael Powell, John Ashcroft, Dick Cheney, Carl Rove and all the other people that come along for the ride. Not to mention that the next president will likely select one or more supreme court justices.
After months of problems with DMA timeouts and lockups caused by using a Highpoint RAID controller and a Promise IDE controller I finally bit the bullet and bought a 3Ware Escalade controller. All the sudden, everything is completely stable.
Do yourself a favor and get a good hardware raid controller and make sure it has good Linux support. Promise sucks. They advertise Linux support on the box - they lie, only with specific 2.4 kernels. 3Ware has good driver support for Linux included with the Linux kernel source code.
-Aaron
I too upgraded to SuSE 9.1 and had a lot of problems, many due to hardware compatibility issues with the 2.6 kernel. Some of the issues were fixed by the online updates, which includes a newer kernel.
I have been a long-time SuSE user and just ordered the 9.2 upgrade, which I expect to be better.
In my long time experience with SuSE, some versions are just not stable.
8.0 was great, 8.1 sucked, 8.2 and 9.0 were good, 9.1 not as good.
I have a SuSE 8.2 server that has been up 451 days without a reboot. I use it for file, NFS, web, FTP, ssh, DNS, imap, smtp, spam filter, and as a remote applications server for running Mozilla.
-Aaron
I agree that C++ can be a good thing. A number of years ago I worked on an Asynchronous Transfer Mode (ATM) networking driver which included the full ATM stack and Lan Emulation. The whole driver, except for the lowest level base driver, was written entirely in C++. C++ kept the code simple and modular and cleaner than the equivalent C code. When I had to expand the LAN Emulation code to handle multiple ELANS, it took very little time, just changing a pointer to the LANE code to an array of pointers.
I had to debug this code at the assembler level in many cases on the platform the code was written (OS/2), and I did not see any significant overhead caused by the use of C++. Then again, this was using the Watcom 16-bit C++ compiler and not G++. (Note: most OS/2 device drivers were 16-bit, even though the OS kernel was mostly 32-bits).
Of course, the C++ data path code was written to be very flat without a lot of inheritance, no multiple inheritance, no templates, exception handling, etc., so as to be lean, fast and transparent. Other less performance critical code used more of the C++ features.
At the same time we had another driver we worked on for Windows NT and Solaris which was written entirely in C. The C++ driver was 100K lines of code. The C driver was over 350K lines of code. The C++ driver was much smaller (binary) and faster, not to mention, it was far more stable and had more features. By using inheritance, the LANE code could emulate either Ethernet or Token Ring or both on a per-elan basis, with the particular network personality being subclassed from the main LANE class.
Both drivers code bases were written to be quite portable as well.
Usually when handling multiple devices you keep passing a pointer to a data structure for the device for each function. In C++, you can think of that as the this pointer in a class.
There is no reason for C++ code to be slow if it is written with performance in mind.
-Aaron
My parents have a Bengal cat. One thing about Bengals is they are also hypoallergenic for most people. A friend of the family was deathly allergic to cats, yet this cat didn't bother him at all, even though the cat would rub all over his head whenever he sat on the couch and sleep on his bed when he stayed over.
Bengals are a lot less than $3500, but they can be quite a handfull in terms of temperament. Bengals are a cross between a housecat and an Asian leopard.
-Aaron
A few years ago I worked in a lab with one other engineer. No matter how many times we complained, it was always way too cold in there.
At this company we made networking adapters. What we did was to install a hook in the ceiling and connect a CAT 3 cable to a networking card. We taped a paper towel to the card and had a coffee mug full of heat facilitating fluid (water) which was applied to the paper towel. By pulling on the CAT3 wire, we could raise and lower the paper towel over the thermostat and hence warm it up in the lab. (CAT5 wire was too stiff).
By cooling the thermostat, we could warm up the lab.
For the first time ever, the heater actually came on. You could smell that odor of burnt dust when a furnace comes on after not having run for a long period of time.
-Aaron
While there are a number of issues with TCP, I think it would be much better in the long run to work on fixing TCP rather than replace it. That way all the existing apps can take advantage of the fixes.
One thing that bothers me is I see ISPs applying policing to their subscriber's bandwidth. Policing is quite unfriendly to TCP, unlike, say, shaping. With policing, a router decides either to pass, drop, or mark a packet based on if it exceeds certain bandwidth constraints. Shaping, on the other hand, will buffer packets and introduce additional latency, thus helping TCP find the sweet spot. Of course shaping will also drop, since nobody provides infinite buffer space.
TCP is relatively easy to extend. There are still some free flag bits and additional fields can be added to the TCP header if needed.
-Aaron
I could just see it, someone makes an auto-TV turn-off device that's sound activated. Everytime everyone cheers, the TV goes off or changes channels to the Home Shopping Network.
-Aaron
At my company we are using Timesys embedded Linux for its real-time options. For telecom, it is absolutely necessary to have real-time in many cases. I.e. if a packet is delayed, it will cause a glitch in the audio or somesuch. The major problem we have with Timesys is that their kernel is a bit out of date (2.4.18) and is missing some critical patches and that they have not applied their changes to the 2.6 kernel tree.
Other things are things like priority inheritance support, to prevent problems caused by priority inversion (which caused problems in the original Mars rover).
For voice support, if you don't mind crappy sound or are only handling one or two calls, you can get away without real-time, but for serious use, it is essential.
Maybe for control path processing it isn't essential, but as soon as it becomes part of the data path, real-time is essential.
Linus has pooh-poohed other important patches in the past as well. That doesn't mean it doesn't go into the kernel, it just means that one needs to demonstrate the advantages of it to Linus and show it doesn't cause other problems.
We are currently successfully using Timesys Linux in an embedded product and are having great success (despite problems with Timesys). Realtime support is essential for telecom and networking applications. It also is quite useful for multimedia, so that video appears smooth and not jumpy. Granted, realtime is often not very useful for a general purpose computer, but in the embedded world, it is extremely useful.
-Aaron
I'd like you to take a look at the kdebindings support. It includes bindings for Perl, Python, C, C#, Java, and more. If that's not enough, I don't know what is.
However, I have yet to find anything that really uses those bindings. I guess one could conceivably write Python, C, or Java KDE programs, but I personally am not aware of any of them. I will say that in my experience that the kdebindings package is huge, having compiled it on Solaris.
Speaking of Gnome, I gave up trying to compile recent Gnome versions for Solaris 8. Too many problems compiling since Sun's X doesn't support the render extension. KDE/QT compile just fine. If Sun is supporting Gnome, they're doing a rather poor job of it. We have been trying for ages to compile a recent version of Evolution for Solaris since they downgraded us to Exchange at work. Our ancient crusty Netscape Calendar/mail worked much better in many ways than Exchange (especially for Unix users), even though the calendar only ran with Netscape 4.6x.
While many agree that nobody is exactly sure what will happen with increased CO2 in the atmosphere, most models predict major problems. Should we wait until someone develops a perfect model, which might not be possible due to limited computing power, or do something about it now. If there's a 50% chance that widespread damage will occur due to global warming (i.e. rising oceans, more hurricanes/tornadoes, stronger weather, more drought, frozen Europe), then doesn't it make a lot of sense to try and limit the damage we cause?
Bush and others claim huge costs with complying with Kyoto. I don't buy it. Some companies (Du Pont, BP) are already complying and have found that they're saving money because they use less energy. Sure, hybrid vehicles cost a bit more to produce than regular vehicles, but I think the extra efficiency will more than make up for it over the life of the vehicle, especially with the rising cost of oil.
And using the fact that China doesn't have to comply is just an excuse. If the rest of the world follows Kyoto, it will help encourage China and those who don't to follow. Not only that, it will make the technology they need to comply cheaper.
I've seen a number of articles about other effects caused by the warming of the oceans. For example, while melting the polar ice cap won't in itself raise the oceans, it will raise the temperature further since ice reflects the sunlight back out into space whereas water absorbs it. A rise in the ocean temperature could cause massive amounts of methane to enter the atmosphere from all the methane hydrides at the ocean floor, and methane is a much bigger greenhouse gas than CO2.
Ignorance of the full global warming effects is no excuse for inaction when we have enough evidence that serious problems are likely.
Maybe we should also try and aid Indonesia with putting out their peat bog fires, which are releasing huge amounts of CO2 as a start.
As it is, today I was contemplating replacing my ancient inefficient refrigerator with a new efficient one. I think I'll go ahead and do it (it helps that Orchard Supply has a tax-free day today when sales tax is 8.75%).
-Aaron
I'm not sure I would call VxWorks a microkernel. In VxWorks, everything is in the same address space. There's basically no memory protection (unless you count the bastardized VxWorks AE). Think of VxWorks as the ultimate monolithic environment, where everything goes into the kernel.
I hate to respond to my own post, but reading the article (what, someone rtfa) it seems that they did add priority inheritance to the kernel mutexes avoid priority inversion. For proper real-time support, hopefully they also made this available to user-space threads.
-Aaron
Where I work we did a project using Timesys Linux which implements true real-time support and has some really cool scheduler options. For example, with Timesys, you can, for example, guarantee that a task will get a minimum of 15.7ms execution time every 31ms. It even allows you to set priorities for interrupts, such that an interrupt can be scheduled at a lower priority than a user thread. And finally, they added support for priority inheritance to avoid the problem of priority inversion, which occurs when a low priority thread has acquired a semaphore and a high priority thread blocks on it.
Not only can you reserve CPU bandwidth, but also network bandwidth. Of course it also has all the other standard features one would expect of a real-time OS.
Sadly, Timesys has not applied their patches yet to the 2.6 kernel at this time.
-Aaron
I most certainly remember Cyberpromo, Agis, and the rest of the early ilk. At the time I was rather inexperienced with the intricacies of email and thought I was being clever to create a filter to forward Cyberpromo's crap back to any email addresses I could find. Unfortunately (or maybe fortunately) it created a mail loop. The company I worked for had a T-1 line. It ended up filling up the T-1 line redirecting email back to Cyberpromo and knocked them off the net for a bit. They called my employer, threatening to sue. I reluctantly turned off my bounce filter.
I don't think any of today's spammers are anywhere near what Cyberpromo and some of the other early spammers were doing. I'd often get over 100 messages a day in my inbox from only a few spam outfits (usually customers of Agis). As far as I can tell, none of the new bottom feeders are anywhere near what Cyberpromo or the other early spammers did.
I remember when they tried to set up a do-not-spam database, which they used as a do-spam database.
I will feel no pity for Spammy if they lock him up and throw away the key. Hopefully he'll share his cell with a nice big inmate named Bubba and they'll feed him nothing but some mystery meat and play Weird Al's Spam song full blast in his cell 24 hours a day and only give earplugs to Bubba.
-Aaron
Huh? If anything, large companies are the ones fighting to eliminate our rights. For example, Disney and a bunch of companies turned around and sued ReplayTV because the Replay allowed one to share what they recorded and it had the capability to automatically skip commercials. Big companies sued over the early MP3 players (and lost) in an effort to prevent the spread of MP3 music. Disney and friends tried to force Replay to record everything the PVR users did (every button press, etc.) in addition to banning the sharing capability and commercial skip feature.
Big companies are only for freedom if it helps their bottom line. In many, if not most cases, they are against expanded freedoms. Who do you think is fighting for the broadcast flag? Large corporations. Who fought for adding the DAT copy protection? Large corporations. Who is fighting P2P networks? Large corporations.
Unfortunately, the large corporations are often writing the laws and winning because they do have the money behind them.
Before 9/11, his administration took its eye off of the terrorist threat. Unlike Clinton, who had daily briefings on the terrorist threat, Bush didn't have one until September. On 9/10, John Ashcroft wanted to reduce funding for terrorism investigation.
Like John Kerry said in the debate, we left much of the Afghanistan operation up to the warlords, who let many of Al Queda slip away, either through incompetence or through bribery.
Even today, the Taliban are returning to Afghanistan and much of the country is no better than it was before the invasion, if you don't count the capital Kabul.
Saddam Hussein was a bad man, but was not at all behind 9/11. By the time we invaded Iraq, the weapons inspectors were back and searching the country, in part due to the threat of force from President Bush. Saddam would not be able to manufacture anything while the inspectors were there. Also, major portions of Iraq were effectively outside the control of Saddam, with the no-fly zones and whatnot. The Kurds were effectively governing themselves.
In fact, I would say that there were many other countries that were far more of a threat than Iraq, like North Korea, who will sell anything to the highest bidder, or Pakastan, who did sell nuclear technology right under our nose.
Saddam was not a threat to the US, nor would he likely become one with the inspectors on the ground. Even so, we should have first finished up in Afghanistan before moving on to Iraq. Besides, what country was more responsible for 9/11? What about Saudi Arabia? Most of the hijackers came from Saudi Arabia as did much of their funding. Much of the fundamentalist intolerant Islam comes out of that country. They set up madrasses to teach their religion in poor countries like Afghanistan where they teach the Koran, but not much else. They don't even teach things like that the Earth is round (many don't know this) and to hate the United States and the West. The Saudi school books teach hatred, and it continued even after 9/11.
Invading Iraq was a major mistake. For all those innocent civilians who are hurt or killed, how many people are turned to hate the US? You kill a mother, a father, a child, brother, or sister, and now you have a whole family mad at you, not to mention all their friends and neighbors.
As bad as Saddam was, he kept Iraq under control and kept it out of the hands of religious fundamentalists, including Al Queda. He was a problem, but he was Iraq's problem, not ours. None of his neighbors was particularly worried about him after '91. None of his neighbors is in the coalition of the willing (Poland does not border Iraq).
Also, I strongly urge you to look at the other reasons we invaded Iraq. A very good article can be found at http://www.harpers.org/BaghdadYearZero.html.
After 9/11 I would much rather see a lot more troops in Afghanistan weeding out Al Queda and the Taliban. I also would want to see very strong action taken against Saudi Arabia who was behind much of the funding and their peddling of hate. We had a strong opportunity to unite the world closer together after 9/11 to stamp out terrorism and Bush totally squandered it. He rejected other countries offering troops and supplies to weed out the terrorists in Afghanistan and went after Iraq instead (who Bush had wanted to invade long before 9/11).
I was recently involved in setting up a corporation and rather than set it up in my own state, there were advantages, other than taxes, for setting it up in Nevada. You will notice, for example, that many companies are incorporated in Deleware or elsewhere, often for the better legal protection provided by that state's laws. For example, trying to sue a Nevada corporation may be more difficult than many other states.
In our case, taxes were not the intent at all. We still pay local state taxes as well, so the savings are not that significant. There is some tax savings since some of the taxes are paid to Nevada instead, but nothing significant.
Now what I do feel bad about is how some companies set up their offices offshore in places like the Caymen islands to avoid federal income taxes or other federal laws. If a US based company does this, then they should not get the benefits of being a US company. I also feel that the federal government should not be allowed to sign contracts with companies that do this. I.e. why should my tax dollars go to Haliburton when the company sets up offices (usually just a mail stop) in places like the Caymen islands or elsewhere to not only avoid paying US taxes, but to also circumvent US laws and do business in places like Iran.
Strangely enough both of my cats avoid my open computer cases and they both wear magnets. I installed a cat door that requires a magnet to be worn on the cat's collar for it to permit entry. The magnet works great for other things as well. My cats always seem to find those little springs and screws I lose in the carpet. Also, my cats seem to have found a never ending supply of bent nails. I have no idea where they pick them up from.
The magnets have worked well to keep out the other neighborhood cats and a couple of raccoons who tried to chase my cat inside.
It can also be comical to watch how the cat reacts when the magnet comes into contact with a magnetic object. I quickly learned to remove spoons from bowls before offering to let my cats pre-rinse them.
Note that Cat 1.0 is not always compatible with Cat 2.0. Often random behavior will occur. Entertainment factor = # of cats squared.