"The reason for this is to give a visual hint as to what is a comment and what is actual code. The rest of the syntax highlighting is less important in this case. While porting, never delete a line of code. Comment it out and make your changes below it. If you need to, comment out whole functions. You are going to find that having an immediate reference to the original code is immensely helpful, but all that text can get very confusing. Syntax highlighting makes everything much easier. "
I couldn't disagree more. Large amounts of commented out code make programs unreadable. Why was it commented out? Was it test code? An older version? Something that wouldn't compile but needs to be fixed? Who knows...often not even the orignal commenter after a few months. Don't think that you will go back in fix it because you will not. There is almost never time to fix badly written code that seems to work because there is always new code to write. You will net get time to look at it again until a bug pops up. And then you will be left wondering what in the world was in that commented code. Writing things the correct way the first time makes life much easier.
Comments should be used for comments. Some text explaining code. They should be used for turning code on and off (use precompiler statments, much more flexible) and they should not be used to save old code. Try using a source repository that has version diff support like cvs. This is a much cleaner way to see what you had before without leaving your old code lying around in massive/* */ blocks.
As for comment syntax highlighting being the only usefull part of syntax highlighting, I have to disagree again. I like to see different parts of the lanugages syntax in different colors. Hence the name "syntax" highlighting, not comment highlighting.
What version of macro economics did you study? Oh... the one where government has no effect on the economy. So you feel that the congress' budget, the FED's policies, the President's domestic and foreign policies, and the court's enforcement of financial and criminal law have nothing to do with our economic prosperity?
Ok, then what in the world does?
Ok, given these crimes, what are you going to do about it? How are you going to get involved in order to make things better? If all the things you mentioned are really happenening and you do nothing, then aren't you really just complaining? Seriously, if the FBI is breaking the law, then you need to let the people that write their checks know. Write to the government. Get lots of people to sign a petition. You know, get really involved with the democratic process.
Last I checked it was our country, not just yours. This means I have just as much a right to voice my opinion as you. If you can not handle a reading a different point of view without resorting to foul name calling then it sounds like you need to go to a place where there are other people that don't like to hear different opinions. I think that China fits that description.
By "chilling out", I did not mean do nothing, I just meant that a person should act rationally. Learn the system, understand the system, and then you can fix the system.
Hello new Frenchie, ex-American!
IMHO, high unemployment is a sign of a lack of overall oppertunity that the US offers. Most people who say that they can not get a job in the US really mean that they can not get a job that they like and want. It does sound like France may have oppertunities in the tech. world but then again, the same can be said for several other countries. But what about a chance of building a successful business for yourself? Moving up in a non-technical field? The US has these oppertunities in large quantities.
As for a transfer, my company is partially owned by a Swiss company and has several Eurpean offices. Maybe I will stop by and see how things are!:-) Then I could confirm or deny the following...
Quick question: Do the French dislike outsiders has much as they are rumored to? If it is anywhere near as bad as the rumors, I would consider that a considerable roadblock. While the US does have its share of racism, it also has a greater mixture of race, religion, and culture than any other country in the world.
Yes,/. is a way of getting involved. But judging from the number of young people who actually vote, there is lot of posting going on but not enough doing. Look at the elderly. They get the vote out in much better percentages. This difference is even greater when you factor in that today's young will have more years to have to live with a bad government and the young tend to be more able bodied than the elderly.
I have nothing against the elderly but as a young person, I want my demographic's view represented as well. Look at social security; baby boomers are rapidly hitting retirement age. That means one of the biggest voting demographics will be pushing for social security benefits to increase or at the the very least stay the same. Where will this money come from? Our paychecks, that's where.
Canidates are afraid of upsetting the old and the rich, but they have no problem with the poor and the young. Why? Because, we don't pay for compaigns and not too many of us vote.
Talk is not enough. And violent, crazy talk is worse. Working within the system can work but only in large numbers. This is where the vocal involvement needs to become real through voting, letter writing, etc. I say "chill out" because violence and angry talk is not the answer. It is these kind of actions that makes the powerful not take us seriously. We to act seriously in order to be taken seriously.
So keep searching for the truth, keep spreading the truth but take it to the next level too!
Ok, keep the government playing by the rules to avoid erosion of basic freedoms. I agree with you there. I just wonder if some people's "militia" like attitudes are the best way to fix it? Perhaps it can still be fixed from the inside. Then again, the next paragraph makes you wonder if anything can be done with some shift-4....
Big corporations and the government using the current prosperity to erect a police state? Hmm... I don't see big corporations and the government having the same agenda, although they are similarities. Big corporations want power to make money while goverment (well, politicions) want money to take power. This brings up the BIG need for campaign finance reform and a deep look into limiting lobbying. This pres. campaign is the most expensive in history by far. The rich (people and corporations) are giving millions of dollors to both canidates to exert just a little bit of influence and maybe sleep over at the white house. Its sad that this office seems to be up for sale.
As a benificiary of the current trend of presperity, I hope that I have not been lulled. Then again, I know I can rely on good ol'/. to show me another view point and get me thinking again.
I agree that there may be places with more freedom than the US. There also may be places with more oppertunity for success. But where is there a place with a good combination of the two?
Every place has its good points and its bad points but on those two issues, I think altough they still need improvement, the US does pretty good on both.
It seems that some people are pretty angry about this. Perhaps you have listened to too much Rage Against the Machine (as I am now)? Those guys can really get you going!:-)
Seriously though, what has the government done to each of you specifically that you are so angry about? I am not a Canivore lover either but I am also not ready to burn the White House.
The way I see it, you have two choices:
1. Work within the sytem to make things better. Vote. Write letters. Write e-mails. Get involved.
2. Move somewhere better. Good luck finding a place with as much oppertunity for success and freedom. When you get there, send us a post so that we can visit you.
Most/. readers are involved with technology and those of us who are out of college know that we all do pretty well economically. Guess who is partly responsible? Yep, the government. While many of us may not remember (or may not want to remember it) there was a time 20-30 years ago when we had double digit inflation and a much larger jobless rate.
My point is that while things are far from perfect, they are much better than they are else where and much better than they were here. We are in the midst of one of the longest streaks of prosperity ever. Lets use this time to keep making things better with constructive ideas.
I don't know about other programmers experiences but when I wrote Alan, I got a response. I wrote him not too long ago with some questions about the Linux I2O drivers he was working on (I was working on some myself for my company's Fibre Channel adapter). I don't think that any of my e-mails took more than a day to get a response.. heck, one was answered in one hour! These were not short, "Don't bother me." relies either; they gave complete answers to my questions.
I have also written Linus about porting to some new hardware my company was developing. Both times I got a full response in under three hours!
While these are just my stories I would find it hard to believe these guys are treating my e-mails as special. So if you have a question or bug fix, let them know. They may not be able to enter the code in write then but you should get a response pretty soon.
Well can the PCI bus handle 480 Mb? Yes. 32-bit 33 Mhz PCI goes up to 132 MB/second, more than enough. 64-bit, 66 Mhz hits 528 MB which blows this away. PCI isn't dead yet.
The current Fibre Channel standard supports up to 542.72 megabytes per second or 4.24 gigabits per second. Single controllers are already going 190-200 MB/sec. Meanwhile gigabit eithernet is only moving at 128 MB/sec (some go a bit faster). They are using Fibre Channel for SANs with multiple RAID arrays connected. Needless to say, it flies! Pretty soon files will be loaded before you even open them!:-)
64-bit 66 Mhz PCI can go 528 megabytes per second. This baby does 480 megabits per second or 120 megabytes per second. Even plain old 32-bit 33 Mhz PCI does 132 megabytes per second. The current Fibre Channel standard supports up to 4.24 gigabits or 542.72 megabytes per second. Single FC adapters are already pushing 190 MB and getting faster. After working with this kind of stuff, I have to admit that USB seems downright slow. Now if only FC would get cheaper...
How was this the haves vs. the have nots? It seemed to be the violent protesters vs. the police. Just because you do not have does not If a guy pushes a police officer off his bike give you the right to destroy public and private property. People too lazy to vote are quick to flip dumpsters in the street. Have nots? These people are well funded. They seem to have.
As for the police being heavy handed...how so? should he sit on the ground and take it? When they press against mounted police spooking horses and thus threatening the police, the horse, and the people nearby should they do nothing? When they destroy property should they do nothing?
You seem to forget theat there were days of peacefull protests that were held without incident.
What was their cause? It seemed to be just to disrupt society and the political process. The system has all sorts of ways to create change. Violence is not they way.
Finally, China (the government) could care less what the US does within its borders. They only car about our foreign policy. Arresting violent protesters is the RIGHT thing to do.
The people that were held for 48 hours without a phone call were not giving their names or otherwise not allowing the police to process them. You have to give to get. If you do not cooporate with the police, why should they cooporate with you? When a cop pulls me over, I might not be happy but I cooporate.
Also, the majority of people that were arrested where violent. Their were no problems until the protesters started getting violent. They started destroying public property. That is wrong. I live in the philadelphia area and I don't appreciate people destroying my city to make a point. How would they like it if I destroyed thier towns? How would you like it?
I went to Seattle shortly after the WTO protests. I talked to the shop owners and small business employees that were crippled with huge damages and lose of business. For what? Because their city hosted the WTO talks? That is just plain wrong.
Comparing this to China is crazy. Peacefull, non-violent protests were not touched by the police. When someone infringes on others rights, destroys property, and endangers other citizens, the police protect those citizens. How does this compare?
Did you even study this before you wrote this? Do you advocate violence as a means of change? I hope this group comes to your town to protest and we will see how you feel.
I was under the impression that you did in fact have to recode to make a process run accross clusters simultaneously. Also, nt4/Win2k sever only supports two nodes, Win2 advanced server supports 4, Win2k data center server may support 8 (when Win2kDCS finally gets out). For the price of Win2k, you would expect much higher levels of clustering. Also, check the cost. Win2kDCS costs more without support than RedHat's offer with support. Next check the machines. One of the nice points of clustering is avoiding large super expensive boxes. Win2kDCS has much higher minimum requirements than other Win2k/nt4 versions. While Linux lets you decide what to run things on. Scalability goes in both directions.
I am bit confused about your WDM ideas. All of Win2k's drivers are WDM. Miniports use a WDM compliant port driver with some new miniport calls. Using "native" nt drivers are what Microsoft calls legacy drivers. The presense of one of these drivers breaks power management on the machine because the legacy driver can not answer the request to change system power state. If the hal (much smaller in win2k), pci, disk, class, scsiport, ndis, etc. are all WDM compliant. WDM is the native model for Win2k. The comparison to Win32 is not correct. While a Win32 call is really just a wrapper for the native nt call, WDM is usually not a wrapper for legacy calls. It is often the opposite; most legacy functions are just macros that call the new WDM functions. The main points of WDM on Win2k are cross platform design (and binary compatibility in many cases) and bringing power managment and plug and play to nt. PM and Pnp are considered very important at MS; they are not just political marketing ideas. Although this new model helps the new developer making products for 9x and win2k, it hurts current nt developers by forcing them to rewrite their drivers. In many ways, WDM actually slowed the developement of drivers for Win2k. Finding classes on a moving spec was a difficult task. My main issue where Win2k is not tuned for performance involve the use of so many general drivers to handle whole classes of devices but none of them well. IMHO MS would have been better off supporting I2O like almost every other OS does. In order to get the best performance, any storage developer will tell you that you have to write a full port drivers to replace scsiport/miniport and disk/class. NDIS is similar. In fact, it is worse because MS will not certify a non-NDIS network driver. Here, the developer is forced into a slow model. NDIS 5 addressed some issues (removal of some locks, some off load support) but left many others. It is this push towards a miniport model that I find hurts performance. Also, the messaging scheme for drivers requires system calls between each layer instead of using a direct function pointer interface. Almost all messages must go all the way through the stack, even when most drivers just blindly pass the call down. While this allows any driver or filter on the stack to change things, it slows performance. Many messages must be handled "on the way up" the driver stack. Whole stacks have to wait on events to be triggered and callbacks called. Microsoft gets extendability through the support of upper and lower filter drivers at the expense of performance. A faster model would reduce the layers in the device stack, use direct function calls rather than system calls for message passing and reduce the need for callbacks and event waiting. As for reading NT 3.0 documentation...well that's great for discussing NT 3.0. Win2k is their current model and it has changed in ways that are more than just wrapper functions and cosmetic changes.
NT "runs all its servers in kernel space"? Do you mean drivers? Services? Services are not run in kernel space, althougth they can be set with a high priority. All drivers are run in kernel space, though. "The kernel has a lot of design concessions that faccilitate a really high I/O rate." Really? Have you looked at the code for network and storage? I have written network and storage drivers for NT4/Win2k and is not designed to be fast. Check out the DDK. Both storage and network use a miniport model (SCSI Miniport and NDIS Miniport) with a port driver doing much of the work. To make matters worse, Win2k use WDM for its drivers. WDM tends to add an additional driver object to the layered model. Both miniport and WDM are designed to be very general and take control away from the driver developer. A call to read a few bytes from the disk goes through so many layers. First, the file system drivers, then class.sys, then disk.sys, then scsiport.sys, then vendorscsiminiport.sys, then hardware. There can also be any number of filter drivers in the mix. WDM allows upper and lower filters for each FDO (Functional Device Object). We got a nice performance boost by not using the SCSIMiniport/Class driver interface. Win2k is not designed to be fast as much as extendable and general. Just my $.02...
No need for virus scanners...just don't run things you don't know. What is it with people's strange need to double click on every single thing that appears in front of them?!! This kind of stuff drives me nuts. You wouldn't talk to strangers but you will open up strange binaries from them? Ug...
Here is the real question though. Do you really think that it costs that much more to make a great car as they charge for it? If they made the profit margins for a Ferrari the same as a Toyota, they would cost a lot less. The same could be said for storage technology and most technology in general. The fact is FC and SCSI have higher profit margins than IDE. RAID cabinets have even higher margins. In this case, what is true for hardware can be true for software. Take Windows 9x vs. NT/2k. Microsoft will tell you that NT Workstation/Win2k Pro is 8X more stable than Win9x. Then why do they even make 9x? They have admitted that it is not stable. Basically, they fixed the bugs in 9x and made it a new product that costs much more. Thanks but no thanks. One of the things that I love about Linux is that there aren't cheap buggy versions and expensive good versions; there is just one inexpensive, great version. When something is fixed it is fixed for all. It would be more ethical if companies just made one good version of a type of product rather than staged levels of quality and cost. It would be better for the people. Sadly, it does not make good business sense and therefore will probably never happen.
RAID comes in three forms: Software, Controller, and External. Software RAID is done by the OS/kernel/file system driver. It can give you the protection of RAID but much of the performance of multiple drives may be lost in host CPU running the RAID. Controller RAID is done by a add-in card usually PCI, usually SCSI or Fibre Channel. There are IDE RAID card but with only four devices and IDE performance...why make the effort? I worked on a Fibre Channel RAID card. Hook this card up to a rack of hard disks and the OS only knew about the virtual RAID disks we presented. We defaulted to RAID 10 by the by. External RAID is the kind done by companies like EMC. Here, a massive cabinet of drives have a target mode controller(s) built in. Plug in a regular controller into the cabinet and the controller only sees the virtual RAID disks. Many of the ideas that you are talking about are already used in RAID today. The placement of the data is important to how fast it can be accessed. Another big issue is cache. Many RAID controllers have 16-64 MB of cache. The often use both read and write cache. Now, to put all of this in small inexpensive unit would be interesting. Some EMC boxes are the 6 ' X 2' X 2' and cost $500,000. I personally think that a good RAID FC controller running RAID 10 connected to Seagate Cheetah drives are your best bet.
I have worked on a Fibre Channel RAID card and I find this whole ATA vs. SCSI discussion strange. It is obvious that SCSI is better. More devices, better throughput. Fibre Channel takes it farther with even more devices and even better throughput. The card I worked on did 190 MB/sec, 23,000 I/Os per second in direct connect mode. It did even better with RAID. 32-bit PCI? I assume you mean 33 Mhz as well. This is worthless in the server market. A good server should have multiple 64-bit 66 Mhz buses. Each bus has a max of 528 MB/sec. Two of our cards can run at max speed with a very high number of drives on it. I have run with over 300 drives on one card in direct connect mode. 32-bit PCI slots should be at least 66 Mhz and should only hang around for legacy stuff. PCI-X and Infiniband are coming soon. In the real world of enterprise I/O, ATA and frankly SCSI, are a joke. 15 drives plus a controller? Please. Never mind the four drives that IDE gives you. As for FireWire and USB...well when I see it in the storage market, I will believe it. Fibre Channel is the king of enterprise storage. SCSI will do it for lower level servers and ATA/IDE will do it for small systems like PCs. It is a step in the right direction for ATA drivers and hardware to support true busmaster DMA. Many IDE type devices eat up CPU by not supporting DMA or only supporting slave DMA. Yuck. Still, if SCSI is better than ATA and FC is better than SCSI, why even make anything other than FC? The answer: Money! I can't stand storage manufactures selling various levels of technology so that they can make a massive profit on their high end stuff. It is no different then Intel releasing a fast totally expensive chip and then a slower but cheaper chip. Does the fast chip cost more to make? Nope. Does FC cost that much more than SCSI? Does SCSI cost that much more than ATA/IDE? This is why good technology takes so long to get into the people's hands.
"This system does run regular software without recompiling. It just doesn't use a lot of CPUs for simultaneous compute unless you change the code to use MPI. But they can access the shared storage at high speeds without any change, and they can get farmed out to separate CPUs without any change."
This is goes to my main point, though. I want a clustering solution that will take full advantage of the processor power available in a transparent way, like SMP does on a traditional multi-proc machine. Imagine being able to just plug in a unit for more processor power. Imagine being able to do it while it is running. Imagine that and you get my idea. I am not saying that PCI/PCI-X/Infiniband is the answer but I think that transparent solutions are often the best solutions. For another example, think about how external RAID systems work. They present the OS with virtual disks that may be composed of many physical disks. The OS does not need to know about RAID at all. It just has a disk that is big and works fast. Below the surface, physical disks (like cluster nodes) can die without the virtual disk dying. I think that this type of solution is nearly perfect. You want speed and failure handling for your disks? Use RAID. You want speed and failure handling for your computer? Someone needs to make a clustering solution like what I am talking about. We need a clustering solution that works without the user needing to know anything about clustering. It should just work.
Many clusters seem to run into limits from the network and storage. Is there anyone playing with some sort of PCI Bridge that would let hardware in one box see hardware in another and communicate through peer to peer PCI? This could avoid a lot of the network and storage waits. Nodes could access each other's memory and hardware as if they were part of of their own. The address mapping would be tricky but other big multi-processor systems seem to have solved this. I guess I am thinking of some middle ground between traditional multi-processor systems and common clusters. I want something that operates like a multi-proc box but uses common hardware, with perhaps one special PCI card/bridge. I want a cluster that can run regular software without special compiling.
"The video core and CPU share the same memory." Color me still unimpressed. A standard PCI adapter can access sytem memory. A standard NT/2k kernel driver can allocate non-paged memory for their devices to use. DMA works this way. Now if the video card does not have to go through any kind of bus to get to system memory, that is an improvement. This isn't a sharing memory issue as much as a bus architecture issue. If they have built a way for the video processor to get to system memory that is faster than PCI or AGP than it is somewhat non-PC. You have to wonder why you wouldn't use it in a PC, though. You also have to wonder what this innovation will cost. A non-standard super video mother board will be more expesive than a regular one. How can MS make this cheap without an MSN based rebate? Or are they going to byte (pardon the pun) the cost and earn it back on games? "I believe the PSX2 is going to retail for $349. I suppose it's not going to sell." Hmm, you got me. I was off by a whole $49. How long do you think that this thing will stay at $349? Standard rule is to always start higher than you are willing to sell it for.
Yes, but all of those distros are based on the same base. Each of MS's OSs only share an API. The outside world can get in the same way but the implementation is different. There are various calls that are the same in WinMe, CE, and 2000 but they are all implemented differently. These differences lead to unpredictability and unreliability. On the Linux side, most distros try to be share as much common kernel code as possible. They want to reduce the piece of the kernel base that they have to maintain. My point was that I am suspicious of yet another MS operating system. I would not be suspicious of a special distro of Win2k, WinCE, or WinMe. I understand the need to change things for a game console in ways that you would not want in a full pc environment. I don't understand the need to start from scratch. If Win2k is better than the 9x kernel (which WinMe is also based on) than what is the point of WinMe? Price? To argue that the prices of WinMe and Win2k are based on the cost to produce them is ridiculous. Supply and demand? Based on that the prices should be reversed. Quality? Ok, but is it ethical to release an inferior product when you have something that works better? This isn't a case of an econo car vs. a luxury car. This is a case of a car that breaks down all the time vs. one that works. MS says that customers should upgrade from Win9x to Win2k Pro because it is more reliable. In other words, your mechanic just told you he can't fix your car but he can sell you one that doesn't have the problem. ME has done this with many of their products. Got a problem with Excel 5? Buy Excel 7 (95). Oops...Excel 7 is broke? Try 97! Still bugs? This time Excel 2k has got them all fixed. No, really. Honest. Oh yeah it is more expensive but worth every penny!
Slashdot Mirror
"The reason for this is to give a visual hint as to what is a comment and what is actual code. The rest of the syntax highlighting is less important in this case. While porting, never delete a line of code. Comment it out and make your changes below it. If you need to, comment out whole functions. You are going to find that having an immediate reference to the original code is immensely helpful, but all that text can get very confusing. Syntax highlighting makes everything much easier. "
/* */ blocks.
I couldn't disagree more. Large amounts of commented out code make programs unreadable. Why was it commented out? Was it test code? An older version? Something that wouldn't compile but needs to be fixed? Who knows...often not even the orignal commenter after a few months. Don't think that you will go back in fix it because you will not. There is almost never time to fix badly written code that seems to work because there is always new code to write. You will net get time to look at it again until a bug pops up. And then you will be left wondering what in the world was in that commented code. Writing things the correct way the first time makes life much easier.
Comments should be used for comments. Some text explaining code. They should be used for turning code on and off (use precompiler statments, much more flexible) and they should not be used to save old code. Try using a source repository that has version diff support like cvs. This is a much cleaner way to see what you had before without leaving your old code lying around in massive
As for comment syntax highlighting being the only usefull part of syntax highlighting, I have to disagree again. I like to see different parts of the lanugages syntax in different colors. Hence the name "syntax" highlighting, not comment highlighting.
What version of macro economics did you study? Oh... the one where government has no effect on the economy. So you feel that the congress' budget, the FED's policies, the President's domestic and foreign policies, and the court's enforcement of financial and criminal law have nothing to do with our economic prosperity?
Ok, then what in the world does?
Ok, given these crimes, what are you going to do about it? How are you going to get involved in order to make things better? If all the things you mentioned are really happenening and you do nothing, then aren't you really just complaining? Seriously, if the FBI is breaking the law, then you need to let the people that write their checks know. Write to the government. Get lots of people to sign a petition. You know, get really involved with the democratic process.
Last I checked it was our country, not just yours. This means I have just as much a right to voice my opinion as you. If you can not handle a reading a different point of view without resorting to foul name calling then it sounds like you need to go to a place where there are other people that don't like to hear different opinions. I think that China fits that description.
By "chilling out", I did not mean do nothing, I just meant that a person should act rationally. Learn the system, understand the system, and then you can fix the system.
Hello new Frenchie, ex-American! :-) Then I could confirm or deny the following...
IMHO, high unemployment is a sign of a lack of overall oppertunity that the US offers. Most people who say that they can not get a job in the US really mean that they can not get a job that they like and want. It does sound like France may have oppertunities in the tech. world but then again, the same can be said for several other countries. But what about a chance of building a successful business for yourself? Moving up in a non-technical field? The US has these oppertunities in large quantities.
As for a transfer, my company is partially owned by a Swiss company and has several Eurpean offices. Maybe I will stop by and see how things are!
Quick question: Do the French dislike outsiders has much as they are rumored to? If it is anywhere near as bad as the rumors, I would consider that a considerable roadblock. While the US does have its share of racism, it also has a greater mixture of race, religion, and culture than any other country in the world.
Yes, /. is a way of getting involved. But judging from the number of young people who actually vote, there is lot of posting going on but not enough doing. Look at the elderly. They get the vote out in much better percentages. This difference is even greater when you factor in that today's young will have more years to have to live with a bad government and the young tend to be more able bodied than the elderly.
I have nothing against the elderly but as a young person, I want my demographic's view represented as well. Look at social security; baby boomers are rapidly hitting retirement age. That means one of the biggest voting demographics will be pushing for social security benefits to increase or at the the very least stay the same. Where will this money come from? Our paychecks, that's where.
Canidates are afraid of upsetting the old and the rich, but they have no problem with the poor and the young. Why? Because, we don't pay for compaigns and not too many of us vote.
Talk is not enough. And violent, crazy talk is worse. Working within the system can work but only in large numbers. This is where the vocal involvement needs to become real through voting, letter writing, etc. I say "chill out" because violence and angry talk is not the answer. It is these kind of actions that makes the powerful not take us seriously. We to act seriously in order to be taken seriously.
So keep searching for the truth, keep spreading the truth but take it to the next level too!
Ok, keep the government playing by the rules to avoid erosion of basic freedoms. I agree with you there. I just wonder if some people's "militia" like attitudes are the best way to fix it? Perhaps it can still be fixed from the inside. Then again, the next paragraph makes you wonder if anything can be done with some shift-4.... /. to show me another view point and get me thinking again.
Big corporations and the government using the current prosperity to erect a police state? Hmm... I don't see big corporations and the government having the same agenda, although they are similarities. Big corporations want power to make money while goverment (well, politicions) want money to take power. This brings up the BIG need for campaign finance reform and a deep look into limiting lobbying. This pres. campaign is the most expensive in history by far. The rich (people and corporations) are giving millions of dollors to both canidates to exert just a little bit of influence and maybe sleep over at the white house. Its sad that this office seems to be up for sale.
As a benificiary of the current trend of presperity, I hope that I have not been lulled. Then again, I know I can rely on good ol'
I agree that there may be places with more freedom than the US. There also may be places with more oppertunity for success. But where is there a place with a good combination of the two?
Every place has its good points and its bad points but on those two issues, I think altough they still need improvement, the US does pretty good on both.
It seems that some people are pretty angry about this. Perhaps you have listened to too much Rage Against the Machine (as I am now)? Those guys can really get you going! :-)
/. readers are involved with technology and those of us who are out of college know that we all do pretty well economically. Guess who is partly responsible? Yep, the government. While many of us may not remember (or may not want to remember it) there was a time 20-30 years ago when we had double digit inflation and a much larger jobless rate.
Seriously though, what has the government done to each of you specifically that you are so angry about? I am not a Canivore lover either but I am also not ready to burn the White House.
The way I see it, you have two choices:
1. Work within the sytem to make things better. Vote. Write letters. Write e-mails. Get involved.
2. Move somewhere better. Good luck finding a place with as much oppertunity for success and freedom. When you get there, send us a post so that we can visit you.
Most
My point is that while things are far from perfect, they are much better than they are else where and much better than they were here. We are in the midst of one of the longest streaks of prosperity ever. Lets use this time to keep making things better with constructive ideas.
I don't know about other programmers experiences but when I wrote Alan, I got a response. I wrote him not too long ago with some questions about the Linux I2O drivers he was working on (I was working on some myself for my company's Fibre Channel adapter). I don't think that any of my e-mails took more than a day to get a response.. heck, one was answered in one hour! These were not short, "Don't bother me." relies either; they gave complete answers to my questions.
I have also written Linus about porting to some new hardware my company was developing. Both times I got a full response in under three hours!
While these are just my stories I would find it hard to believe these guys are treating my e-mails as special. So if you have a question or bug fix, let them know. They may not be able to enter the code in write then but you should get a response pretty soon.
Well can the PCI bus handle 480 Mb? Yes. 32-bit 33 Mhz PCI goes up to 132 MB/second, more than enough. 64-bit, 66 Mhz hits 528 MB which blows this away. PCI isn't dead yet.
The current Fibre Channel standard supports up to 542.72 megabytes per second or 4.24 gigabits per second. Single controllers are already going 190-200 MB/sec. Meanwhile gigabit eithernet is only moving at 128 MB/sec (some go a bit faster). They are using Fibre Channel for SANs with multiple RAID arrays connected. Needless to say, it flies! Pretty soon files will be loaded before you even open them! :-)
64-bit 66 Mhz PCI can go 528 megabytes per second. This baby does 480 megabits per second or 120 megabytes per second. Even plain old 32-bit 33 Mhz PCI does 132 megabytes per second. The current Fibre Channel standard supports up to 4.24 gigabits or 542.72 megabytes per second. Single FC adapters are already pushing 190 MB and getting faster. After working with this kind of stuff, I have to admit that USB seems downright slow. Now if only FC would get cheaper...
How was this the haves vs. the have nots? It seemed to be the violent protesters vs. the police. Just because you do not have does not If a guy pushes a police officer off his bike give you the right to destroy public and private property. People too lazy to vote are quick to flip dumpsters in the street. Have nots? These people are well funded. They seem to have.
As for the police being heavy handed...how so? should he sit on the ground and take it? When they press against mounted police spooking horses and thus threatening the police, the horse, and the people nearby should they do nothing? When they destroy property should they do nothing?
You seem to forget theat there were days of peacefull protests that were held without incident.
What was their cause? It seemed to be just to disrupt society and the political process. The system has all sorts of ways to create change. Violence is not they way.
Finally, China (the government) could care less what the US does within its borders. They only car about our foreign policy. Arresting violent protesters is the RIGHT thing to do.
The people that were held for 48 hours without a phone call were not giving their names or otherwise not allowing the police to process them. You have to give to get. If you do not cooporate with the police, why should they cooporate with you? When a cop pulls me over, I might not be happy but I cooporate.
Also, the majority of people that were arrested where violent. Their were no problems until the protesters started getting violent. They started destroying public property. That is wrong. I live in the philadelphia area and I don't appreciate people destroying my city to make a point. How would they like it if I destroyed thier towns? How would you like it?
I went to Seattle shortly after the WTO protests. I talked to the shop owners and small business employees that were crippled with huge damages and lose of business. For what? Because their city hosted the WTO talks? That is just plain wrong.
Comparing this to China is crazy. Peacefull, non-violent protests were not touched by the police. When someone infringes on others rights, destroys property, and endangers other citizens, the police protect those citizens. How does this compare?
Did you even study this before you wrote this? Do you advocate violence as a means of change? I hope this group comes to your town to protest and we will see how you feel.
I was under the impression that you did in fact have to recode to make a process run accross clusters simultaneously.
Also, nt4/Win2k sever only supports two nodes, Win2 advanced server supports 4, Win2k data center server may support 8 (when Win2kDCS finally gets out). For the price of Win2k, you would expect much higher levels of clustering.
Also, check the cost. Win2kDCS costs more without support than RedHat's offer with support. Next check the machines. One of the nice points of clustering is avoiding large super expensive boxes. Win2kDCS has much higher minimum requirements than other Win2k/nt4 versions. While Linux lets you decide what to run things on. Scalability goes in both directions.
I am bit confused about your WDM ideas. All of Win2k's drivers are WDM. Miniports use a WDM compliant port driver with some new miniport calls. Using "native" nt drivers are what Microsoft calls legacy drivers. The presense of one of these drivers breaks power management on the machine because the legacy driver can not answer the request to change system power state.
If the hal (much smaller in win2k), pci, disk, class, scsiport, ndis, etc. are all WDM compliant. WDM is the native model for Win2k. The comparison to Win32 is not correct. While a Win32 call is really just a wrapper for the native nt call, WDM is usually not a wrapper for legacy calls. It is often the opposite; most legacy functions are just macros that call the new WDM functions.
The main points of WDM on Win2k are cross platform design (and binary compatibility in many cases) and bringing power managment and plug and play to nt. PM and Pnp are considered very important at MS; they are not just political marketing ideas. Although this new model helps the new developer making products for 9x and win2k, it hurts current nt developers by forcing them to rewrite their drivers. In many ways, WDM actually slowed the developement of drivers for Win2k. Finding classes on a moving spec was a difficult task.
My main issue where Win2k is not tuned for performance involve the use of so many general drivers to handle whole classes of devices but none of them well. IMHO MS would have been better off supporting I2O like almost every other OS does. In order to get the best performance, any storage developer will tell you that you have to write a full port drivers to replace scsiport/miniport and disk/class.
NDIS is similar. In fact, it is worse because MS will not certify a non-NDIS network driver. Here, the developer is forced into a slow model. NDIS 5 addressed some issues (removal of some locks, some off load support) but left many others.
It is this push towards a miniport model that I find hurts performance. Also, the messaging scheme for drivers requires system calls between each layer instead of using a direct function pointer interface. Almost all messages must go all the way through the stack, even when most drivers just blindly pass the call down. While this allows any driver or filter on the stack to change things, it slows performance. Many messages must be handled "on the way up" the driver stack. Whole stacks have to wait on events to be triggered and callbacks called. Microsoft gets extendability through the support of upper and lower filter drivers at the expense of performance.
A faster model would reduce the layers in the device stack, use direct function calls rather than system calls for message passing and reduce the need for callbacks and event waiting.
As for reading NT 3.0 documentation...well that's great for discussing NT 3.0. Win2k is their current model and it has changed in ways that are more than just wrapper functions and cosmetic changes.
NT "runs all its servers in kernel space"? Do you mean drivers? Services? Services are not run in kernel space, althougth they can be set with a high priority. All drivers are run in kernel space, though.
"The kernel has a lot of design concessions that faccilitate a really high I/O rate." Really? Have you looked at the code for network and storage? I have written network and storage drivers for NT4/Win2k and is not designed to be fast. Check out the DDK. Both storage and network use a miniport model (SCSI Miniport and NDIS Miniport) with a port driver doing much of the work. To make matters worse, Win2k use WDM for its drivers. WDM tends to add an additional driver object to the layered model. Both miniport and WDM are designed to be very general and take control away from the driver developer. A call to read a few bytes from the disk goes through so many layers. First, the file system drivers, then class.sys, then disk.sys, then scsiport.sys, then vendorscsiminiport.sys, then hardware. There can also be any number of filter drivers in the mix. WDM allows upper and lower filters for each FDO (Functional Device Object). We got a nice performance boost by not using the SCSIMiniport/Class driver interface. Win2k is not designed to be fast as much as extendable and general.
Just my $.02...
No need for virus scanners...just don't run things you don't know. What is it with people's strange need to double click on every single thing that appears in front of them?!!
This kind of stuff drives me nuts. You wouldn't talk to strangers but you will open up strange binaries from them? Ug...
Here is the real question though. Do you really think that it costs that much more to make a great car as they charge for it? If they made the profit margins for a Ferrari the same as a Toyota, they would cost a lot less.
The same could be said for storage technology and most technology in general. The fact is FC and SCSI have higher profit margins than IDE. RAID cabinets have even higher margins. In this case, what is true for hardware can be true for software. Take Windows 9x vs. NT/2k. Microsoft will tell you that NT Workstation/Win2k Pro is 8X more stable than Win9x. Then why do they even make 9x? They have admitted that it is not stable. Basically, they fixed the bugs in 9x and made it a new product that costs much more. Thanks but no thanks.
One of the things that I love about Linux is that there aren't cheap buggy versions and expensive good versions; there is just one inexpensive, great version. When something is fixed it is fixed for all.
It would be more ethical if companies just made one good version of a type of product rather than staged levels of quality and cost. It would be better for the people. Sadly, it does not make good business sense and therefore will probably never happen.
RAID comes in three forms: Software, Controller, and External.
Software RAID is done by the OS/kernel/file system driver. It can give you the protection of RAID but much of the performance of multiple drives may be lost in host CPU running the RAID.
Controller RAID is done by a add-in card usually PCI, usually SCSI or Fibre Channel. There are IDE RAID card but with only four devices and IDE performance...why make the effort? I worked on a Fibre Channel RAID card. Hook this card up to a rack of hard disks and the OS only knew about the virtual RAID disks we presented. We defaulted to RAID 10 by the by.
External RAID is the kind done by companies like EMC. Here, a massive cabinet of drives have a target mode controller(s) built in. Plug in a regular controller into the cabinet and the controller only sees the virtual RAID disks.
Many of the ideas that you are talking about are already used in RAID today. The placement of the data is important to how fast it can be accessed. Another big issue is cache. Many RAID controllers have 16-64 MB of cache. The often use both read and write cache.
Now, to put all of this in small inexpensive unit would be interesting. Some EMC boxes are the 6 ' X 2' X 2' and cost $500,000. I personally think that a good RAID FC controller running RAID 10 connected to Seagate Cheetah drives are your best bet.
I have worked on a Fibre Channel RAID card and I find this whole ATA vs. SCSI discussion strange. It is obvious that SCSI is better. More devices, better throughput. Fibre Channel takes it farther with even more devices and even better throughput.
The card I worked on did 190 MB/sec, 23,000 I/Os per second in direct connect mode. It did even better with RAID.
32-bit PCI? I assume you mean 33 Mhz as well. This is worthless in the server market. A good server should have multiple 64-bit 66 Mhz buses. Each bus has a max of 528 MB/sec. Two of our cards can run at max speed with a very high number of drives on it. I have run with over 300 drives on one card in direct connect mode. 32-bit PCI slots should be at least 66 Mhz and should only hang around for legacy stuff. PCI-X and Infiniband are coming soon.
In the real world of enterprise I/O, ATA and frankly SCSI, are a joke. 15 drives plus a controller? Please. Never mind the four drives that IDE gives you.
As for FireWire and USB...well when I see it in the storage market, I will believe it. Fibre Channel is the king of enterprise storage. SCSI will do it for lower level servers and ATA/IDE will do it for small systems like PCs. It is a step in the right direction for ATA drivers and hardware to support true busmaster DMA. Many IDE type devices eat up CPU by not supporting DMA or only supporting slave DMA. Yuck. Still, if SCSI is better than ATA and FC is better than SCSI, why even make anything other than FC? The answer: Money!
I can't stand storage manufactures selling various levels of technology so that they can make a massive profit on their high end stuff. It is no different then Intel releasing a fast totally expensive chip and then a slower but cheaper chip. Does the fast chip cost more to make? Nope. Does FC cost that much more than SCSI? Does SCSI cost that much more than ATA/IDE? This is why good technology takes so long to get into the people's hands.
"This system does run regular software without recompiling. It just doesn't use a lot of CPUs for simultaneous compute unless you change the code to use MPI. But they can access the shared storage at high speeds without any change, and they can get farmed out to separate CPUs without any change."
This is goes to my main point, though. I want a clustering solution that will take full advantage of the processor power available in a transparent way, like SMP does on a traditional multi-proc machine. Imagine being able to just plug in a unit for more processor power. Imagine being able to do it while it is running. Imagine that and you get my idea.
I am not saying that PCI/PCI-X/Infiniband is the answer but I think that transparent solutions are often the best solutions.
For another example, think about how external RAID systems work. They present the OS with virtual disks that may be composed of many physical disks. The OS does not need to know about RAID at all. It just has a disk that is big and works fast. Below the surface, physical disks (like cluster nodes) can die without the virtual disk dying. I think that this type of solution is nearly perfect. You want speed and failure handling for your disks? Use RAID. You want speed and failure handling for your computer? Someone needs to make a clustering solution like what I am talking about. We need a clustering solution that works without the user needing to know anything about clustering. It should just work.
Many clusters seem to run into limits from the network and storage. Is there anyone playing with some sort of PCI Bridge that would let hardware in one box see hardware in another and communicate through peer to peer PCI? This could avoid a lot of the network and storage waits. Nodes could access each other's memory and hardware as if they were part of of their own. The address mapping would be tricky but other big multi-processor systems seem to have solved this.
I guess I am thinking of some middle ground between traditional multi-processor systems and common clusters. I want something that operates like a multi-proc box but uses common hardware, with perhaps one special PCI card/bridge. I want a cluster that can run regular software without special compiling.
"The video core and CPU share the same memory."
Color me still unimpressed. A standard PCI adapter can access sytem memory. A standard NT/2k kernel driver can allocate non-paged memory for their devices to use. DMA works this way.
Now if the video card does not have to go through any kind of bus to get to system memory, that is an improvement. This isn't a sharing memory issue as much as a bus architecture issue. If they have built a way for the video processor to get to system memory that is faster than PCI or AGP than it is somewhat non-PC. You have to wonder why you wouldn't use it in a PC, though. You also have to wonder what this innovation will cost. A non-standard super video mother board will be more expesive than a regular one. How can MS make this cheap without an MSN based rebate? Or are they going to byte (pardon the pun) the cost and earn it back on games?
"I believe the PSX2 is going to retail for $349. I suppose it's not going to sell."
Hmm, you got me. I was off by a whole $49. How long do you think that this thing will stay at $349? Standard rule is to always start higher than you are willing to sell it for.
Yes, but all of those distros are based on the same base. Each of MS's OSs only share an API. The outside world can get in the same way but the implementation is different.
There are various calls that are the same in WinMe, CE, and 2000 but they are all implemented differently. These differences lead to unpredictability and unreliability.
On the Linux side, most distros try to be share as much common kernel code as possible. They want to reduce the piece of the kernel base that they have to maintain.
My point was that I am suspicious of yet another MS operating system. I would not be suspicious of a special distro of Win2k, WinCE, or WinMe. I understand the need to change things for a game console in ways that you would not want in a full pc environment. I don't understand the need to start from scratch.
If Win2k is better than the 9x kernel (which WinMe is also based on) than what is the point of WinMe? Price? To argue that the prices of WinMe and Win2k are based on the cost to produce them is ridiculous. Supply and demand? Based on that the prices should be reversed. Quality? Ok, but is it ethical to release an inferior product when you have something that works better? This isn't a case of an econo car vs. a luxury car. This is a case of a car that breaks down all the time vs. one that works. MS says that customers should upgrade from Win9x to Win2k Pro because it is more reliable. In other words, your mechanic just told you he can't fix your car but he can sell you one that doesn't have the problem.
ME has done this with many of their products. Got a problem with Excel 5? Buy Excel 7 (95). Oops...Excel 7 is broke? Try 97! Still bugs? This time Excel 2k has got them all fixed. No, really. Honest. Oh yeah it is more expensive but worth every penny!