I like the Hurd's principal, but all the good ideas die because they aren't mainstream as quick as the mediocre ones. It's quite hard to edge out the market once a standard has been set. (Evidence: Windows, Monolithic Kernels, X86, X11, 2 Party system in America, Saddam Hussain, Communism) Microkernels are great at security, scalability, ease of programming, backwards compatibility, and portability. Come on, does the small added speed of one less context switch really matter that much? Certainly not anymore. Besides, hardware isn't really optimized for context switches yet (no real demand). If it were, perhaps microkernels could outperform monolithic.
The point is, for this fairytale view of magical computers fixing themselves to come true, there is a huge barrier that would have to be crossed, and I don't think Linus is willing to cross it. Thus my question is, will the open source community ever really accept such a fundamental change in the kernel, or will it be MS, Sun, and IBM that pioneer the future yet again? It just seems to me that Linux will be playing catch up if it doesn't. If Microsoft is really serious about security and really does do a good job, then what is linux going to have that Windows doesn't other than price?
I guess the issue that concerns me is if open source can make large transitions as easily as a company. One might say yes, but you have to think of the coordination issues involved with so many developers. It's important to know if large scale open source systems can keep up or surpass their commercial equivalents. It's relevant because sometimes the technically superior still can't get enough support amongst open source communities in some cases pertaining to the OS. (Think about Rieser's plite. He had to turn to MP3.com for funding for a great product.) Can Linux move quick enough? Does it have too much inertia?
cvarc.org how to build a full wave directional wire antenna (my favorite kind... I used to build a lot of them when I was 15-18 Very good for ham, cb, FM, and TV (built one for a friend to pick up the super bowl for his party))
softcom.net Javascript calculator for the lengths of the wire
other than that, you just hook a couple antennas together and all is well. There aren't any links to the copper ring antenna or the double quad design that I'm aware of. I kinda made those from my head and bits of other designs and techniques(horizontal stacking free.fr is has pictures of double quads (stacked quads) for UHF, 2.4Ghz is only 6 inches in its longest direction.). If you want detailed instructions just email me and I'll forward you what I've sent to the other guy who asked. If there are more than 5 who care, I'll actually take pictures and put it on a web-page.
If you build quads and need a refelctor (the big metal piece behind them in the pictures) use a cd that is about 1" behind it. You can use that in conjunction with a satelite dish, and it's much better than the coffee can satelite feed that I've seen around:)
And if you are just crazy, you can build an antenna array like this with stacking : bigskyspaces.com
every time you double the number of antenna's you use, you get 3db more gain. 2=3db 4=6db 8=9db etc. also keep in mind that the antenna's in the wireless devices are about -9 db, so just about anything you do will make them better. I bet wrapping wire around it would improve it. It isn't even concievable that linksys even tried to make an antenna for the PCMCIA cards and the USB adapters. They are so terrible, that a spoon would probably have worked better. The WAP's on the other hand are good dipole antenna's. This makes the radiation pattern circle the antenna. If you aren't on the same level as the WAP, you'ld be better off pointing one antenna to the side. Of course I modified mine to be a quad, and my USB (only externally) so that I could get 75-85% signal strength through two floors (up to my bedroom).
It's definatly an addictive hobby.
Bad laptop antenna's + repeater
on
Cringely's Bank Shot
·
· Score: 4, Informative
I've done some work myself on making a passive repeater for other purposes. I've found that even duct-taping a copper circle of one wave length onto my usb wireless adapter for my laptop will improve link quality more than 10% when you go through a few obstacles. I've been lazy, but if you want to do something pretty cool build a directional antenna (any with good gain) and run the cable to a copper circle of length 11.168cm(Ch 6) (don't connect the ends to each other, just to the coaxial cable). This should give you much better gain and distance on your laptop:) you could build the double quad antenna (double the wave length in length, looks like and you connect the coax to the center such that it ends up being two stacked quads), and it would give you at least 3db gain more than a single quad and be omnidirectional so you can move your laptop around:) There are lots of documentation on how to build these antenna's. Build a couple and connect them to each other and viola, you've got a passive repeater.
You're exactly right, which is my point really. Sun isn't a charity. I don't expect that they should be. It's still a bit one-sided the way they do things. While their products are mostly free as in beer, they generally are very proprietary. They have a good reason (ala the M$ reference), but they still shouldn't be so misleading(in at least some sence of the word). I like the way IBM handles things better than Sun. IBM charges for DB2 and their AppServer, but they don't pretend that either are open. Sun invites Jakarta and the likes to Java stardards commity, but it's really only there so that they can say "We love open source." Sun "opens" OpenOffice, but they are only using it to fuel the next version of StarOffice for free. It really blurs the lines of open source vs commercial. I'm not saying lets all hate Sun, because Java really isn't that bad anymore. J2EE is getting better. CMP is working better to same extent. They close the standard so people like M$ can't steer the band-wagon. StarOffice is good (albeit slow on older machines). Sun does some amazingly good things, but they have much more power than anyone gives them credit for. A lot of people use Java, and a lot of open source people have some kind of trust for them that I think might be misplaced. I'm still going to use Java, just with caution..Net is the greater evil.
Does this mean they are going to be nicer to the Jakarta folks???
http://jakarta.apache.org/site/news.html under "30 January 2002 - That flaming fireball in the sky..."
Sun's always been friendly to OSS as long as it gives them good press to be so. I'm not certain they are so good at heart. Maybe they were just scarred by microsoft changing the meaning of Java that they don't trust an ad-hoc group of unpaid developers to not do the same.
yeah, he's a fucked up kid, and he deserves punishment. I think you as well as most of the slashdot community have been preaching to the quire even to the kid. He admitted doing wrong, and he didn't say that he doesn't deserve punishment for his hacking, he basically said it was neccisary evil _for him_. He did not say that he shouldn't be punished for his hacking, just that if he got jail time, as opposed to community service and large fines, that it would be because of his opinions. The fact is, his bomb making instructions can be found in chemistry books, but Addison-Wesley won't be infiltrated by the FBI. Also, I've known a few hackers in my day, and usually, they get off with a slap on the rist by the time it's over. The biggest point is, his crimes can not be compared to terrorism or murder, because his were non-violent. His only harm to anyone has been vandalism, and his only crime cracking. He needs to be treated like all the other misguided teens, slapped on the wrist with a fine, community service, and maybe even a psyciatrist to help him deal with his problems.
I think its a perfectly fine reason to feel sorry for a *'d up kid just because they are *'d up.
Don't try to put yourself above this guy, you do the same thing everyday most likely. I doubt you can go the day without checking slashdot a couple of times, perhaps even posting. Add that up over the last few years, and you'll have a much larger sum of REAL costs than the state. Your benifits in entertainment are no different than his only potential monitary benifits.
If your boss knew you checked slashdot on company time which of the following would be his first reaction:
A) Ask you what's new.
B) Tell you not on company time.
C) Fire you.
D) Charge you with a criminal suit that constitutes up to 120y in prison.
They chose D because they thought they could milk money out of the man. They were horribly wrong, and with their intentions should have let the case continue, won, and been sued for reasonable attorney fees + lost wages. It's not too much to ask someone at work who leaves the faucets running to turn them all off before you accuse them of stealing 800,000$ worth of water. Hell, the machines he ran it on probably aren't worth that much. Thats 400 top of the line machines. (2000 e-machine quality computers) Bandwidth consumption is negligable. dnet downloads enough of the key-space in one second to keep the computer occupied for 8 hours of continous work at 56K.
If the guy got off light, they would have repremanded him. They went the full force of the law to try to take everything he owned and put him behind bars for the rest of his life, without showing one cent of real losses. They had absolutly no proof of enough damages to make it to small claims court, so they exagerate the number and make it a criminal case by adding some bogus other charges. Maybe he should have been fired for not acting in the best interests of the state, but 2100$ and 9 years of probation isn't justice, and probably hurts the state more to find a replacement because they did that. I sure the H311 don't want to work for Georgia if I could get that kind of treatment for playing Tetris in my spare time. (Tetris was using 10% of the CPU, that's 35c*60s*60m*10% = 126$ of CPU time you owe us per hour you played + legal fees + your wages. The installation time of Windows would cost more than the computer itself. I wished Microsoft used those kinds of estimates in their Total Cost of Ownership estimates.)
Get real... We're talking about criminal charges for dumpster fishing of CPU cycles. You think it's getting off easy to pay 2100$, legal fees, 80 hours of community service, 9 years of probation, and the loss of two jobs. I know some people who have been running solitare for 30 years about 20% of the time they are at work. That's about the same amount of CPU time. Go lock them up. Anyone could have installed dnet, he just happened to be the admin. The hacking charges don't hold up in any way.
Your definition of misappropriation would include 9 years probation for people that dumpster-fish. Sorry to say it, but I think it requires a bit more than using something the state is squandering to be considered theft in any sence of the word.
You are correct where a time limit is impossed(always) and functionality is more important than documentation(not very often). The big problem occurs when you are writting a library or object that you will have the urge to break compatibility with other code. This is why documentation should come first in most cases. After the documentation, you should implement test cases that force that the code work the way you said it will. Obviously unit tests are valuable (see eXtreme Programming for the reasons such as automated testing). The point is, design of projects, while not being static, needs to be consistent and backwards compatible. GREP won't even save you much time finding all occurances of a given method to fix how it is used when you aren't the only author. The best solution is never tamper with the documentation unless it's really flawed; Only add unit tests or functionality. If you start mutating the code to fit the design you could break more than fix. If you have up to date documentation, you can find what you need much quicker than browsing millions of lines of code. Any project large enough will save time from good documentation from the beginning. (assuming stability is an absolute requirement)
Oh great, my radeon will be obsolete by then!:) NVidia will have released SpaceHeater 4 GTS (Using the plasma state of mass to increase core temperature and power draw by another 120%. Gonna have to run a drive connector to the vidcard) Then ATI will have released something decent to try to compete, but won't have time to write the drivers for it either. Maybe the Radeon drivers will be good enough to make up the difference:)
Being a big freak about Dvorak keyboard(yes! it's better!!), I happen to know that linux supports the Dvorak one-handed keyboard layouts that are required by law in educational institutes for one-handed students. Such an easy keyboard layout to learn, that they often reach 60wpm with just one hand. Some can type up to 80wpm with one hand. That made me feal better that if I ever lost a hand somehow, I could still program effectively:)
Since linux practically runs on every D@#$ thing in the world, is free, has a metric ton of developer tools... I have only one question...
Why doesn't anyone ship a cd that boots linux,X11 4.01, and then runs their 3D game? If a game developer did this would they not be able to run it on Mac,PC(3D accel required), PS2(special ver.), XBOX(sure that's next if not already), Toaster Oven, etc. Seems like if a game developer really wanted to hit the entire market, they could use BSD or Linux pretty easily... I don't know how the GPL plays with shipping a binary linux kernel with a commercial product, but BSD license is all peachy. Even GPL, can you not just ship the source code on the same CD for everything but your game. (Linux distro's do this, so I don't see why game developers can't.) I don't see what I'm missing. Seems to me a larger market base for games + more games for linux is a win-win situation for the Linux community and the game publishers. Is the compiler for those processors just not optimized? Is there no OpenGL X11 support for the video of the PS2? Is Sony intentionally not wanting to compete with computers? Any fine game developers/legal experts/people generally smarter than me want to point out the flaw in my thinking? (Assuming it's flawed because someone else would have already jumped the idea if it was feasible. Then again, it could be at least a new goal for linux. I know a lot of gamers that would enjoy the power of linux. I know linux wasn't developed for realtime processes, but 2.4 is decent. I'm sure a game dev would tweak the scheduler.)
I don't think I agree with the full analogy. Obviously Code is not randomly generated or selected for mutation. We use intelligence to know what's wrong, then we use knowledge to improve. What really evolves in good software is the design, not the code. Sometimes you have to start from scratch again to implement design changes.
We could probably design a new, better human, but sheer evolution will _NEVER_ result in perfection. Design can perfect many small peices of code. Combining these smaller pieces, one can achieve near perfection in a lot less time than sheer luck. Evolution produces local minima's, where design can find the absolute minimum error, and move toward it much quicker. (Think if multiple layer perceptron networks.)
Indeed moire is caused by lack of sampling, but that's really a generic answer for everything though isn't it? If your graphics aren't cool, it's because there aren't enough samples.. etc. Originally, there was the scaling of textures that would cause god-awful appearations in the textures. The original fix was to calculate the true color of a pixel based on the surrounding pixels(or overlapping). The work around was bilinear filtering. When a skew into a third dimension was combined, there were more problems introduced, then intensified by perspective correctness. The solution was obviously trilinear filtering. It does indeed remove moire, but at the cost of bluring the texture beyond recognition in some implementations. The obvious correct way would be some form of tricubic filtering, as would the obvious way to fix two dimensional scaling would be bicubic filtering. I don't even know what the third dimension would do to the equation, but it would obviously be the "correct" way to do things. This doesn't require any more samples(just more math), but would be the theoretic best you could do with a texture of a given size.
The banding that is present in 16-bit of depth is not precievable by the faulty human eye. To test this theory, you could draw one half the screen in a color such as 11110 000000 00000 and the other in 11111 000000 00000 and ask a group of people which half of the screen is brighter. The answer is obviously that no one will see a difference. The problem with banding that you see in most games of 16-bit color is caused when the game requests transparencies greater than 1 bit. (the extra green bit is used as an alpha channel by default, some cards allow more) Even then, the banding due to color is minimal compared to the severe change in alpha levels that one bit makes. Choosing the appropriate color is impossible when you are trying to draw an exlosion, fire, or lightening effect over a texture and only have one bit to set opaque or translucent. 24 bits would have been enough. (6 bits of each RGBA) The use of 32 bits I think is a bit wasteful at this point, but in the future, we would all like to see floating point color implemented in hardware as well(already in place in OpenGL).
No matter how you argue the AA issue, the simple truth is, if you built a display device that can draw the samples you take, then there is absolutely no point in AA. The only place I've seen AA matter is reflections where you can't build a display device to take the samples that a raytracer would. Then again, no FSAA algorithm to date takes reflective AA into account. The point is, if you have a display device that is capable of the same resolution as the human eye (maybe a couple more) then the human eye will do the AA for you. I don't think the rod/cone density is so high that we won't reach that in the next 2 years at the current pace.
Maya and PRMan may not be classified as raytracers, but technically, 3dStudio isn't really a raytracer. They all take shortcuts to emulate reality beyond human perception without taking a month to do so. (where as povray will take an hour to render something perfect that you can't tell a difference in anyhow) While they may not be Ray-Tracers in their full capacity, they definately use the principals of ray-tracing to calculate things that no 3d accelerator to date has. (Caustics etc.) the nVidia demo was not realtime nor at a resolution that many gamers would accept. Given the movie has much better physics than most games (arguably more important to realism than rendering), I can't help but to see where ray-tracing would make things at least appear more organic.
I want to see people turn red from bloodflow to their head when they get out of breath.(also missing from many games is stamina) I want to see a room as cluttered as mine, and pavement that isn't perfectly flat. These require polygons(or more CSG for raytracers). I don't care about AA as much an an environment that is more indepth. I'm not focusing on the telephone pole in the distance, but rather the landscape in my immediate proximity. I think its quite amusing to see pictures an Anandtech comparing different AA methods in games that have perfectly smooth buildings and roads. Where's the litter blowing in the wind?:)
I'm not quite so optimistic about raytracing on a chip. I know a bit of math, and I'm afraid out of my wits to think of the calculations that a raytracer does. 3dStudio is not made by mortals, for them to estimate raytracing so effectively. I fear that polygons are about the best I'll see done in hardware in my lifetime (perhaps a nurb). I can't concieve how someone can trace the path of a simulated photon a million times or more for just one frame in hardware, much less anything realtime. Even raytracers are just now dealing with caustics in a more natural way. Traditional raytracing traced the path from the camera to the light sources. The new twist is emulating photons themselves. I think graphics is infinetly expanding. Thus NVidia is pretty well justified in pumping out as much power as they can. It will be used!:) Maybe Doom3 won't max it out, but I'm sure something will, and I assure you it won't take long. I think there will be a focus shift soon from pretty graphics to the long neglected physics. I hated to see in FF the movie where a character could shift balance in an impossible way for a real human. Or in games, when players are injured, they don't have wounds.
Moire is caused by bilinear filtering... Trilinear filtering will remove that. I would think the advantage would be more polygons instead of more resolution. A more in-depth world would be possible with more polygons, where-as actual visual enhancements are almost negligable anymore. The only real advantage to 32 bit color was more alpha levels. No one had taken much advantage of this until recently. 32-bit textures are pointless really, unless you are using multi-texturing. If you've played Giants: Citizen Kabuto, then you can see the need for more power. Mostly because it tries to render the entire landscape with no fogging. The landscape is very detailed, so there are very many polygons. The result is 25 FPS at 640x480 on a Geforce2. I would really like to see a clutered room or hair in a game, but it's quite dificult without bogging down the card with too many polygons. It's always a misconception amongst gamers that the FPS is based on resolution and bit-depth. The biggest contributor to the hardware push is more realistic environment, not more resolution. We've pretty much maxed out resolution, bitdepth, and frame rate. We need hair, a responsive detailed environment, and better texturing. I want to see some more raytracing-ish features such as realistic lighting, caustics, and better photon estimation. For example, when you shine a light on skin, the surrounding skin ilimunates, not reflects more of that dull pastel color they call "skin tone".
You're absolutely right about AA though. I think there is probably a better way to handle jaggies though. Current AA is just as capable of loosing a telephone pole as just raw rendering if by chance the samples miss the pole. AA does great for slanted lines, but I think 2x the resolution would do more without inventing more math. Good Ole display devices just can't handle it. Until the resolution of the device is the resolution of the retina, someone will still be displeased. (maybe further with fools thinking that 200 FPS is somehow better than 85)
BTW: Good graphics do not a game make. (Nethack, Chess, Final Fantasy II (US)) MMMmmm FF2, the best game I've ever played used sprites. I still play it.
It's sad, but Square Soft still has to use a Ray-Tracer for cinematics because NVidia is still cracking that one. I want a raytracer an a chip, but thats just not going to happen until quantum computing:)
The most irritating thing I see is people wasting polygons. You shouldn't use squares(or quads as they call them) for drawing a sphere. The graphics engine immediatly converts them to triangles. These triangles are not on a plane perpendicular to the line intersecting the center of the sphere and the center of the triangle. They are alse not equilateral, therefor they are less efficient than a fractal triangular sphere. (a right tetrahedron with a serpinski triangle fractal of increasing distance from the center. I've implemented one on my own if anyone wants the code, email me). If you're not using trinagles, you probably aren't thinking about it correctly:) (except cubes of course)
Yes, Win2k does very well at fixing problems. So does fsck, but that's a whole other issue. NTFS on Win2k is mighty good, but it's hosed my system twice. I thought it was my fault until I had a couple of friends go through the FS loss.. Not just files, the whole system. Win2k is excellent at fixing errors on the system if you can make it that far. Ext2 hasn't ever corrupted to the point that I couldn't fix it with fsck to get my system back up. I've had to reinstall windows a few times now, but I have to use it to make money. I've found a way to keep things pretty safe with windows. I have a linux system with Samba and Reiser where I keep my really important mission critical data. I keep a copy (for performance) on a FAT32 drive of mine (so I can access in windows and linux... I've heard that there is a windows device driver for Ext2 but I can't find it, only an explorer app.)
Anyhow, saying windows can fix errors on the system is only comparing ScanDisk(TM) with fsck. I was merely pointing out that linux filesystems don't really loose the whole fs like Win2k has done to me and a few friends. As I understand it, Win2K can still corrupt the MFT, but even this kernel bug doesn't hose the filesystem.
(I run linux on the same computer, and I've never had a problem even with 2.4.15 (running it for a few weeks now) Guess I should patch to be on the safe side:) )
My view point is still the same on this kernel though... I've considered anything since the new VM developmental. I still love the kernel hackers even if they've been drinking Decaf. for the past few months:) I'm not trying to justify buggy kernels in a stable tree, but this whole issue is blown out of proportion. Most people are using ReiserFS, EXT3, or XFS. I don't think this bug effects them. I can find no information nor cases where the FS has been corrupted on a journalled filesystem. I'm not a subscriber to the LKML and no one has posted any info about that except of uncertainty. I've done quite a bit of research on ReiserFS (http://www.namesys.com for more info on ReiserFS) and we all know about Ext2(have no link). I can see easily how Ext2 would be effected if Inodes where not written back from cache, but I'm not so certain about ReiserFS. If the Inodes aren't written back, the data should still be valid, just outdated since "metadata" is re-written instead of over-written.
I think that ReiserFS might be able to cope with it. The transaction log is not stored in/as inodes, therefore the transaction log replay should fix the B*Tree of data if it's broken. I'm not sure of other journalling filesystems... I'm glad I use raw partitions for my MySQL/InnoDB databases:) I could be wrong about the ReiserFS thing, so if anyone has any info to the contrary, it would be much appreciated. It only seems logical that FS's that were written for the possibility that Inodes wouldn't be flushed would be just fine.
BTW... has anyone ever seen NTFS corrupt. I've seen it happen to many before. It's not pretty... you can't recover a single file. Still its better than FAT (with all the corruption there, I wonder why it hasn't formed it's own lifeform from the randomness) Still... this is no excuse for not testing software/major changes in a stable tree.
Those who say that no one should have downloaded the new kernel just aren't thinking. We all should for this particular reason. We need to find the bugs fast and keep our beloved kernel developers on their toes:) I'ld rather know now than when I actually need to run the kernel.
Do you really think that if 170 patches are going to give minimal increase in performance that a few more are going to make it any better? I think the benchmarks show a lack of justification for switching to a kernel that still hasn't been fully tested. At least not on a production system.
Also I think the author of the benchmark brings up a good topic about the schedular being as important as the VM. IMHO, I think the the schedular needs a couple of tweeks. The only real way to reduce page faults is to schedual processes that will likely reference pages already in memory over those that have a greater possibility to access swapped pages. In the interim of executing these other processes, you could issue the DMA commands to swap pages so long as the processes you schedualed are unlikely to use parts of the disk that are not cached. (Thus blocking anyhow, just select another process) I'm uncertain of how this might affect old PIO drives, or if it is completely feasible, but it doesn't seam that hard of a task.
it wouldn't change the minds of the leaders, but I don't think they would have nearly the following in their countries if the US played nice. We have crazy fanatics in america, they just don't have the followers because people aren't that upset with the government. Bringing people together for insane acts requires a lot of hatred to play off of to begin with. Hitler made his own hatred, Bin Laden uses our involvement in the Mid-East to unite his people.
Re:Actually...
on
Globalization
·
· Score: 1, Offtopic
With coorperations like Nike et al. its pretty obvious why a lot don't like the US. The Avg Joe just doesn't put it together. We slaved out Cuba for sugar after the Spanish American war. Castro used that to round up supporters. As you all know, we almost had a nukes fired at us.
The old linux VM seamed incredibly insiteful to me. It is far superior to any other operating systems that I've seen out there. It keeps good stats on how long its been since the page was last accessed and swaps pages accessed least recently first and gives a bonus to clean pages.
The problems they describe is a fault of the schedular for picking a process that had a lot of pages just swapped. If the memory gets low, you need to try to give processes that have all their pages in memory and aren't blocking priority so they can do their thing and hopefully free some ram up. Thats why the guy made the kill patch it sounds like, but it should rather suspend than kill. There is nothin you can do if you are just running too many processes anyhow though. You have to swap out a bunch of pages for one process to swap in the pages from another. It's called thrashing, and the only thing you can really do is give one priority over the other, or get more ram. Hey, rams cheap, buy more if your gonna use that much:) Don't cause so much strife by switching key elements of the kernel in a stable tree.
I don't know how the guy pulled it off with the fact that paging is VERY arcitecture dependant. (the kernel pretty much ignores some of the features of X86 (segmentation) to make it more portable) I think their both right, but maybe linus should call it 2.6 so people can have some time to make their unofficial patches still work with it and have some time for testing before its used in a production environment.
This is a very good idea. They are doing a great job of getting accelerated drivers. Also, I think it would be very easy to port opengl to it as well. all of opengl is in libGL and libGLU. The window creation and manipulation code is in glx, which can probably easily be recreated for directFB. It would also be easy to port to other operating systems, because all video cards have a framebuffer. You could probably keep the closed source nvidia driver and still have hardware 3d acceleration, but you'll still have to write the 2d for nvidia if they wouldn't.
However, they would rather be creating a GUI/wm for DirectFB more than DirectFB. It would be nice to see DirectFB used instead of X to begin with. They might even decide to keep it that way so that it could be easily ported and the driver database kept current.
Almost all unix users, linux users, as well as developers that work directly with the X API, are disgusted by the complexity of X. X is nice because I can connect across the internet to it, but the X11 protocols for doing that are no faster than VNC. It's compatible with X Terminals, but they are all gone now save a few that could be replaced for 200$. The configuration to support accelerated video, USB mice, and anything else, is all done in one nice file that requires an experienced X user to fix if something goes wrong on auto detection.
This leaves me wondering why X still exists. I would be extremely happy if the BlueOS team could implement a replacement for it. There isn't a concievable way that it couldn't be better than X. I was kinda disappointed that they were gonna start with X. They could probably replace X quicker than they could learn and develop for it.
Maybe, they can make an interface worthy of teaching my mother:) Ahh, the mother test. Its harder than the turing test. Linux isn't ready for the desktop until I can teach it to my mother. Now I have hope that she'll still be able to see the monitor when that day comes:) (Reaching into my big basket of dreams and optimtism...) Maybe they can make an interface thats as easy to use as the Mac is noted to be, and handles fonts and media in a resonable way.
to bugs. Bugless programs don't need error checking, just input bounds checking. Use Lisp and prove your program with mathmatical induction, or optionally, you can keep the same mindset in C. Just don't let the user screw up. If you have a finite set of inputs, you can very easily see that your program won't fail. I find it much easier to create functions to test that certain things work as expected than putting try blocks around things.
Now that I've given all the tips that I'm aware of, its time for the justification of my own faulty behavior that can't be justified:)
I think open source software does well for bug handling though. The bigest things I can think of that a lot of open source projects have faults with were never meant to be mission critical || are v1.0 || miss coordination caused some negative synergey. As for the first two, you should expect failures. The last is going to happen to even the best. I think it is a testament to OSS still. With such little time to invest, all the products I've seen get better every day.
And here come the excuses...:)
I really wouldn't call it laziness, but more a lack of motivation. The bulk of OSS is written in a geeks spare time, which in itself is small if the geek attends college and works. You have to account for all the reading a geek has to do on a daily basis. (Slashdot, Freshmeat, Changelogs, Anandtech et al, Pricewatch & EBay) Then account for all the time A geek spends perfecting his own system. (New kernel, apt-get, compiling his special favorite programs(MySQL, Apache, PostgreSQL, XBill)) By the time you get done with all the things you try to stay on top, you really don't have much time left. From there on out, your sleepy and are working purely on caffiene. You will enevitably make a few mistakes.:) It's not laziness... just tired, and you aren't really getting paid for the work, so why try as hard as you do for paid work? I'm not even metioning games which I believe are essential. All programmers need to take their frustrations out on some helpless AI creature, or else they would buckle under the stress.
Before someone says it, I know the rewards of OSS programming. If there were no rewards, then no one would do it in the first place.
Even granted the fact that Microsoft is gaining ground in technical side of the aspect(less crashing) , they are loosing it more rapidly in the feedom and privacy arena, which until a month ago was becoming ever increasingly important to the average Joe.
Microsoft is not friendly to developers as the artical suggests. There will always be people like Adobe that have to rewrite their applications for other operating systems, and they will suffer from Microsoft's unwillingness to cooperate. The things 3rd party developers must worry about are sometimes as menial as how windows doesn't handle fonts the same as a Mac, to the enevitability that the X-Box won't support OpenGL out of the box. (NVidia's version aside, also, I'm sure someone will play XBill on it in a week:)) A project I was working on for windows involving TAPI and Mail Merging in particular was twice as hard as it should have been. At one point I contemplated merging manually into HTML or postscript. Did you know that office quietly truncates SQL queries from the COM interface of mail merge to 512 bytes over 2 seperate 256 byte fields?? Also take a look at TAPI sometime; in order to fully use it properly, you must convelude your code such that you are ashamed to have written it.
On the other side of things:
OSS can't compete:
The one thing that I notice about all of open source software is the complete lack of good documentation. I don't know about many people on here, but if you've worked with MSDN, then you know that something is definately missing from OSS documentation. No, man doesn't count. There is a lot of documentation on how to use various tools, but its very hard to even find out how to create a window in X without using SDL or GGI. You can't expect a relatively new programmer to grep 1G of source to understand all the API calls to create a graphical version of FTP that takes all of a day to write in VB or Borland Builder/Delphi for windows. The OSS community could make things much more enticing for new developers by giving them a standard that if the software follows it is gauranteed to run on any distrabution without a headache (Quake3 is an excellent example, ID doesn't want to make another version of their software for linux due to tech support issues) Sun does the same for Java and the numbers speak for them, not by users choice, but the convenience to developers. Linux is also prohibitive in the fact that it almost certainly requires hardware manufacturers to release more to the community than windows does, or pay developers to maintain the drivers functionality with every OS change (NVidia chooses to do their own driver, and I can tell they struggle... Promise tries as well, but the SCSI driver code base changes with almost every revisionof the kernel). The result is very poor hardware support, even with IBM's help.
But, then again, OSS software maight get a bit of a kick from the commercial entities:
Microsoft's success or failure might lie in the hands of Apple. Apple's ability to make a stable, secure, OSS underlying OS that is easy for the average person to use, easy for the average programmer to make inexpensive or free software for, and easy for coorperations to adopt without loosing functionality or money, is a variable that still gives me hope that I won't have to run XP on anything but a test bed. Macs are more expensive because of the proprietary nature of the hardware, but if they release a X86 version of the GUI, then they would have much more market. Most of the software I have to use Windows for has a Mac counterpart. Mac OS's reign in compatibility with itself. Also many companies have a few macs and are open to experimentation with them.
The bottom line is: With Bush as president, MS is pretty much given free reign to be as monopolistic and anti-privacy as they wish. Votes tallied with MS Election.NET next term?
I've been running Diablo II on Linux for a while now... You have to run Game.exe of a no cd crack. Running Diablo.exe is useless, and the game won't see what it expects when it tries to read the CDROM:) Be sure to tell wine to use a desktop of 640x480
Half-Life already works too
Quake is available for linux already as well as UT.
This leads me to the question of what is this company expect to give the linux community... there are a lot of game projects out there for linux too as well as the ability to run a lot of windows games as well. If I donated $5 to anyone, it would be to the fine developers of wine:) (or one of the OS game projects)
Slashdot Mirror
Hmm... sounds like a micro-kernel :)
I like the Hurd's principal, but all the good ideas die because they aren't mainstream as quick as the mediocre ones. It's quite hard to edge out the market once a standard has been set. (Evidence: Windows, Monolithic Kernels, X86, X11, 2 Party system in America, Saddam Hussain, Communism) Microkernels are great at security, scalability, ease of programming, backwards compatibility, and portability. Come on, does the small added speed of one less context switch really matter that much? Certainly not anymore. Besides, hardware isn't really optimized for context switches yet (no real demand). If it were, perhaps microkernels could outperform monolithic.
The point is, for this fairytale view of magical computers fixing themselves to come true, there is a huge barrier that would have to be crossed, and I don't think Linus is willing to cross it. Thus my question is, will the open source community ever really accept such a fundamental change in the kernel, or will it be MS, Sun, and IBM that pioneer the future yet again? It just seems to me that Linux will be playing catch up if it doesn't. If Microsoft is really serious about security and really does do a good job, then what is linux going to have that Windows doesn't other than price?
I guess the issue that concerns me is if open source can make large transitions as easily as a company. One might say yes, but you have to think of the coordination issues involved with so many developers. It's important to know if large scale open source systems can keep up or surpass their commercial equivalents. It's relevant because sometimes the technically superior still can't get enough support amongst open source communities in some cases pertaining to the OS. (Think about Rieser's plite. He had to turn to MP3.com for funding for a great product.) Can Linux move quick enough? Does it have too much inertia?
cvarc.org
:)
how to build a full wave directional wire antenna (my favorite kind... I used to build a lot of them when I was 15-18 Very good for ham, cb, FM, and TV (built one for a friend to pick up the super bowl for his party))
softcom.net
Javascript calculator for the lengths of the wire
other than that, you just hook a couple antennas together and all is well. There aren't any links to the copper ring antenna or the double quad design that I'm aware of. I kinda made those from my head and bits of other designs and techniques(horizontal stacking free.fr is has pictures of double quads (stacked quads) for UHF, 2.4Ghz is only 6 inches in its longest direction.). If you want detailed instructions just email me and I'll forward you what I've sent to the other guy who asked. If there are more than 5 who care, I'll actually take pictures and put it on a web-page.
If you build quads and need a refelctor (the big metal piece behind them in the pictures) use a cd that is about 1" behind it. You can use that in conjunction with a satelite dish, and it's much better than the coffee can satelite feed that I've seen around
And if you are just crazy, you can build an antenna array like this with stacking : bigskyspaces.com
every time you double the number of antenna's you use, you get 3db more gain. 2=3db 4=6db 8=9db etc. also keep in mind that the antenna's in the wireless devices are about -9 db, so just about anything you do will make them better. I bet wrapping wire around it would improve it. It isn't even concievable that linksys even tried to make an antenna for the PCMCIA cards and the USB adapters. They are so terrible, that a spoon would probably have worked better. The WAP's on the other hand are good dipole antenna's. This makes the radiation pattern circle the antenna. If you aren't on the same level as the WAP, you'ld be better off pointing one antenna to the side. Of course I modified mine to be a quad, and my USB (only externally) so that I could get 75-85% signal strength through two floors (up to my bedroom).
It's definatly an addictive hobby.
I've done some work myself on making a passive repeater for other purposes. I've found that even duct-taping a copper circle of one wave length onto my usb wireless adapter for my laptop will improve link quality more than 10% when you go through a few obstacles. I've been lazy, but if you want to do something pretty cool build a directional antenna (any with good gain) and run the cable to a copper circle of length 11.168cm(Ch 6) (don't connect the ends to each other, just to the coaxial cable). This should give you much better gain and distance on your laptop :) you could build the double quad antenna (double the wave length in length, looks like and you connect the coax to the center such that it ends up being two stacked quads), and it would give you at least 3db gain more than a single quad and be omnidirectional so you can move your laptop around :) There are lots of documentation on how to build these antenna's. Build a couple and connect them to each other and viola, you've got a passive repeater.
You're exactly right, which is my point really. Sun isn't a charity. I don't expect that they should be. It's still a bit one-sided the way they do things. While their products are mostly free as in beer, they generally are very proprietary. They have a good reason (ala the M$ reference), but they still shouldn't be so misleading(in at least some sence of the word). I like the way IBM handles things better than Sun. IBM charges for DB2 and their AppServer, but they don't pretend that either are open. Sun invites Jakarta and the likes to Java stardards commity, but it's really only there so that they can say "We love open source." Sun "opens" OpenOffice, but they are only using it to fuel the next version of StarOffice for free. It really blurs the lines of open source vs commercial. I'm not saying lets all hate Sun, because Java really isn't that bad anymore. J2EE is getting better. CMP is working better to same extent. They close the standard so people like M$ can't steer the band-wagon. StarOffice is good (albeit slow on older machines). Sun does some amazingly good things, but they have much more power than anyone gives them credit for. A lot of people use Java, and a lot of open source people have some kind of trust for them that I think might be misplaced. I'm still going to use Java, just with caution. .Net is the greater evil.
Does this mean they are going to be nicer to the Jakarta folks???
http://jakarta.apache.org/site/news.html under "30 January 2002 - That flaming fireball in the sky..."
Sun's always been friendly to OSS as long as it gives them good press to be so. I'm not certain they are so good at heart. Maybe they were just scarred by microsoft changing the meaning of Java that they don't trust an ad-hoc group of unpaid developers to not do the same.
yeah, he's a fucked up kid, and he deserves punishment. I think you as well as most of the slashdot community have been preaching to the quire even to the kid. He admitted doing wrong, and he didn't say that he doesn't deserve punishment for his hacking, he basically said it was neccisary evil _for him_. He did not say that he shouldn't be punished for his hacking, just that if he got jail time, as opposed to community service and large fines, that it would be because of his opinions. The fact is, his bomb making instructions can be found in chemistry books, but Addison-Wesley won't be infiltrated by the FBI. Also, I've known a few hackers in my day, and usually, they get off with a slap on the rist by the time it's over. The biggest point is, his crimes can not be compared to terrorism or murder, because his were non-violent. His only harm to anyone has been vandalism, and his only crime cracking. He needs to be treated like all the other misguided teens, slapped on the wrist with a fine, community service, and maybe even a psyciatrist to help him deal with his problems.
I think its a perfectly fine reason to feel sorry for a *'d up kid just because they are *'d up.
What ever happened to no harm no foul?
Don't try to put yourself above this guy, you do the same thing everyday most likely. I doubt you can go the day without checking slashdot a couple of times, perhaps even posting. Add that up over the last few years, and you'll have a much larger sum of REAL costs than the state. Your benifits in entertainment are no different than his only potential monitary benifits.
If your boss knew you checked slashdot on company time which of the following would be his first reaction:
A) Ask you what's new.
B) Tell you not on company time.
C) Fire you.
D) Charge you with a criminal suit that constitutes up to 120y in prison.
They chose D because they thought they could milk money out of the man. They were horribly wrong, and with their intentions should have let the case continue, won, and been sued for reasonable attorney fees + lost wages. It's not too much to ask someone at work who leaves the faucets running to turn them all off before you accuse them of stealing 800,000$ worth of water. Hell, the machines he ran it on probably aren't worth that much. Thats 400 top of the line machines. (2000 e-machine quality computers) Bandwidth consumption is negligable. dnet downloads enough of the key-space in one second to keep the computer occupied for 8 hours of continous work at 56K.
If the guy got off light, they would have repremanded him. They went the full force of the law to try to take everything he owned and put him behind bars for the rest of his life, without showing one cent of real losses. They had absolutly no proof of enough damages to make it to small claims court, so they exagerate the number and make it a criminal case by adding some bogus other charges. Maybe he should have been fired for not acting in the best interests of the state, but 2100$ and 9 years of probation isn't justice, and probably hurts the state more to find a replacement because they did that. I sure the H311 don't want to work for Georgia if I could get that kind of treatment for playing Tetris in my spare time. (Tetris was using 10% of the CPU, that's 35c*60s*60m*10% = 126$ of CPU time you owe us per hour you played + legal fees + your wages. The installation time of Windows would cost more than the computer itself. I wished Microsoft used those kinds of estimates in their Total Cost of Ownership estimates.)
Get real... We're talking about criminal charges for dumpster fishing of CPU cycles. You think it's getting off easy to pay 2100$, legal fees, 80 hours of community service, 9 years of probation, and the loss of two jobs. I know some people who have been running solitare for 30 years about 20% of the time they are at work. That's about the same amount of CPU time. Go lock them up. Anyone could have installed dnet, he just happened to be the admin. The hacking charges don't hold up in any way.
Your definition of misappropriation would include 9 years probation for people that dumpster-fish. Sorry to say it, but I think it requires a bit more than using something the state is squandering to be considered theft in any sence of the word.
You are correct where a time limit is impossed(always) and functionality is more important than documentation(not very often). The big problem occurs when you are writting a library or object that you will have the urge to break compatibility with other code. This is why documentation should come first in most cases. After the documentation, you should implement test cases that force that the code work the way you said it will. Obviously unit tests are valuable (see eXtreme Programming for the reasons such as automated testing). The point is, design of projects, while not being static, needs to be consistent and backwards compatible. GREP won't even save you much time finding all occurances of a given method to fix how it is used when you aren't the only author. The best solution is never tamper with the documentation unless it's really flawed; Only add unit tests or functionality. If you start mutating the code to fit the design you could break more than fix. If you have up to date documentation, you can find what you need much quicker than browsing millions of lines of code. Any project large enough will save time from good documentation from the beginning. (assuming stability is an absolute requirement)
Oh great, my radeon will be obsolete by then! :) NVidia will have released SpaceHeater 4 GTS (Using the plasma state of mass to increase core temperature and power draw by another 120%. Gonna have to run a drive connector to the vidcard) Then ATI will have released something decent to try to compete, but won't have time to write the drivers for it either. Maybe the Radeon drivers will be good enough to make up the difference :)
Being a big freak about Dvorak keyboard(yes! it's better!!), I happen to know that linux supports the Dvorak one-handed keyboard layouts that are required by law in educational institutes for one-handed students. Such an easy keyboard layout to learn, that they often reach 60wpm with just one hand. Some can type up to 80wpm with one hand. That made me feal better that if I ever lost a hand somehow, I could still program effectively :)
Since linux practically runs on every D@#$ thing in the world, is free, has a metric ton of developer tools... I have only one question...
Why doesn't anyone ship a cd that boots linux,X11 4.01, and then runs their 3D game? If a game developer did this would they not be able to run it on Mac,PC(3D accel required), PS2(special ver.), XBOX(sure that's next if not already), Toaster Oven, etc. Seems like if a game developer really wanted to hit the entire market, they could use BSD or Linux pretty easily... I don't know how the GPL plays with shipping a binary linux kernel with a commercial product, but BSD license is all peachy. Even GPL, can you not just ship the source code on the same CD for everything but your game. (Linux distro's do this, so I don't see why game developers can't.) I don't see what I'm missing. Seems to me a larger market base for games + more games for linux is a win-win situation for the Linux community and the game publishers. Is the compiler for those processors just not optimized? Is there no OpenGL X11 support for the video of the PS2? Is Sony intentionally not wanting to compete with computers? Any fine game developers/legal experts/people generally smarter than me want to point out the flaw in my thinking? (Assuming it's flawed because someone else would have already jumped the idea if it was feasible. Then again, it could be at least a new goal for linux. I know a lot of gamers that would enjoy the power of linux. I know linux wasn't developed for realtime processes, but 2.4 is decent. I'm sure a game dev would tweak the scheduler.)
I don't think I agree with the full analogy. Obviously Code is not randomly generated or selected for mutation. We use intelligence to know what's wrong, then we use knowledge to improve. What really evolves in good software is the design, not the code. Sometimes you have to start from scratch again to implement design changes.
We could probably design a new, better human, but sheer evolution will _NEVER_ result in perfection. Design can perfect many small peices of code. Combining these smaller pieces, one can achieve near perfection in a lot less time than sheer luck. Evolution produces local minima's, where design can find the absolute minimum error, and move toward it much quicker. (Think if multiple layer perceptron networks.)
Indeed moire is caused by lack of sampling, but that's really a generic answer for everything though isn't it? If your graphics aren't cool, it's because there aren't enough samples.. etc. Originally, there was the scaling of textures that would cause god-awful appearations in the textures. The original fix was to calculate the true color of a pixel based on the surrounding pixels(or overlapping). The work around was bilinear filtering. When a skew into a third dimension was combined, there were more problems introduced, then intensified by perspective correctness. The solution was obviously trilinear filtering. It does indeed remove moire, but at the cost of bluring the texture beyond recognition in some implementations. The obvious correct way would be some form of tricubic filtering, as would the obvious way to fix two dimensional scaling would be bicubic filtering. I don't even know what the third dimension would do to the equation, but it would obviously be the "correct" way to do things. This doesn't require any more samples(just more math), but would be the theoretic best you could do with a texture of a given size. :)
:) Maybe Doom3 won't max it out, but I'm sure something will, and I assure you it won't take long. I think there will be a focus shift soon from pretty graphics to the long neglected physics. I hated to see in FF the movie where a character could shift balance in an impossible way for a real human. Or in games, when players are injured, they don't have wounds.
The banding that is present in 16-bit of depth is not precievable by the faulty human eye. To test this theory, you could draw one half the screen in a color such as 11110 000000 00000 and the other in 11111 000000 00000 and ask a group of people which half of the screen is brighter. The answer is obviously that no one will see a difference. The problem with banding that you see in most games of 16-bit color is caused when the game requests transparencies greater than 1 bit. (the extra green bit is used as an alpha channel by default, some cards allow more) Even then, the banding due to color is minimal compared to the severe change in alpha levels that one bit makes. Choosing the appropriate color is impossible when you are trying to draw an exlosion, fire, or lightening effect over a texture and only have one bit to set opaque or translucent. 24 bits would have been enough. (6 bits of each RGBA) The use of 32 bits I think is a bit wasteful at this point, but in the future, we would all like to see floating point color implemented in hardware as well(already in place in OpenGL).
No matter how you argue the AA issue, the simple truth is, if you built a display device that can draw the samples you take, then there is absolutely no point in AA. The only place I've seen AA matter is reflections where you can't build a display device to take the samples that a raytracer would. Then again, no FSAA algorithm to date takes reflective AA into account. The point is, if you have a display device that is capable of the same resolution as the human eye (maybe a couple more) then the human eye will do the AA for you. I don't think the rod/cone density is so high that we won't reach that in the next 2 years at the current pace.
Maya and PRMan may not be classified as raytracers, but technically, 3dStudio isn't really a raytracer. They all take shortcuts to emulate reality beyond human perception without taking a month to do so. (where as povray will take an hour to render something perfect that you can't tell a difference in anyhow) While they may not be Ray-Tracers in their full capacity, they definately use the principals of ray-tracing to calculate things that no 3d accelerator to date has. (Caustics etc.) the nVidia demo was not realtime nor at a resolution that many gamers would accept. Given the movie has much better physics than most games (arguably more important to realism than rendering), I can't help but to see where ray-tracing would make things at least appear more organic.
I want to see people turn red from bloodflow to their head when they get out of breath.(also missing from many games is stamina) I want to see a room as cluttered as mine, and pavement that isn't perfectly flat. These require polygons(or more CSG for raytracers). I don't care about AA as much an an environment that is more indepth. I'm not focusing on the telephone pole in the distance, but rather the landscape in my immediate proximity. I think its quite amusing to see pictures an Anandtech comparing different AA methods in games that have perfectly smooth buildings and roads. Where's the litter blowing in the wind?
I'm not quite so optimistic about raytracing on a chip. I know a bit of math, and I'm afraid out of my wits to think of the calculations that a raytracer does. 3dStudio is not made by mortals, for them to estimate raytracing so effectively. I fear that polygons are about the best I'll see done in hardware in my lifetime (perhaps a nurb). I can't concieve how someone can trace the path of a simulated photon a million times or more for just one frame in hardware, much less anything realtime. Even raytracers are just now dealing with caustics in a more natural way. Traditional raytracing traced the path from the camera to the light sources. The new twist is emulating photons themselves. I think graphics is infinetly expanding. Thus NVidia is pretty well justified in pumping out as much power as they can. It will be used!
Moire is caused by bilinear filtering... Trilinear filtering will remove that. I would think the advantage would be more polygons instead of more resolution. A more in-depth world would be possible with more polygons, where-as actual visual enhancements are almost negligable anymore. The only real advantage to 32 bit color was more alpha levels. No one had taken much advantage of this until recently. 32-bit textures are pointless really, unless you are using multi-texturing. If you've played Giants: Citizen Kabuto, then you can see the need for more power. Mostly because it tries to render the entire landscape with no fogging. The landscape is very detailed, so there are very many polygons. The result is 25 FPS at 640x480 on a Geforce2. I would really like to see a clutered room or hair in a game, but it's quite dificult without bogging down the card with too many polygons. It's always a misconception amongst gamers that the FPS is based on resolution and bit-depth. The biggest contributor to the hardware push is more realistic environment, not more resolution. We've pretty much maxed out resolution, bitdepth, and frame rate. We need hair, a responsive detailed environment, and better texturing. I want to see some more raytracing-ish features such as realistic lighting, caustics, and better photon estimation. For example, when you shine a light on skin, the surrounding skin ilimunates, not reflects more of that dull pastel color they call "skin tone".
:)
:) (except cubes of course)
You're absolutely right about AA though. I think there is probably a better way to handle jaggies though. Current AA is just as capable of loosing a telephone pole as just raw rendering if by chance the samples miss the pole. AA does great for slanted lines, but I think 2x the resolution would do more without inventing more math. Good Ole display devices just can't handle it. Until the resolution of the device is the resolution of the retina, someone will still be displeased. (maybe further with fools thinking that 200 FPS is somehow better than 85)
BTW: Good graphics do not a game make. (Nethack, Chess, Final Fantasy II (US)) MMMmmm FF2, the best game I've ever played used sprites. I still play it.
It's sad, but Square Soft still has to use a Ray-Tracer for cinematics because NVidia is still cracking that one. I want a raytracer an a chip, but thats just not going to happen until quantum computing
The most irritating thing I see is people wasting polygons. You shouldn't use squares(or quads as they call them) for drawing a sphere. The graphics engine immediatly converts them to triangles. These triangles are not on a plane perpendicular to the line intersecting the center of the sphere and the center of the triangle. They are alse not equilateral, therefor they are less efficient than a fractal triangular sphere. (a right tetrahedron with a serpinski triangle fractal of increasing distance from the center. I've implemented one on my own if anyone wants the code, email me). If you're not using trinagles, you probably aren't thinking about it correctly
Yes, Win2k does very well at fixing problems. So does fsck, but that's a whole other issue. NTFS on Win2k is mighty good, but it's hosed my system twice. I thought it was my fault until I had a couple of friends go through the FS loss.. Not just files, the whole system. Win2k is excellent at fixing errors on the system if you can make it that far. Ext2 hasn't ever corrupted to the point that I couldn't fix it with fsck to get my system back up. I've had to reinstall windows a few times now, but I have to use it to make money. I've found a way to keep things pretty safe with windows. I have a linux system with Samba and Reiser where I keep my really important mission critical data. I keep a copy (for performance) on a FAT32 drive of mine (so I can access in windows and linux... I've heard that there is a windows device driver for Ext2 but I can't find it, only an explorer app.)
:) )
:) I'm not trying to justify buggy kernels in a stable tree, but this whole issue is blown out of proportion. Most people are using ReiserFS, EXT3, or XFS. I don't think this bug effects them. I can find no information nor cases where the FS has been corrupted on a journalled filesystem. I'm not a subscriber to the LKML and no one has posted any info about that except of uncertainty. I've done quite a bit of research on ReiserFS (http://www.namesys.com for more info on ReiserFS) and we all know about Ext2(have no link). I can see easily how Ext2 would be effected if Inodes where not written back from cache, but I'm not so certain about ReiserFS. If the Inodes aren't written back, the data should still be valid, just outdated since "metadata" is re-written instead of over-written.
:)
Anyhow, saying windows can fix errors on the system is only comparing ScanDisk(TM) with fsck. I was merely pointing out that linux filesystems don't really loose the whole fs like Win2k has done to me and a few friends. As I understand it, Win2K can still corrupt the MFT, but even this kernel bug doesn't hose the filesystem.
(I run linux on the same computer, and I've never had a problem even with 2.4.15 (running it for a few weeks now) Guess I should patch to be on the safe side
My view point is still the same on this kernel though... I've considered anything since the new VM developmental. I still love the kernel hackers even if they've been drinking Decaf. for the past few months
Still would like some more info!!
I think that ReiserFS might be able to cope with it. The transaction log is not stored in/as inodes, therefore the transaction log replay should fix the B*Tree of data if it's broken. I'm not sure of other journalling filesystems... I'm glad I use raw partitions for my MySQL/InnoDB databases :) I could be wrong about the ReiserFS thing, so if anyone has any info to the contrary, it would be much appreciated. It only seems logical that FS's that were written for the possibility that Inodes wouldn't be flushed would be just fine.
:) I'ld rather know now than when I actually need to run the kernel.
BTW... has anyone ever seen NTFS corrupt. I've seen it happen to many before. It's not pretty... you can't recover a single file. Still its better than FAT (with all the corruption there, I wonder why it hasn't formed it's own lifeform from the randomness) Still... this is no excuse for not testing software/major changes in a stable tree.
Those who say that no one should have downloaded the new kernel just aren't thinking. We all should for this particular reason. We need to find the bugs fast and keep our beloved kernel developers on their toes
Do you really think that if 170 patches are going to give minimal increase in performance that a few more are going to make it any better? I think the benchmarks show a lack of justification for switching to a kernel that still hasn't been fully tested. At least not on a production system.
Also I think the author of the benchmark brings up a good topic about the schedular being as important as the VM. IMHO, I think the the schedular needs a couple of tweeks. The only real way to reduce page faults is to schedual processes that will likely reference pages already in memory over those that have a greater possibility to access swapped pages. In the interim of executing these other processes, you could issue the DMA commands to swap pages so long as the processes you schedualed are unlikely to use parts of the disk that are not cached. (Thus blocking anyhow, just select another process) I'm uncertain of how this might affect old PIO drives, or if it is completely feasible, but it doesn't seam that hard of a task.
it wouldn't change the minds of the leaders, but I don't think they would have nearly the following in their countries if the US played nice. We have crazy fanatics in america, they just don't have the followers because people aren't that upset with the government. Bringing people together for insane acts requires a lot of hatred to play off of to begin with. Hitler made his own hatred, Bin Laden uses our involvement in the Mid-East to unite his people.
With coorperations like Nike et al. its pretty obvious why a lot don't like the US. The Avg Joe just doesn't put it together. We slaved out Cuba for sugar after the Spanish American war. Castro used that to round up supporters. As you all know, we almost had a nukes fired at us.
The old linux VM seamed incredibly insiteful to me. It is far superior to any other operating systems that I've seen out there. It keeps good stats on how long its been since the page was last accessed and swaps pages accessed least recently first and gives a bonus to clean pages.
:) Don't cause so much strife by switching key elements of the kernel in a stable tree.
The problems they describe is a fault of the schedular for picking a process that had a lot of pages just swapped. If the memory gets low, you need to try to give processes that have all their pages in memory and aren't blocking priority so they can do their thing and hopefully free some ram up. Thats why the guy made the kill patch it sounds like, but it should rather suspend than kill. There is nothin you can do if you are just running too many processes anyhow though. You have to swap out a bunch of pages for one process to swap in the pages from another. It's called thrashing, and the only thing you can really do is give one priority over the other, or get more ram. Hey, rams cheap, buy more if your gonna use that much
I don't know how the guy pulled it off with the fact that paging is VERY arcitecture dependant. (the kernel pretty much ignores some of the features of X86 (segmentation) to make it more portable) I think their both right, but maybe linus should call it 2.6 so people can have some time to make their unofficial patches still work with it and have some time for testing before its used in a production environment.
This is a very good idea. They are doing a great job of getting accelerated drivers. Also, I think it would be very easy to port opengl to it as well. all of opengl is in libGL and libGLU. The window creation and manipulation code is in glx, which can probably easily be recreated for directFB. It would also be easy to port to other operating systems, because all video cards have a framebuffer. You could probably keep the closed source nvidia driver and still have hardware 3d acceleration, but you'll still have to write the 2d for nvidia if they wouldn't.
However, they would rather be creating a GUI/wm for DirectFB more than DirectFB. It would be nice to see DirectFB used instead of X to begin with. They might even decide to keep it that way so that it could be easily ported and the driver database kept current.
Almost all unix users, linux users, as well as developers that work directly with the X API, are disgusted by the complexity of X. X is nice because I can connect across the internet to it, but the X11 protocols for doing that are no faster than VNC. It's compatible with X Terminals, but they are all gone now save a few that could be replaced for 200$. The configuration to support accelerated video, USB mice, and anything else, is all done in one nice file that requires an experienced X user to fix if something goes wrong on auto detection.
:) Ahh, the mother test. Its harder than the turing test. Linux isn't ready for the desktop until I can teach it to my mother. Now I have hope that she'll still be able to see the monitor when that day comes :) (Reaching into my big basket of dreams and optimtism...) Maybe they can make an interface thats as easy to use as the Mac is noted to be, and handles fonts and media in a resonable way.
This leaves me wondering why X still exists. I would be extremely happy if the BlueOS team could implement a replacement for it. There isn't a concievable way that it couldn't be better than X. I was kinda disappointed that they were gonna start with X. They could probably replace X quicker than they could learn and develop for it.
Maybe, they can make an interface worthy of teaching my mother
to bugs. Bugless programs don't need error checking, just input bounds checking. Use Lisp and prove your program with mathmatical induction, or optionally, you can keep the same mindset in C. Just don't let the user screw up. If you have a finite set of inputs, you can very easily see that your program won't fail. I find it much easier to create functions to test that certain things work as expected than putting try blocks around things.
:)
:)
:) It's not laziness... just tired, and you aren't really getting paid for the work, so why try as hard as you do for paid work? I'm not even metioning games which I believe are essential. All programmers need to take their frustrations out on some helpless AI creature, or else they would buckle under the stress.
Now that I've given all the tips that I'm aware of, its time for the justification of my own faulty behavior that can't be justified
I think open source software does well for bug handling though. The bigest things I can think of that a lot of open source projects have faults with were never meant to be mission critical || are v1.0 || miss coordination caused some negative synergey. As for the first two, you should expect failures. The last is going to happen to even the best. I think it is a testament to OSS still. With such little time to invest, all the products I've seen get better every day.
And here come the excuses...
I really wouldn't call it laziness, but more a lack of motivation. The bulk of OSS is written in a geeks spare time, which in itself is small if the geek attends college and works. You have to account for all the reading a geek has to do on a daily basis. (Slashdot, Freshmeat, Changelogs, Anandtech et al, Pricewatch & EBay) Then account for all the time A geek spends perfecting his own system. (New kernel, apt-get, compiling his special favorite programs(MySQL, Apache, PostgreSQL, XBill)) By the time you get done with all the things you try to stay on top, you really don't have much time left. From there on out, your sleepy and are working purely on caffiene. You will enevitably make a few mistakes.
Before someone says it, I know the rewards of OSS programming. If there were no rewards, then no one would do it in the first place.
Even granted the fact that Microsoft is gaining ground in technical side of the aspect(less crashing) , they are loosing it more rapidly in the feedom and privacy arena, which until a month ago was becoming ever increasingly important to the average Joe.
:)) A project I was working on for windows involving TAPI and Mail Merging in particular was twice as hard as it should have been. At one point I contemplated merging manually into HTML or postscript. Did you know that office quietly truncates SQL queries from the COM interface of mail merge to 512 bytes over 2 seperate 256 byte fields?? Also take a look at TAPI sometime; in order to fully use it properly, you must convelude your code such that you are ashamed to have written it.
Microsoft is not friendly to developers as the artical suggests. There will always be people like Adobe that have to rewrite their applications for other operating systems, and they will suffer from Microsoft's unwillingness to cooperate. The things 3rd party developers must worry about are sometimes as menial as how windows doesn't handle fonts the same as a Mac, to the enevitability that the X-Box won't support OpenGL out of the box. (NVidia's version aside, also, I'm sure someone will play XBill on it in a week
On the other side of things:
OSS can't compete:
The one thing that I notice about all of open source software is the complete lack of good documentation. I don't know about many people on here, but if you've worked with MSDN, then you know that something is definately missing from OSS documentation. No, man doesn't count. There is a lot of documentation on how to use various tools, but its very hard to even find out how to create a window in X without using SDL or GGI. You can't expect a relatively new programmer to grep 1G of source to understand all the API calls to create a graphical version of FTP that takes all of a day to write in VB or Borland Builder/Delphi for windows. The OSS community could make things much more enticing for new developers by giving them a standard that if the software follows it is gauranteed to run on any distrabution without a headache (Quake3 is an excellent example, ID doesn't want to make another version of their software for linux due to tech support issues) Sun does the same for Java and the numbers speak for them, not by users choice, but the convenience to developers. Linux is also prohibitive in the fact that it almost certainly requires hardware manufacturers to release more to the community than windows does, or pay developers to maintain the drivers functionality with every OS change (NVidia chooses to do their own driver, and I can tell they struggle... Promise tries as well, but the SCSI driver code base changes with almost every revisionof the kernel). The result is very poor hardware support, even with IBM's help.
But, then again, OSS software maight get a bit of a kick from the commercial entities:
Microsoft's success or failure might lie in the hands of Apple. Apple's ability to make a stable, secure, OSS underlying OS that is easy for the average person to use, easy for the average programmer to make inexpensive or free software for, and easy for coorperations to adopt without loosing functionality or money, is a variable that still gives me hope that I won't have to run XP on anything but a test bed. Macs are more expensive because of the proprietary nature of the hardware, but if they release a X86 version of the GUI, then they would have much more market. Most of the software I have to use Windows for has a Mac counterpart. Mac OS's reign in compatibility with itself. Also many companies have a few macs and are open to experimentation with them.
The bottom line is: With Bush as president, MS is pretty much given free reign to be as monopolistic and anti-privacy as they wish. Votes tallied with MS Election.NET next term?
I've been running Diablo II on Linux for a while now... You have to run Game.exe of a no cd crack. Running Diablo.exe is useless, and the game won't see what it expects when it tries to read the CDROM :) Be sure to tell wine to use a desktop of 640x480
:) (or one of the OS game projects)
Half-Life already works too
Quake is available for linux already as well as UT.
This leads me to the question of what is this company expect to give the linux community... there are a lot of game projects out there for linux too as well as the ability to run a lot of windows games as well. If I donated $5 to anyone, it would be to the fine developers of wine