UDP is commonly used in games and other time sensitive environments precisely because it lacks reliabilty. With time sensitive data (such as streaming video or unit positions in a game), if the data gets dropped it's not worth it to retransmit, because it woudl be out of date. Therefore, the program just transmits the next update and the user sees a small skip. This is better than getting "out of sync".
TCP is designed to make unreliable networks (like the internet, which only gives "best effort delivery") reliable by ensuring that a stream can be reassembled, in order, with no missing pieces. Read the RFC for more info here. This reliability makes it good for things that need zero corruption (file transfers for example), and aren't time criticial.
You could try finding an old bottle vendor (like the old coke machines) that have a door and allow you to take one, and only one, bottle when the door is opened.
Get yourself a converter to turn your VGA signal into a Component signal. Component video is a common method of analog interconnection between DVD players and other high resolution devices to HDTVs that utilizes three separate coax cables (usually with RCA connectors at each end).
I can't seem to find any right now, but I know I've seen them before on pricewatch. Make sure your video card can output a mode compatible with HDTV (the converters I saw usually did 1024x768).
A converter like this will allow you to run at significantly higher resolutions than your average S-Video cable can seem to provide.
VaxWars has an insteresting story about that very thing. Rather humerous read. After you read that, I reccomend you check out the VaxBar for another good laugh.
This is getting horribly offtopic, but I'll continue anyway. You're free to load stuff into the kernel that isn't GPLed (nvidia does it all the time), but you can't use stuff that the authors of GPL kernel components have said they don't want non GPL code to be able to use. There are GPLed portions of the kernel that aren't exported GPLONLY because the developer thought it was OK, or better in the interest of compatibilty. Non-GPL kernel modules can still use these things, and are free to load into the kernel.
You might say the purpose of that warning is to say that if a symbol doesn't resolve later, itmight be that the symbol is exported GPLONLY and the module isn't GPL. You can safely ignore the warning if you don't mind tainting your kernel.
Of the membrane keyboards I have (I loved the old IBM mechanical things, but they're too loud!), I do prefer my $50 IBM Active Response one better, and it's what I use on my main desktop. But in the other room, I keep a couple of $9 CompUSA branded things. They feel *almost* as good as my IBM, and they're a helluva lot cheaper. Try them out, CompUSA (if you have them in your area), often has a whole pallet of them (I kid you not) stacked at least 3-4 feet high stuck next to the pole at the end of the keyboard isle. They're great for keeping around as spares or got bundling with white box systems.
For me though? I won't be getting rid of my IBM ActiveTouch any time soon. It lacks drain holes, not that it would do well even if it did being a membrane keyboard, but it's firm tactile feedback while still being a quiet membrane keyboard is unmatched by anything else I've used. As a plus, it also has the skinny enter bar with the big slash and pipe key, which I'm sooooo used to now that I can't use keyboards with the other design (my Sun type 4 always throws me off...)
The kernel tainting thing was put into linux because RMS complained as I recall. The GPL doesn't like letting other liscenses get a free ride, so they had to make the GPLONLY thing available to developers.
One could argue that linux also does this for stability though. Before, the only linux kernel drivers you'd ever see were written by linux kernel hackers and were GPLed or under a similar liscense that made the source available. Now, it's not uncommon to see binary only kernel modules (often using wrappers to load them to take into account differing kernel versions). These drivers aren't under the control of any linux developer and haven't been checked over for stability. Basically, the kernel taint message is warning you that "by loading this, your kernel may become unstable and we can't do a thing about it, so don't bother complaining" since kernel modules can do basically anything to the, well, kernel (when loaded, a kernel module becomes a part of the kernel).
Actually, that's probably a bug. I'm pretty sure ipchains is GPLed (it was in 2.2.x), so that compatibilty module is probably old and unupdated and needs to declare it's liscense as GPLed.
Also, I guess you've never had Win2k bitch you out for loading an unsigned driver until you dig around and disable the warning? I bet there's a command line flag for insmod/modprobe to disable that, or just use an older version of modutils that still works and doesn't include the kernel tainting info:)
Remember that loading proprietary modules now "taints" your kernel and that non GPLed modules can't use GPLONLY symbols. This means that if MS wants to do something like load an entire binary compatibility layer, they might very well have to write a lot of stuff themseleves and end up with a rather large kernel module (which linux people generally do not like). Also, sicne they "taint" the kernel, they automagically get the backlash from users when their system crashes since it COULD be the Microsoft stuff. See the problems people have with the crummy (stability/OSS wise, not speed wise) nVidia drivers.
iptables is included with the kernel. It is GPLed and can even be compiled in statically if you don't like modules. It's hardly proprietary to a company (proprietary in the sense that it won't load on other OSes, but would you expect a linux network card driver to load on anything but a linux kernel?)
"strongarm Intel" eh? I thought Intel was the one with the strongarm, not Microsoft applying it against Intel. Oh wait, that was the StrongARM processor...
If you complain/voice your opinion, be nice about it. The last thing your representative wants is to read a bunch of garbage from "Joe USAizain" who is just blabbing off about how he things this bill sucks because everyone at/. does.
Voice your opinions about why you think it's bad, not what others say. If you think it might lock OSS out of PCs, say so. If you think it might make a black market, say so. etc, etc. Try not to go off on tangents without tying them back in to your original topic (SSSCA or whatever it's called these days) and using it to furthur your argument.
Also, KISS. Your rep gets lots of mail and doesn't have time to read 10 page rants. Keep it concise, and offer to provide more info should they be interested (put it on a webpage that they can visit at their leisure so they don't even have to contact you for it). Make sure you don't alienate the peopel who are trying to help you!
While the server may be about as efficeient as you can get at serving up static pages, the web today revolves around serving up dynamic content. If the server can serve up PHP or Perl pages faster, more power to it.
I don't have a cell phone, but that's partly because I can't get good analog service. I've used digital cell phones. While the sound quality may be good when it's there, it's either there or it's not there. When it drops it drops completely. I much preferred good 'ol analog where you might have some "static" noise, but at least you wouldn't completely drop the call very often.
SeaGate SCSI drives perhaps? They use these REALLY TINY jumpers that you practically have to use tweezers to handle. However, check your old pile of Quantum and Maxtor hard drives. Quantum Lightning and the older Maxtors (my 800MB one used them) used the two types of tiny jumpers like are used on some SeaGate SCSI drives.
Your local mom&pop shop probably has thousands of them. I stopped by one that was oging out of business and they gave me a small box (about the size of the ice bucket in your average freezer, which is what I'm now using) full of various sized computer screws. They probably had the same thing only with jumpers:)
Caused no problems for me, other than the traceroute issue you mention (since 10.0.0.0/8 is dropped by most internet routers as it's not supposed to ever be there) and windows doesn't like using a router in the subnet other than it's primary IP as a default route. For example, I give my windows box a public IP, but the router is 10.0.0.1 (to conserve an address), I have to go to the command line and add a static route every time I boot up. Note there might be an easier way to do this in NT, but in 9x this seems to be the only way around that problem. Of course in most unices we have startup scripts to do this for us (networking isn't up when autoexec.bat is run under 9x). I guess you might be able to script soemthign to do this under 9x too though (at least I should hope you can).
I'm not familiar with solaris, but I know schilly cdrecord claims to work on DVD media (right on the front page), and linux has UDF support (the filesystem used on most DVDs, though ISO9660 is used too). However, like a CD, you can put any filesystem you want on it as long as the OS can read it.
It appears that linux can handle larger hard drives now. According to the referenced slashdot article (and the "stuff" contained therein), linux can now handle IDE hard drives in excess of 100 petabytes. That's a pretty large hard drive.
Slashdot Mirror
Sorry about the confusion there, I meant to say that it protects againts corruption due to out of order packet assembly, not packet "damage".
--MonMotha
UDP is commonly used in games and other time sensitive environments precisely because it lacks reliabilty. With time sensitive data (such as streaming video or unit positions in a game), if the data gets dropped it's not worth it to retransmit, because it woudl be out of date. Therefore, the program just transmits the next update and the user sees a small skip. This is better than getting "out of sync".
TCP is designed to make unreliable networks (like the internet, which only gives "best effort delivery") reliable by ensuring that a stream can be reassembled, in order, with no missing pieces. Read the RFC for more info here. This reliability makes it good for things that need zero corruption (file transfers for example), and aren't time criticial.
Hope this helps.
--MonMotha
Exactly, he'll just whip out his BFG10k and suddenly we'll hear this deep voice saying "Quad Damage"...then "Vinny" will be a pile of, well, nothing.
Oh, this isn't in Quake? Nevermind.
You could try finding an old bottle vendor (like the old coke machines) that have a door and allow you to take one, and only one, bottle when the door is opened.
--MonMotha
Get yourself a converter to turn your VGA signal into a Component signal. Component video is a common method of analog interconnection between DVD players and other high resolution devices to HDTVs that utilizes three separate coax cables (usually with RCA connectors at each end).
I can't seem to find any right now, but I know I've seen them before on pricewatch. Make sure your video card can output a mode compatible with HDTV (the converters I saw usually did 1024x768).
A converter like this will allow you to run at significantly higher resolutions than your average S-Video cable can seem to provide.
--MonMotha
VaxWars has an insteresting story about that very thing. Rather humerous read. After you read that, I reccomend you check out the VaxBar for another good laugh.
--MonMotha
I'm pretty sure mine cost $49.99, but I could be wrong. Either way, it's significantly more than the $9 CompUSA brand.
This is getting horribly offtopic, but I'll continue anyway. You're free to load stuff into the kernel that isn't GPLed (nvidia does it all the time), but you can't use stuff that the authors of GPL kernel components have said they don't want non GPL code to be able to use. There are GPLed portions of the kernel that aren't exported GPLONLY because the developer thought it was OK, or better in the interest of compatibilty. Non-GPL kernel modules can still use these things, and are free to load into the kernel.
You might say the purpose of that warning is to say that if a symbol doesn't resolve later, itmight be that the symbol is exported GPLONLY and the module isn't GPL. You can safely ignore the warning if you don't mind tainting your kernel.
Of the membrane keyboards I have (I loved the old IBM mechanical things, but they're too loud!), I do prefer my $50 IBM Active Response one better, and it's what I use on my main desktop. But in the other room, I keep a couple of $9 CompUSA branded things. They feel *almost* as good as my IBM, and they're a helluva lot cheaper. Try them out, CompUSA (if you have them in your area), often has a whole pallet of them (I kid you not) stacked at least 3-4 feet high stuck next to the pole at the end of the keyboard isle. They're great for keeping around as spares or got bundling with white box systems.
For me though? I won't be getting rid of my IBM ActiveTouch any time soon. It lacks drain holes, not that it would do well even if it did being a membrane keyboard, but it's firm tactile feedback while still being a quiet membrane keyboard is unmatched by anything else I've used. As a plus, it also has the skinny enter bar with the big slash and pipe key, which I'm sooooo used to now that I can't use keyboards with the other design (my Sun type 4 always throws me off...)
--MonMotha
The kernel tainting thing was put into linux because RMS complained as I recall. The GPL doesn't like letting other liscenses get a free ride, so they had to make the GPLONLY thing available to developers.
One could argue that linux also does this for stability though. Before, the only linux kernel drivers you'd ever see were written by linux kernel hackers and were GPLed or under a similar liscense that made the source available. Now, it's not uncommon to see binary only kernel modules (often using wrappers to load them to take into account differing kernel versions). These drivers aren't under the control of any linux developer and haven't been checked over for stability. Basically, the kernel taint message is warning you that "by loading this, your kernel may become unstable and we can't do a thing about it, so don't bother complaining" since kernel modules can do basically anything to the, well, kernel (when loaded, a kernel module becomes a part of the kernel).
--MonMotha
Actually, that's probably a bug. I'm pretty sure ipchains is GPLed (it was in 2.2.x), so that compatibilty module is probably old and unupdated and needs to declare it's liscense as GPLed.
:)
Also, I guess you've never had Win2k bitch you out for loading an unsigned driver until you dig around and disable the warning? I bet there's a command line flag for insmod/modprobe to disable that, or just use an older version of modutils that still works and doesn't include the kernel tainting info
--MonMotha
Remember that loading proprietary modules now "taints" your kernel and that non GPLed modules can't use GPLONLY symbols. This means that if MS wants to do something like load an entire binary compatibility layer, they might very well have to write a lot of stuff themseleves and end up with a rather large kernel module (which linux people generally do not like). Also, sicne they "taint" the kernel, they automagically get the backlash from users when their system crashes since it COULD be the Microsoft stuff. See the problems people have with the crummy (stability/OSS wise, not speed wise) nVidia drivers.
--MonMotha
iptables is included with the kernel. It is GPLed and can even be compiled in statically if you don't like modules. It's hardly proprietary to a company (proprietary in the sense that it won't load on other OSes, but would you expect a linux network card driver to load on anything but a linux kernel?)
--MonMotha
Can you hear me NOW? No? Hum, get a crew out here...we need another tower.
Ah, can you hear me NOW? Good!
--MonMotha
"strongarm Intel" eh? I thought Intel was the one with the strongarm, not Microsoft applying it against Intel. Oh wait, that was the StrongARM processor...
It's funny, laugh!
--MonMotha
If you complain/voice your opinion, be nice about it. The last thing your representative wants is to read a bunch of garbage from "Joe USAizain" who is just blabbing off about how he things this bill sucks because everyone at /. does.
Voice your opinions about why you think it's bad, not what others say. If you think it might lock OSS out of PCs, say so. If you think it might make a black market, say so. etc, etc. Try not to go off on tangents without tying them back in to your original topic (SSSCA or whatever it's called these days) and using it to furthur your argument.
Also, KISS. Your rep gets lots of mail and doesn't have time to read 10 page rants. Keep it concise, and offer to provide more info should they be interested (put it on a webpage that they can visit at their leisure so they don't even have to contact you for it). Make sure you don't alienate the peopel who are trying to help you!
--MonMotha
Two Words: Dynamic Content
While the server may be about as efficeient as you can get at serving up static pages, the web today revolves around serving up dynamic content. If the server can serve up PHP or Perl pages faster, more power to it.
--MonMotha
I don't have a cell phone, but that's partly because I can't get good analog service. I've used digital cell phones. While the sound quality may be good when it's there, it's either there or it's not there. When it drops it drops completely. I much preferred good 'ol analog where you might have some "static" noise, but at least you wouldn't completely drop the call very often.
--MonMotha
Microsoft: The Lunchbox, Microsoft: The Doll, Microsoft: The Flame Thrower...*flames*
Then later we see Microsoft: The Toliet Paper, Microsoft: The Menu...
Hold on, you say Mel Brooks ISN'T the CEO of Microsoft? Oh, then nevermind.
--MonMotha
SeaGate SCSI drives perhaps? They use these REALLY TINY jumpers that you practically have to use tweezers to handle. However, check your old pile of Quantum and Maxtor hard drives. Quantum Lightning and the older Maxtors (my 800MB one used them) used the two types of tiny jumpers like are used on some SeaGate SCSI drives.
:)
Your local mom&pop shop probably has thousands of them. I stopped by one that was oging out of business and they gave me a small box (about the size of the ice bucket in your average freezer, which is what I'm now using) full of various sized computer screws. They probably had the same thing only with jumpers
--MonMotha
I thought AMD already discovered this... --MonMotha
I can defeat your compression algorithm. I will set my homepage to return a 1 byte file.
:)
Now, my homepage URL is "http://monmothas.shacknet.nu" (it's down now, my server needs a new mobo). That URL is 29 bytes long.
No where near 100:1 compression. In fact, the ratio was 1:29. Your algorithm is terrible!
Ah, who am I kidding? No one will even see this post probably
--MonMotha
Caused no problems for me, other than the traceroute issue you mention (since 10.0.0.0/8 is dropped by most internet routers as it's not supposed to ever be there) and windows doesn't like using a router in the subnet other than it's primary IP as a default route. For example, I give my windows box a public IP, but the router is 10.0.0.1 (to conserve an address), I have to go to the command line and add a static route every time I boot up. Note there might be an easier way to do this in NT, but in 9x this seems to be the only way around that problem. Of course in most unices we have startup scripts to do this for us (networking isn't up when autoexec.bat is run under 9x). I guess you might be able to script soemthign to do this under 9x too though (at least I should hope you can).
--MonMotha
I'm not familiar with solaris, but I know schilly cdrecord claims to work on DVD media (right on the front page), and linux has UDF support (the filesystem used on most DVDs, though ISO9660 is used too). However, like a CD, you can put any filesystem you want on it as long as the OS can read it.
--MonMotha
It appears that linux can handle larger hard drives now. According to the referenced slashdot article (and the "stuff" contained therein), linux can now handle IDE hard drives in excess of 100 petabytes. That's a pretty large hard drive.
--MonMotha