Re:Crusoe Schmuso (Yeah, I know I'll get flamed...
on
RLX Gets Denser
·
· Score: 4, Informative
One thing that is overlooked in rackmount (especially 1U) servers is power drain. Computer centers don't have unlimited power (in fact since you have to both power the equiptment AND power the refrigeration units to draw the heat from your machines out of the CC the power requirements of your servers becomes something of an issue). The ability to run dozens of servers in a rack without creating a power situtation is a big bonus for computing centers.
One of the big trends in computing recently has been for servers to grow smaller and consume lots more power. Just look at your average P4 or Athlon compared to the old Pentiums and Pentium Pros. It's getting so bad that PCs are drawing
more power than even energy hogging Alpha based machines in some cases. This problem is compunded by everybody putting servers in the smallest boxes prossible, 1U frequently, so that these energy hogs can be stacked all the way to the top of the rack and draw an enormous amount of power from the circut designated for that tile.
Actually, if you read Eric's Commentary you'll discover that he didn't "win". CRC won. CRC screwed Eric hard, and didn't even have the common courtesy to give him a reach around. At least the website will be back up for awhile until CRC decides to issue another printing of the book and shut them down again.
Yeah, Niven will seem soo passe once we start building Ringworlds or Dyson Spheres, and he'll look like a 50's pulp fiction author once we find some actual living smoke rings.
I think Niven will still be read years from now because his ideas are so far out there that they will still be exciting (if completely unachievable) to the people of the future.
Granted some of his stuff probably won't survive, but I don't think there has been any major author throughout history that has churned out nothing but classics, including any of the fine fellows on this list.
Amen to that. I always hated going to the paintball field and discovering that the other team is a bunch of tournament jerks. For some reason these guys try to bend/break every rule they think they can get away with, and couldn't care less if they suck all of the fun right out of the game.
Most flagrantly, if they were hit somewhere out of site of the refs (almost all of the field) they'd just walk behind the next wall and start playing again. Or if they were seen they'd walk around the field and point out the other team members to their buds (or take potshots when nobody was looking).
Sometimes they'd get caught, but usually there weren't nearly enough refs to handle those jerks. The worst part is when they get all arrogant and cocky about taking only one or two losses (in games of 12 on 12 or so) and spend all the between-game time telling you how much you suck.
A lot of consumer broadband connections have upload caps that will make them almost certainly unusable for nethack. If your local DSL provider or cable company hasn't set a cap though I'd think they'd be sufficent.
How many film cameras these days still work if they run out of power? Other than some of the very cheap point-n-shoot varieties and some of low end SLR bodies not too many. Have you looked a CF cards? 128MB or 256MB models are reasonably priced (
If you are in the middle of nowherezistan then it probably makes sense to carry around film, but I'd consider that the exception rather than the rule.
Heh, I guess it was too obvious that I pulled that URL directly from the ports tree. FreeBSDers who really want to see the game in action (although I suspect many people will have the above sentiment: This used to think this was fun?!?) can:
# cd/usr/ports/games/xrobotsAdjust if your ports tree is elsewhere of course # make install clean
Or you can play the text version (it's part of the base BSD games): % cd/usr/games %./robots
Or you can install the gnome versions (not recommended unless you already have gnome installed, no point installing all of gnome for a couple of lousy games after all): # cd/usr/ports/games/gnomegames # make install clean
This should be enough robot on robot action for anybody.
Gah. This will teach me to check my links before sending them. The above link is dead (probably has been for a long time). xrobots can be downloaded from the FreeBSD project ftp site. It should compile on just about any platform with X. Gnome also ships with a version called "gnobots" and an enhanced version called "gnobots2".
You play on a simple field (often times an 80x25 terminal window). Your intreped hero has no weapons (sometimes he has a single shot sonic screwdriver) and if any robot touches him he dies. You can only kill the robots by making them run into each other (where they'll leave piles of debris behind--and running into debris is also deadly for the robots!). The robots are very stupid, they always head in a bee-line to the player (at least as much as they can being constrained to moving in ordinal directions). Your hero's primary weapon is a teleporter that randomly teleports him to somewhere else in the playing field (including occasionally next to a robot). The game is over when a robot touches the player (usually when you teleport in right next to a robot). A somewhat feature lacking version can be compiled from here: Hpux
One thing that has been bothering me for some time is the amount of energy spent fighting convection in a standard PC case. Just look how often your primary exaust is a power supply fan--where the intake is about 2/3 of the way up in the case. Any air above the intake fan (HOT air) has to be drawn back down into the exaust fan or stagnate. I suspect that in most cases that hot air just stagnates (right up where you CD burner, DVD and possibly even hard drives sit).
Sorry for not including a small ascii art diagram that would have made the entire layout very obvious, but the lameness filter wouldn't even let a very dumbed down version of it through. Instead I'll try to spell it out here: (Ironically this is much lamer than just including a simple ASCII art diagram).
The case has the power supply mounted in the normal location, however it is mounted such that the power cord connecter is on the top and the intake vents point downward. In front of the power supply (lining the top of the case) are the drive bays. The top of the case is an open mesh to allow air to escape easily. CD-ROM trays will open upwards (like the Apple Cube) and the floppies will drop in from the top. Each drive will have at least half an inch of space on either side of it to allow air to flow around it. The case will have three (possibly more or less) fans mounted in the bottom blowing upwards. The bottom of the case will be on raised legs that allow air to be pulled in from underneath the case. The motherboard will be mounted normally (since case manufacturers can't really do anything about it). Ideally though you would find some way to mount the PCI and AGP cards vertically.
The rest of this post is an attempt to explain why I think this will work, and an attempt to avoid the lame lameness filter (why can't people who have good records, with say 35 or 40+ karma, get around the lameness filter?).
Anyway, my primary intent with this case design is it reduce the turbulance in the case and to use the natural tendancy for heat to rise in my favor. The input fans at the bottom of the case should keep the entire case at a slightly positive pressure. One thing you must do with this case is sit it a couple of inches off of the ground espeically if the "ground" here is shag carpeting. The drives should not be as close together as they are in a standard PC case, and they should allow air to flow freely between them. In these days of 10k and 15k RPM drives cooling your HDs is perhaps one of the most often overlooked aspects if case design. The only big problem I have with this case is the PCI slots. PCI slots are generally too close together for my liking and they are almost invariably mounted horizontally, guarenteeing that any hot card will create a hot spot on the card above it. Unfortunatly there is little a case manufacturer can do about this so I'm leaving it as a caveat. Having the power supply mounted vertically will mean that the power cord will attach to the top of your computer. I recommend either a specialy modified power supply or mounting the power supply horizontally (with the intake repositioned to the bottom of the power supply) and lengthing the case somewhat or simply leaving the power cords on the top of the case. A crafty case manufacturer might even create a little box for the ends of the power cords on the top of the case that will conceal them from general view. If you need more 3 1/2 bays, you can run them down the front of the case (in front of the motherboard) vertically. The final caveat with this configuration is that your users must remember to never stick objects on the top of this case (especially things that can spill, like coffee cups). I recommend making the top rounded or triangular or some other shape unsuited for sitting things on (but remember to leave access for things like CD-ROMs and floppy drives!).
A similar construction (albiet with more specialized hardware) to this was already used for that Apple Cube (and look how good the cooling was on that, no fan needed!) but it seems like PC manufacturuers still havn't got it. Look at practially any professional server and you'll see similar concepts in play (I'm definatly not claiming I invented any of this) almost exclusivly. Nothing here is new in the slightest and yet nearly every PC case manufacturer insists on the same general layout and same general poor quality construction. Sometimes it feels like the only thing cases are manufactured for is low cost, and all other considerations are secondary. It is this attitude of cutting every corner possible that leads to the air circulation nightmare we have in almost all modern cases. I also believe that high heat lowers the life of PC components, be it through shrinking and expanding or just plain mild constant overheating. This goes double for devices like hard drives which have actual mechanical components and thin layers of oil to worry about.
Have you ever actually used the command line in Win2K? It's horrible. Everything has a space in it (space is generally reserved as a delimiter character on the command line!) and you end up having to put quotes around practically ever CD command. Even worse, if you have your prompt set up to display the current directory it doesn't take very long for your prompt to become long enough to start wrapping on almost every command. I consider most directory names in Win2K to be needlessly verbose personally, and I'd love to see them trim some of the directory names down a bit. I can't be the only person who gets annoyed at humongous paths like: C:\Documents and Settings\Administrator\Start Menu\Programs\Microsoft Office Too
ls>
Yeah, but make sure it's one of the surge protectors with the anti-gravity unit built in, just a regular surge protector probably won't prevent a two foot fall from damaging a Nomad...especially if the Nomad hits it instead of the carpeted floor.
Um, Xosl is configurable from the boot screen (IIRC, it's been a year since I used it last).
Lets see what I can remember about this.
It is very nice looking, and even has optional fading effets
It is easy to install
It comes with a partition manager (Ranish Partition Manager)
It's quite configurable, you can do pretty much anything with it that you can do with any other first stage boot loader
It requires a FAT parition on your hard drive. This is bad for people who want to dual boot between Linux/FreeBSD and Win2k, and doubly bad if your FAT partition gets corrupted.
It tends to mark things "unavailable" if they disappear temporarily and never brings them back (you have to delete and recreate the partition). This was a big annoyance when I had a flaky SCSI card.
Ranish Parition Manager is not exactly pretty or easy to use
It's not so good for systems with fixed frequncy monitors, fortunatly this isn't a big deal anymore, but I used to have a 1280x1024 ONLY monitor attached to my system (not even a text mode) and my video card's best VESA mode was 800x600.
Despite what the docs say, you pretty much need a mouse to use it. I was never able to get the keyboard shortcuts working correctly for the configuration screens.
That's pretty much all I remember about it... I hope that gives you and idea of what Xosl is like.
When someone asks me, "What is the best computer?" I always give them the same response:
The one you build yourself. If you build it yourself you open up whole new worlds of possilibites that most OEMs never even consider. Want to use high quality Power Supplies, Motherboards, Cases, etc...? No problem. Want an all IEEE 1394 PC? No problem? Want a super nice sound card and a low end/no video card? Hey, you can do that.
Even better is how these days most connections are keyed (especially on the high quality stuff) and most components ship with sane defaults (cable select on IDE devices for instance) so you can just plug it all in and go. Generally I find I tend to pay a bit more for a machine I build myself, but that's because I tend to buy the more expensive parts and avoid cutting corners. I figure it's much cheaper than replacing that ultra-cheap noname HD 6 months down the road.
Re:Tiny operating systems
on
Tiny Apps
·
· Score: 1
Or from personal experiance:
Redhat Disk On Chip support fixit floppy. I had a heck of a time getting Redhat (7.0) working correctly with these M-Systems chips we had. The problem was getting just enough files on the small DOC to get RH bootable, but not so many that we couldn't fit what we needed. It was quite a challenge because the installer was designed for 6.2 and I didn't have a lot of RH expertise at the time (I would have much preferred a different distro, but that's what we had to use). Anyway, FreeBSD's picobsd support is very customizable, you specify exactly what you want in your kernel and userland (though easy to edit text files) and run build, a few minutes later it spits out a floppy with everything you need. What's even more impressive is how much you can include on a single floppy, I had most of/bin a good bit of/sbin, a kernel that had networking, DOC, ext2fs, and various other support, and I still had room left over for a few useful utilties.
I know lots of people who already have trouble keeping track of all of their CDs. I can only imagine what troubles they'll have if the media is only the size of a quarter.
The only big wins I see with this technology are
Portable players. Imagine a matchbox sized unit that holds a full album worth of music. Portable CD players long ago reached the point where the size of the CD is the limiting factor in how small they can be made.
Massive storage units. If you could put these CDs in "rolls" or some other method, you can store a whole lot of them in a standard sized consumer audio unit, as opposed to the 5 or 10 CD changers that are common now.
And that's about it. For just about everything else a regular CD is just better. The consumer-hostile content control is just the icing on the cake IMHO.
Er, that analogy doesn't make much sense. That would be like running back to the store and stealing a couple extra copies of WinXP, then installing them on seperate partitions on your HD, which really doesn't make sense. The license is more like allowing your wife to drive the car you bought from MS, even though when you bought the car the dealer told you (in the fine print) that if she wanted to drive she'd have to come back and buy another car for her instead. But then these car analogies are dumb because when you buy a car you actually own the thing, as opposed to software were you mearly buy the permission from the company to use it in some limited fashion. Now the analogy is more like you rent the car from the dealer and he tells you that you must not let anyone else drive it becaue that will cut into his rental business.
Man, why is it when companies build in NICs on motherboards they always choose the crappiest one they can find? Bill Paul has some choice words to say about this card (taken from if_rl.c in the FreeBSD source tree).
/*
* The RealTek 8139 PCI NIC redefines the meaning of 'low end.' This is
* probably the worst PCI ethernet controller ever made, with the possible
* exception of the FEAST chip made by SMC. The 8139 supports bus-master
* DMA, but it has a terrible interface that nullifies any performance
* gains that bus-master DMA usually offers.
*
* For transmission, the chip offers a series of four TX descriptor
* registers. Each transmit frame must be in a contiguous buffer, aligned
* on a longword (32-bit) boundary. This means we almost always have to
* do mbuf copies in order to transmit a frame, except in the unlikely
* case where a) the packet fits into a single mbuf, and b) the packet
* is 32-bit aligned within the mbuf's data area. The presence of only
* four descriptor registers means that we can never have more than four
* packets queued for transmission at any one time.
*
* Reception is not much better. The driver has to allocate a single large
* buffer area (up to 64K in size) into which the chip will DMA received
* frames. Because we don't know where within this region received packets
* will begin or end, we have no choice but to copy data from the buffer
* area into mbufs in order to pass the packets up to the higher protocol
* levels.
*
* It's impossible given this rotten design to really achieve decent
* performance at 100Mbps, unless you happen to have a 400Mhz PII or
* some equally overmuscled CPU to drive it.
*
* On the bright side, the 8139 does have a built-in PHY, although
* rather than using an MDIO serial interface like most other NICs, the
* PHY registers are directly accessible through the 8139's register
* space. The 8139 supports autonegotiation, as well as a 64-bit multicast
* filter.
*
* The 8129 chip is an older version of the 8139 that uses an external PHY
* chip. The 8129 has a serial MDIO interface for accessing the MII where
* the 8139 lets you directly access the on-board PHY registers. We need
* to select which interface to use depending on the chip type.
*/
The worst part is, it's not that expensive to build decent 10/100 chips these days. NetGear and LinkSys sell decent cards for as little as $5 a pop. There's really no reason to go with the RealTeks anymore.
The DTLA is one of the affected drives, the IC35L040 is one of the new 60GXPs, I've heard mixed reports on them, but so far they don't appear to be as bad as the 75GXPs. Also, you have command tagged queueing turned on, for some reason this seems to reduce the failure rate (at least on our local machines; we used to get these failures every couple of months on the drives, but we havn't seen it since turning on CTQ).
That's what they want you to think, but copyright law specifically allows you to make backups, so long as you do not give those backups (or the originals!) away. MP3s themselves are not illegal, however they may contain illegaly copied copywritten content.
In general, take any legal advice from RIAA or the MPAA with a grain of salt. They tend to subtly twist the truth to whatever makes them the most money with little regard for weather it's within the spirit (or sometimes the letter) of the law.
Well, when I tested data over Iridium the latency was around 1 second round trip. Pretty good for satellite communications but not what poeple these days are really expecting. Also, the data rate was 2400 baud and the connection was pretty lossy (voice calls were only tolerable on the phones). Iridium is not a solution for something like a wireless lan. Besides, you can't buy airtime on the constellation as an individual anymore (and it's way more expensive than you'd want anyway)
It's fairly slow to read, although much better than the writes
It tends to wear out after only a few tens of thousands of writes. Even the fancy new adaptors that spread writes out across the entire memory space get bitten by this
It's more expensive than RAM (quite a bit more currently, but that may be an economy of scale).
Most of them use PIO0 for access (at least the ones I've seen, some of them may support DMA, but I've never seen them). This means your processor has to spend a lot of time handling disk reads and writes. This is purely an engineering problem at the moment that would go away if anybody really tried to sell these as HD replacements, but it is still a problem for people using them today.
Slashdot Mirror
One thing that is overlooked in rackmount (especially 1U) servers is power drain. Computer centers don't have unlimited power (in fact since you have to both power the equiptment AND power the refrigeration units to draw the heat from your machines out of the CC the power requirements of your servers becomes something of an issue). The ability to run dozens of servers in a rack without creating a power situtation is a big bonus for computing centers.
One of the big trends in computing recently has been for servers to grow smaller and consume lots more power. Just look at your average P4 or Athlon compared to the old Pentiums and Pentium Pros. It's getting so bad that PCs are drawing more power than even energy hogging Alpha based machines in some cases. This problem is compunded by everybody putting servers in the smallest boxes prossible, 1U frequently, so that these energy hogs can be stacked all the way to the top of the rack and draw an enormous amount of power from the circut designated for that tile.
Actually, if you read Eric's Commentary you'll discover that he didn't "win". CRC won. CRC screwed Eric hard, and didn't even have the common courtesy to give him a reach around. At least the website will be back up for awhile until CRC decides to issue another printing of the book and shut them down again.
Yeah, Niven will seem soo passe once we start building Ringworlds or Dyson Spheres, and he'll look like a 50's pulp fiction author once we find some actual living smoke rings.
I think Niven will still be read years from now because his ideas are so far out there that they will still be exciting (if completely unachievable) to the people of the future.
Granted some of his stuff probably won't survive, but I don't think there has been any major author throughout history that has churned out nothing but classics, including any of the fine fellows on this list.
Amen to that. I always hated going to the paintball field and discovering that the other team is a bunch of tournament jerks. For some reason these guys try to bend/break every rule they think they can get away with, and couldn't care less if they suck all of the fun right out of the game.
Most flagrantly, if they were hit somewhere out of site of the refs (almost all of the field) they'd just walk behind the next wall and start playing again. Or if they were seen they'd walk around the field and point out the other team members to their buds (or take potshots when nobody was looking).
Sometimes they'd get caught, but usually there weren't nearly enough refs to handle those jerks. The worst part is when they get all arrogant and cocky about taking only one or two losses (in games of 12 on 12 or so) and spend all the between-game time telling you how much you suck.
So do you want to use Gnumeric because it IS Excel, or because it ISN'T Excel?
Could you set up the local accounts in a jail?
A lot of consumer broadband connections have upload caps that will make them almost certainly unusable for nethack. If your local DSL provider or cable company hasn't set a cap though I'd think they'd be sufficent.
How many film cameras these days still work if they run out of power? Other than some of the very cheap point-n-shoot varieties and some of low end SLR bodies not too many. Have you looked a CF cards? 128MB or 256MB models are reasonably priced (
If you are in the middle of nowherezistan then it probably makes sense to carry around film, but I'd consider that the exception rather than the rule.
Heh, I guess it was too obvious that I pulled that URL directly from the ports tree. FreeBSDers who really want to see the game in action (although I suspect many people will have the above sentiment: This used to think this was fun?!?) can: # cd /usr/ports/games/xrobots Adjust if your ports tree is elsewhere of course
/usr/games ./robots
/usr/ports/games/gnomegames
# make install clean
Or you can play the text version (it's part of the base BSD games):
% cd
%
Or you can install the gnome versions (not recommended unless you already have gnome installed, no point installing all of gnome for a couple of lousy games after all):
# cd
# make install clean
This should be enough robot on robot action for anybody.
Gah. This will teach me to check my links before sending them. The above link is dead (probably has been for a long time). xrobots can be downloaded from the FreeBSD project ftp site. It should compile on just about any platform with X. Gnome also ships with a version called "gnobots" and an enhanced version called "gnobots2".
You play on a simple field (often times an 80x25 terminal window). Your intreped hero has no weapons (sometimes he has a single shot sonic screwdriver) and if any robot touches him he dies. You can only kill the robots by making them run into each other (where they'll leave piles of debris behind--and running into debris is also deadly for the robots!). The robots are very stupid, they always head in a bee-line to the player (at least as much as they can being constrained to moving in ordinal directions). Your hero's primary weapon is a teleporter that randomly teleports him to somewhere else in the playing field (including occasionally next to a robot). The game is over when a robot touches the player (usually when you teleport in right next to a robot). A somewhat feature lacking version can be compiled from here:
Hpux
One thing that has been bothering me for some time is the amount of energy spent fighting convection in a standard PC case. Just look how often your primary exaust is a power supply fan--where the intake is about 2/3 of the way up in the case. Any air above the intake fan (HOT air) has to be drawn back down into the exaust fan or stagnate. I suspect that in most cases that hot air just stagnates (right up where you CD burner, DVD and possibly even hard drives sit).
Sorry for not including a small ascii art diagram that would have made the entire layout very obvious, but the lameness filter wouldn't even let a very dumbed down version of it through. Instead I'll try to spell it out here: (Ironically this is much lamer than just including a simple ASCII art diagram).
The case has the power supply mounted in the normal location, however it is mounted such that the power cord connecter is on the top and the intake vents point downward. In front of the power supply (lining the top of the case) are the drive bays. The top of the case is an open mesh to allow air to escape easily. CD-ROM trays will open upwards (like the Apple Cube) and the floppies will drop in from the top. Each drive will have at least half an inch of space on either side of it to allow air to flow around it. The case will have three (possibly more or less) fans mounted in the bottom blowing upwards. The bottom of the case will be on raised legs that allow air to be pulled in from underneath the case. The motherboard will be mounted normally (since case manufacturers can't really do anything about it). Ideally though you would find some way to mount the PCI and AGP cards vertically.
The rest of this post is an attempt to explain why I think this will work, and an attempt to avoid the lame lameness filter (why can't people who have good records, with say 35 or 40+ karma, get around the lameness filter?).
Anyway, my primary intent with this case design is it reduce the turbulance in the case and to use the natural tendancy for heat to rise in my favor. The input fans at the bottom of the case should keep the entire case at a slightly positive pressure. One thing you must do with this case is sit it a couple of inches off of the ground espeically if the "ground" here is shag carpeting. The drives should not be as close together as they are in a standard PC case, and they should allow air to flow freely between them. In these days of 10k and 15k RPM drives cooling your HDs is perhaps one of the most often overlooked aspects if case design. The only big problem I have with this case is the PCI slots. PCI slots are generally too close together for my liking and they are almost invariably mounted horizontally, guarenteeing that any hot card will create a hot spot on the card above it. Unfortunatly there is little a case manufacturer can do about this so I'm leaving it as a caveat. Having the power supply mounted vertically will mean that the power cord will attach to the top of your computer. I recommend either a specialy modified power supply or mounting the power supply horizontally (with the intake repositioned to the bottom of the power supply) and lengthing the case somewhat or simply leaving the power cords on the top of the case. A crafty case manufacturer might even create a little box for the ends of the power cords on the top of the case that will conceal them from general view. If you need more 3 1/2 bays, you can run them down the front of the case (in front of the motherboard) vertically. The final caveat with this configuration is that your users must remember to never stick objects on the top of this case (especially things that can spill, like coffee cups). I recommend making the top rounded or triangular or some other shape unsuited for sitting things on (but remember to leave access for things like CD-ROMs and floppy drives!).
A similar construction (albiet with more specialized hardware) to this was already used for that Apple Cube (and look how good the cooling was on that, no fan needed!) but it seems like PC manufacturuers still havn't got it. Look at practially any professional server and you'll see similar concepts in play (I'm definatly not claiming I invented any of this) almost exclusivly. Nothing here is new in the slightest and yet nearly every PC case manufacturer insists on the same general layout and same general poor quality construction. Sometimes it feels like the only thing cases are manufactured for is low cost, and all other considerations are secondary. It is this attitude of cutting every corner possible that leads to the air circulation nightmare we have in almost all modern cases. I also believe that high heat lowers the life of PC components, be it through shrinking and expanding or just plain mild constant overheating. This goes double for devices like hard drives which have actual mechanical components and thin layers of oil to worry about.
Have you ever actually used the command line in Win2K? It's horrible. Everything has a space in it (space is generally reserved as a delimiter character on the command line!) and you end up having to put quotes around practically ever CD command. Even worse, if you have your prompt set up to display the current directory it doesn't take very long for your prompt to become long enough to start wrapping on almost every command. I consider most directory names in Win2K to be needlessly verbose personally, and I'd love to see them trim some of the directory names down a bit. I can't be the only person who gets annoyed at humongous paths like:
C:\Documents and Settings\Administrator\Start Menu\Programs\Microsoft Office Too
ls>
Yeah, but make sure it's one of the surge protectors with the anti-gravity unit built in, just a regular surge protector probably won't prevent a two foot fall from damaging a Nomad...especially if the Nomad hits it instead of the carpeted floor.
Lets see what I can remember about this.
That's pretty much all I remember about it... I hope that gives you and idea of what Xosl is like.
When someone asks me, "What is the best computer?" I always give them the same response:
The one you build yourself. If you build it yourself you open up whole new worlds of possilibites that most OEMs never even consider. Want to use high quality Power Supplies, Motherboards, Cases, etc...? No problem. Want an all IEEE 1394 PC? No problem? Want a super nice sound card and a low end/no video card? Hey, you can do that.
Even better is how these days most connections are keyed (especially on the high quality stuff) and most components ship with sane defaults (cable select on IDE devices for instance) so you can just plug it all in and go. Generally I find I tend to pay a bit more for a machine I build myself, but that's because I tend to buy the more expensive parts and avoid cutting corners. I figure it's much cheaper than replacing that ultra-cheap noname HD 6 months down the road.
Or from personal experiance: /bin a good bit of /sbin, a kernel that had networking, DOC, ext2fs, and various other support, and I still had room left over for a few useful utilties.
Redhat Disk On Chip support fixit floppy. I had a heck of a time getting Redhat (7.0) working correctly with these M-Systems chips we had. The problem was getting just enough files on the small DOC to get RH bootable, but not so many that we couldn't fit what we needed. It was quite a challenge because the installer was designed for 6.2 and I didn't have a lot of RH expertise at the time (I would have much preferred a different distro, but that's what we had to use). Anyway, FreeBSD's picobsd support is very customizable, you specify exactly what you want in your kernel and userland (though easy to edit text files) and run build, a few minutes later it spits out a floppy with everything you need. What's even more impressive is how much you can include on a single floppy, I had most of
The only big wins I see with this technology are
- Portable players. Imagine a matchbox sized unit that holds a full album worth of music. Portable CD players long ago reached the point where the size of the CD is the limiting factor in how small they can be made.
- Massive storage units. If you could put these CDs in "rolls" or some other method, you can store a whole lot of them in a standard sized consumer audio unit, as opposed to the 5 or 10 CD changers that are common now.
And that's about it. For just about everything else a regular CD is just better. The consumer-hostile content control is just the icing on the cake IMHO.Er, that analogy doesn't make much sense. That would be like running back to the store and stealing a couple extra copies of WinXP, then installing them on seperate partitions on your HD, which really doesn't make sense. The license is more like allowing your wife to drive the car you bought from MS, even though when you bought the car the dealer told you (in the fine print) that if she wanted to drive she'd have to come back and buy another car for her instead. But then these car analogies are dumb because when you buy a car you actually own the thing, as opposed to software were you mearly buy the permission from the company to use it in some limited fashion. Now the analogy is more like you rent the car from the dealer and he tells you that you must not let anyone else drive it becaue that will cut into his rental business.
On board Realtek 8139C
/*
Man, why is it when companies build in NICs on motherboards they always choose the crappiest one they can find? Bill Paul has some choice words to say about this card (taken from if_rl.c in the FreeBSD source tree).
* The RealTek 8139 PCI NIC redefines the meaning of 'low end.' This is
* probably the worst PCI ethernet controller ever made, with the possible
* exception of the FEAST chip made by SMC. The 8139 supports bus-master
* DMA, but it has a terrible interface that nullifies any performance
* gains that bus-master DMA usually offers.
*
* For transmission, the chip offers a series of four TX descriptor
* registers. Each transmit frame must be in a contiguous buffer, aligned
* on a longword (32-bit) boundary. This means we almost always have to
* do mbuf copies in order to transmit a frame, except in the unlikely
* case where a) the packet fits into a single mbuf, and b) the packet
* is 32-bit aligned within the mbuf's data area. The presence of only
* four descriptor registers means that we can never have more than four
* packets queued for transmission at any one time.
*
* Reception is not much better. The driver has to allocate a single large
* buffer area (up to 64K in size) into which the chip will DMA received
* frames. Because we don't know where within this region received packets
* will begin or end, we have no choice but to copy data from the buffer
* area into mbufs in order to pass the packets up to the higher protocol
* levels.
*
* It's impossible given this rotten design to really achieve decent
* performance at 100Mbps, unless you happen to have a 400Mhz PII or
* some equally overmuscled CPU to drive it.
*
* On the bright side, the 8139 does have a built-in PHY, although
* rather than using an MDIO serial interface like most other NICs, the
* PHY registers are directly accessible through the 8139's register
* space. The 8139 supports autonegotiation, as well as a 64-bit multicast
* filter.
*
* The 8129 chip is an older version of the 8139 that uses an external PHY
* chip. The 8129 has a serial MDIO interface for accessing the MII where
* the 8139 lets you directly access the on-board PHY registers. We need
* to select which interface to use depending on the chip type.
*/
The worst part is, it's not that expensive to build decent 10/100 chips these days. NetGear and LinkSys sell decent cards for as little as $5 a pop. There's really no reason to go with the RealTeks anymore.
The DTLA is one of the affected drives, the IC35L040 is one of the new 60GXPs, I've heard mixed reports on them, but so far they don't appear to be as bad as the 75GXPs. Also, you have command tagged queueing turned on, for some reason this seems to reduce the failure rate (at least on our local machines; we used to get these failures every couple of months on the drives, but we havn't seen it since turning on CTQ).
Because they're cheap and you can create IDE RAID arrays for a fraction of the price of any other solution.
Also, they're cheap.
That's what they want you to think, but copyright law specifically allows you to make backups, so long as you do not give those backups (or the originals!) away. MP3s themselves are not illegal, however they may contain illegaly copied copywritten content.
In general, take any legal advice from RIAA or the MPAA with a grain of salt. They tend to subtly twist the truth to whatever makes them the most money with little regard for weather it's within the spirit (or sometimes the letter) of the law.
Well, when I tested data over Iridium the latency was around 1 second round trip. Pretty good for satellite communications but not what poeple these days are really expecting. Also, the data rate was 2400 baud and the connection was pretty lossy (voice calls were only tolerable on the phones). Iridium is not a solution for something like a wireless lan. Besides, you can't buy airtime on the constellation as an individual anymore (and it's way more expensive than you'd want anyway)
I hope this was helpful.