A lot of mid-high to high end projectors designed for home theater support 1080p, but note that (I beleve I am correct on this but it's possible I misuderstood something), you'll see some broadcasts in 1080i, but never in 1080p because there is not enough bandwidth allocated to any HDTV channel to support a 1080p broadcast. Perhaps once we get past the digital drop-dead date (yeah right), we'll have some satellite providers helping out with beautiful 1080p feeds.
The main reason for me is unlinks (deletes) on XFS are dog slow. A file deletion happens whenever a message is retrieved from a maildir (most pop3 clients issue a DELE after each message).. It adds up when you have a few hundred (or thousand) users connecting concurrently since it means that people have to be connected longer. The biggest advantage to reiser in a maildir setup, though, is the sector tailing. Messages are routinely a lot smaller than the sector size of a disk. You can save tens of gigs easily on a large mailserver with reiserfs.
As others have pointed out, this 32 bit PCI card pretty much sucks. It outperforms, what, a single U160 drive on a 32 bit controller? By only a factor of two?!?
You could easily smash the performance with a little U160 RAID on a 64 bit pci controller, and perhaps even with a single drive, though seek time would hurt a bit. At least it'd cost you a hell of a lot less.
It's not battery backed either, which is pretty useless for anything this might need to do. Heck, without battery backup, it proabbly can't even survive a reboot to get that precious data back after a system crash!
This reeks of an EE or Embedded Systems course assignment. It's barely a real product.
Good christ, you'd think that by the time you outgrew a QUAD XEON mailserver with only 5000 users, you'd have been reevaluating performance before plunking down what must have been close to 10-15 grand or more at the time on a second one!
Your mailbox format is all wrong. Storing all messages in a single file is pretty much the worst way to do anything useful. You want to explore some alternative storage format such as mbx or maildir. I personally use maildir on ReiserFS on Linux and have good luck. (The filesystem is VERY important for maildirs. ReiserFS's block tail support and directory indexing give it major disk space and speed advantages for a maildir mailserver application, while running something like maildirs on XFS would instantly kill your server. I hear mbx is pretty good too, if you're stuck on some sort of standard filesystem since it uses indexing and fewer files than maildir. The downside is that it's not as immediately parseable as maildir or mbox... Ie you couldnt write a script to say... delete extremely high scoring spam messages from any user who hasn't checked their mail in over 3 months, or other things ISP's might routinely do to maintain their servers.
Finally, if you plan to scale way up there (60,000+), you need to start looking at better cluster systems than just a couple machines. Specialize the tasks of several machines to do mail storage or talk POP3. Look at something like POPular for specialized POP3 server clustering software.
Yes, red LED's are cheaper, but there is another reason, too.. They also want to use the cheapest CCD available. That's going to be a monochrome CCD that's sensitive to larger wavelengths. A red LED is going to work better on that.
With that in mind, an Infrared LED would probably work great with optical mice and their cheap CCD's.. maybe even better than red. You might have to remove an IR filter from in front of the CCD, and be wary of using them in a room with flourescent lighting, but it'd be good to try. How cool is an optical mouse with apparently NO light?!?!
If you want the "cool" blue look for whatever twisted ass reason, just use a blue LED and an IR LED in paralell. You might have to play with different led's/led voltages to get the right balance between a responsive mouse and the cool blue glow your riceboy heart desires, but again, it should work.
Look, as everybody else has said, if someone's going to steal your car, they're going to steal it. If they're going to break into it, they're going to break into it.
To get into the car, the thief has several options: Intercept, record and duplicate the keyless entry's signal; jimmy the door lock by yanking the lock wire or an interior switch (very hard on most new cars, but still possible on a great many); go after the keyed lock on the door itself with a pick/what have you; or break the window. Anyway, the point is that even if the thief has all the gear to get into your car via the keyless entry system, it's not going to be very effective to use versus any of the other methods. It's the same with robbing houses. Too many people leave the door between the garage and the house unlocked, and the garage doors operate on a (generally much less secure than car) wireless RF remote.... But you don't see burglers breaking into houses via the garage door very often. It's just easier for them to crowbar through the back door or find a house with an open window.
Anyway, what it boiled down to for me was that: 1) pushing a button to unlock the car is a hell of a lot quicker/easier than turning a key. 2) I am an idiot and forget where I parked a lot. I can make the car beep at me, and it helps. 3) Having a car alarm of some sort brings down my car insurance a lot... i guess mainly because of my age and the car i drive.. but no matter what it will usually knock your rate down a little bit.
So anyway, I'd say go for it, but make sure that the system has some sort of alarm feature such as a glass break mic or ignition cutoff that would classify it as a tiny bit more than a remote lock/unlock button, or it's not going to help you with your insurance.
Yeah, I know. I have fooled with the Win32 compile of GAIM off and on. It's still very dodgy, though Gimp/win32 is pretty stable -- there are some other GTK apps that feel OK on windows too, but which ones elude me...
Anyway, nothing against GTK, I like it fine and I enjoy writing Perl/GTK very much, but QT does just have that good native widget feel and is stable on all architectures, even os x! TT does a good job.
What you have done here is fix your problem without realizing what it was. Don't worry, though, the Higher-Ups will probably buy your grounding/capacitance bullshit anyway if you have to justify the downtime to them. It sounds impressive.
Motherboard replacement involves handling, reseating, and resocketing every single component in the system, except sometimes drives. Anyway, here are the problems you might actually have been having:
1) Grounding something that wasn't supposed to be grounded: If you didnt get the new board on exactly square, or you accidentally left a metal riser where the old board needed one but the new board didn't, you might have been grounding something that shouldn't have been. How many times did you take the new board out and check before changing the risers?
2) Card seating. Your Asus board might have had PCI/AGP slots manufactured by a different company than your old motherboard. Perhaps the 2mm difference made a huge difference in the way cards seated in them (The AGP slot would be a likely contender to exhibit weird behavior on a 2mm height difference since it has two rows of pins.) This happened sometimes to me on EISA boards where I would not get a card seated all the way (You had to push HARD) This explination seems the most likely
3) Regular gremlins. Reseating everything: processor, cards, cables, etc. can sometimes do wonders for a system that's causing you to bang your head on the wall. If it just so happened that whatever device was giving you the problems was twiddled in just the right fashon when you switched the risers, then it would seem only natural that the risers fixed the problem, only they probably just forced you to unplug everything looking for it, and as a side effect you got some contact to scrape up against some peice of metal just so..
If by "almost identical" you mean "pales by comparison" then I'll take your word for it. In fact, the only features trillian has above gaim are pretty useless - skinning and MDI window capabilities. I guess you might include translucent windows, but who in the fuck uses that useless piece of crap? Until the *background* can be translucent, the *text and borders* solid, and buttons fade toward opaque, window translucency will be nothing more than a stupid hack.
Anyway, that being said, Gaim: 1) suppports more protocols and plugin authors at least have the ability to write support for protocols it doesnt support 2) is scriptable via perl and 3) is monetarily free keeping it out of that shitty grey area where cerullian studios makes money on a product that would not function without the continued work of aol/yahoo/msn, some of whom, at least, lose revenue when their users don't use their own clients.
Anyway, trillian sucks. The SDK sucks, it wouldn't be worth using if someone wrote a Win32 native GUI for GAIM. Too bad GAIM wasn't started on Qt...
Ditto that case. I picked up the steel version nearly three years ago when Antec started making (or OEM branding) them.. Sturdy, roomy, and very easy to work with. I have about 15 of them now, and haven't really found a better substitute.
Note that I have also found that the aluminum versions of this case, although very light, are absolute shit for shipping. They fall apart. Trust your shipping needs to steel!
Actually, the first thing I'm going to point the finger at is the use of a 16 bit WiFi PC Card rather than a CardBus one. That'd probably fix you up right quick. Still, the WinME can't help, but really, CardBus is to PCMCIA as PCI is to ISA (well, not exactly, but from the end user perspective the pitfalls are about the same)
Yeah, if the page is a script that gives out different content based on some parameter, you could easily do this. I would imagine that the module lets you *configure it*.. Gee, imagine being able to change a parameter?!?!
I suppose this will mean that in order to get LSB 1.2 certified, your software must run on the listed architectures?
Would an application vendor seeking LSB certification be required to release versions of software that worked on the different architectures? IMO, that would be a good thing in an ideal world, but in the real world, it's going to hinder adoption of the LSB because there are going to be companies that it is just not cost effective to develop, maintain, and support their software on multiple architectures. I don't think that's a good idea at all.
While that is true, there are now open source encoders and decoders for those older (3.11) DiVX formats. It is presumable that someone could create decoding hardware for the 3.x format files based upon these known algorithms.
By "roaming" all they really mean is that they are forcing you to pay inflated prices because you are in a captive market when you are on an airplane. How do they propose to block the normal frequencies without active jamming? If they are using active jamming, then why would it be unsafe to use normal phones anyway? Smells like a scam to me. Creating more *reliable* service with a new range of in-air cellular frequencies is one thing, but doing it under the auspices of safety like this is pretty tounge-in-cheek.
All phones look pretty much the same... Something you hold to your ear alongside your face. How do they plan on identifying "normal" phones from the phones that can be used in-flight unless they are in some way actively blocking the other frequencies from use?
Anyway, the only really truly nice thing that can come from this is that whatever technology they develop to shoehorn people into paying inflated prices for cellular service while they sit in an airplane cabin could likely be adopted to things like movie theaters - your phone rings in the movies, for instance, and you owe the theater's private cellular network $5 and $2/min while you blabber. It would make people think twice before they allow their obnoxiousness to annoy everyone else, but still provide for emergency use and whatnot.
While they are at it with adding some cell standards, they should also allow for some sort of device that would force a cell phone that is in-range of the device to vibrate, go silent (if it can't vibrate) - or at least switch to the lowest ring volume. Now *that* would be a nice idea.
I love wireless as much as the next guy (probably more since it has made me quite a bit of money), but if you are talking a 200ft run across the parking lot (probably your own parking lot, right?), then there is a LOT to be said for running a little piece of conduit with a couple pieces of fiber in it. Your workers will likely not be happy running VoIP and the network over consumer grade 802.11b equipment, not to mention the headaches that you will face with QoS or interfacing the IP phones into your existing system.
Unless you plan on replacing your entire phone system, you're probably going to have problems interfacing the IP phones with your existing stuff, and even if you can do that without trouble, you're going to get really stuck when you put it all on that 802.11b link and start having problems. By the time you get enough *good* wireless hardware and routers to handle the link properly, you'll have spent an awful lot of money. Also, as others have mentioned you'll have problems meeting fire and safety codes without land-lines over there anyway, and building-to-building calls would be free with local phone service anyway.
Anyway, my main point is that in the long run, it will not be cost effective to bridge the parking lot with wireless and VoIP versus a buried fiber "extension cord" type setup. If you were a mile away or running across a major street or something, it would be highly cost effective.
If you decide to go for it anyway, here's a couple of suggestions:
1) Don't think VoIP phones. Think VoIP bridges. A couple of channel banks on either end of whatever network link you have should allow you to interface remote phones directly with your PBX. Contact your vendor about it. Hint: Don't mention wireless; say ethernet and fiber instead. Adtran has some inexpensive stuff in this arena.
2) Don't think 802.11b. It's not going to cut it. Get something along the line of the WyLAN or Proxim equipment that's designed to be a point-to-point ethernet bridge. You'll get performance almost an order of magnitude higher than 802.11b with a something like the 12mb proxim system vs 11mb 802.11b. This is a very important consideration. Do not ignore this suggestion.
3) Put real routers at either end of your bridge. Do not rely on crap 802.11b linksys network bridges or something similar. You will need real routers to do the very important job of QoS. If you have voice running over the network, they MUST be able to properly prioritize all traffic. All of the inexpensive consumer wireless products that advertise QoS will likely be unsuitable for your applicaiton. Get some cisco equipment on either end and be happy.
Their IDE-RAID is actually software RAID. The SCSI myth can go off the shelf, sure, but don't take the RAID myth down.
The promise FastTrak and Highpoint and a few others are not actually hardware RAID controllers. They are regular controlers with enough firmware to allow BIOS calls to do drive access via software RAID (located in the firmware of the controller), and OS drivers that implement the company's own software RAID implementation at the driver level, thereby doing things like making only one device appear to the OS. Some of the chips have some performance improvements over a purely software RAID solutions, such as the ability to do data comparisons between two drives in a mirror during reads, but that's about it. If you ever boot them into a new install of windows without preloading their "drivers", guess what? Your "RAID" of 4 drives is just 4 drives. The hardware recovery options they have are also pretty damned worthless when it comes to a comparison with real RAID controllers - be they IDE or SCSI.
A good solution to the IDE RAID debacle are the controllers by 3Ware (very fine) or the Adaptec AAA series controllers (also pretty fine). These are real hardware controllers with onboard cache, hardware XOR acceleration for RAID 5 and the whole bit.
Anyway, I'm not really all that taken aback that this webserver is floundering a bit, but seems really responsive when the page request "gets through," so to speak. If it's not running low on physical RAM, it's probably got a lot of processes stuck in D state due to the shit promise controller. A nice RAID controller would probably have everything the disks are thrashing on in a RAM cache at this point.
You forgot at least one step. Pick one to add but not both:
4.5. Just because we're using a mac webserver, doesn't mean we're free from the responsibility of properly tuning our configuration. Anyone can buy a box of any type that's preconfigured to run apache when you first plug it in. Anyway, we tune the heck out of our Apache so that it will stand up to the load we're expecting.
or
7. Wonder what is going wrong when we realize we have no grasp of how our computer or applications actually work.
Yes. I think for the size it could not do that. If an accelerometer were mounted at the top of a very long antenna or mast of some sort to sense the forward/backward tilting, then I believe it could, but I don't think that you're going to get enough data to extrapolate a tilt angle if you mount one on even the very top of the mindstorm brick.
Now, a *very very* sensitive 2d accelerometer with a very very high sample rate might be able to do it, but that's probably a very ineffective method.
I don't know where the reply came from, since it doesn't really apply to my post, but accelerometers will work for what I am talking about - allowing the robot to determine after a fall if it has landed on its front or back (two wheels in contact with the ground) or landed on its side where only one wheel can contact the ground.
It probably has less to do with incompetent idiots cracking the cpu dye because they can't put on a heatsink properly and more to do with the problems of heatsinks falling off during shipping. It's happened to two systems I have sent recently (to the DC area interestingly enough - i guess the FedEx and UPS guys are moving with a bit more haste these days) Trying to explain how to properly install a heatsink to someone over the telephone is *very* difficult, and since most must be installed with some kind of thermal compound, this is a really really large problem.
Large computer manufacturers usually use an external baffle/clip system to hold the heatsinks in place so they won't fall off during shipping.
Anyway, I'm glad for AMD's move to a bolt-down CPU heatsink. I've been missing the luxury of those secure slot processors...
After you spend all that time to get the stickers off of it, you might want to consider that if you want to buy the same DVD cases, you can get them for 10 to 30 cents each...
Slashdot Mirror
A lot of mid-high to high end projectors designed for home theater support 1080p, but note that (I beleve I am correct on this but it's possible I misuderstood something), you'll see some broadcasts in 1080i, but never in 1080p because there is not enough bandwidth allocated to any HDTV channel to support a 1080p broadcast. Perhaps once we get past the digital drop-dead date (yeah right), we'll have some satellite providers helping out with beautiful 1080p feeds.
The main reason for me is unlinks (deletes) on XFS are dog slow. A file deletion happens whenever a message is retrieved from a maildir (most pop3 clients issue a DELE after each message).. It adds up when you have a few hundred (or thousand) users connecting concurrently since it means that people have to be connected longer. The biggest advantage to reiser in a maildir setup, though, is the sector tailing. Messages are routinely a lot smaller than the sector size of a disk. You can save tens of gigs easily on a large mailserver with reiserfs.
As others have pointed out, this 32 bit PCI card pretty much sucks. It outperforms, what, a single U160 drive on a 32 bit controller? By only a factor of two?!?
You could easily smash the performance with a little U160 RAID on a 64 bit pci controller, and perhaps even with a single drive, though seek time would hurt a bit. At least it'd cost you a hell of a lot less.
It's not battery backed either, which is pretty useless for anything this might need to do. Heck, without battery backup, it proabbly can't even survive a reboot to get that precious data back after a system crash!
This reeks of an EE or Embedded Systems course assignment. It's barely a real product.
Good christ, you'd think that by the time you outgrew a QUAD XEON mailserver with only 5000 users, you'd have been reevaluating performance before plunking down what must have been close to 10-15 grand or more at the time on a second one!
Your mailbox format is all wrong. Storing all messages in a single file is pretty much the worst way to do anything useful. You want to explore some alternative storage format such as mbx or maildir. I personally use maildir on ReiserFS on Linux and have good luck. (The filesystem is VERY important for maildirs. ReiserFS's block tail support and directory indexing give it major disk space and speed advantages for a maildir mailserver application, while running something like maildirs on XFS would instantly kill your server. I hear mbx is pretty good too, if you're stuck on some sort of standard filesystem since it uses indexing and fewer files than maildir. The downside is that it's not as immediately parseable as maildir or mbox... Ie you couldnt write a script to say... delete extremely high scoring spam messages from any user who hasn't checked their mail in over 3 months, or other things ISP's might routinely do to maintain their servers.
Finally, if you plan to scale way up there (60,000+), you need to start looking at better cluster systems than just a couple machines. Specialize the tasks of several machines to do mail storage or talk POP3. Look at something like POPular for specialized POP3 server clustering software.
~GoRK
What, no mention of Blogdex, the Media Diffusion Index? It's several years old, and is pretty good at picking current trends.
~GoRK
Yes, red LED's are cheaper, but there is another reason, too.. They also want to use the cheapest CCD available. That's going to be a monochrome CCD that's sensitive to larger wavelengths. A red LED is going to work better on that.
With that in mind, an Infrared LED would probably work great with optical mice and their cheap CCD's.. maybe even better than red. You might have to remove an IR filter from in front of the CCD, and be wary of using them in a room with flourescent lighting, but it'd be good to try. How cool is an optical mouse with apparently NO light?!?!
If you want the "cool" blue look for whatever twisted ass reason, just use a blue LED and an IR LED in paralell. You might have to play with different led's/led voltages to get the right balance between a responsive mouse and the cool blue glow your riceboy heart desires, but again, it should work.
~GoRK
Look, as everybody else has said, if someone's going to steal your car, they're going to steal it. If they're going to break into it, they're going to break into it.
... But you don't see burglers breaking into houses via the garage door very often. It's just easier for them to crowbar through the back door or find a house with an open window.
.. but no matter what it will usually knock your rate down a little bit.
To get into the car, the thief has several options: Intercept, record and duplicate the keyless entry's signal; jimmy the door lock by yanking the lock wire or an interior switch (very hard on most new cars, but still possible on a great many); go after the keyed lock on the door itself with a pick/what have you; or break the window. Anyway, the point is that even if the thief has all the gear to get into your car via the keyless entry system, it's not going to be very effective to use versus any of the other methods. It's the same with robbing houses. Too many people leave the door between the garage and the house unlocked, and the garage doors operate on a (generally much less secure than car) wireless RF remote.
Anyway, what it boiled down to for me was that:
1) pushing a button to unlock the car is a hell of a lot quicker/easier than turning a key.
2) I am an idiot and forget where I parked a lot. I can make the car beep at me, and it helps.
3) Having a car alarm of some sort brings down my car insurance a lot... i guess mainly because of my age and the car i drive
So anyway, I'd say go for it, but make sure that the system has some sort of alarm feature such as a glass break mic or ignition cutoff that would classify it as a tiny bit more than a remote lock/unlock button, or it's not going to help you with your insurance.
~GoRK
Yeah, I know. I have fooled with the Win32 compile of GAIM off and on. It's still very dodgy, though Gimp/win32 is pretty stable -- there are some other GTK apps that feel OK on windows too, but which ones elude me...
Anyway, nothing against GTK, I like it fine and I enjoy writing Perl/GTK very much, but QT does just have that good native widget feel and is stable on all architectures, even os x! TT does a good job.
What you have done here is fix your problem without realizing what it was. Don't worry, though, the Higher-Ups will probably buy your grounding/capacitance bullshit anyway if you have to justify the downtime to them. It sounds impressive.
Motherboard replacement involves handling, reseating, and resocketing every single component in the system, except sometimes drives. Anyway, here are the problems you might actually have been having:
1) Grounding something that wasn't supposed to be grounded: If you didnt get the new board on exactly square, or you accidentally left a metal riser where the old board needed one but the new board didn't, you might have been grounding something that shouldn't have been. How many times did you take the new board out and check before changing the risers?
2) Card seating. Your Asus board might have had PCI/AGP slots manufactured by a different company than your old motherboard. Perhaps the 2mm difference made a huge difference in the way cards seated in them (The AGP slot would be a likely contender to exhibit weird behavior on a 2mm height difference since it has two rows of pins.) This happened sometimes to me on EISA boards where I would not get a card seated all the way (You had to push HARD) This explination seems the most likely
3) Regular gremlins. Reseating everything: processor, cards, cables, etc. can sometimes do wonders for a system that's causing you to bang your head on the wall. If it just so happened that whatever device was giving you the problems was twiddled in just the right fashon when you switched the risers, then it would seem only natural that the risers fixed the problem, only they probably just forced you to unplug everything looking for it, and as a side effect you got some contact to scrape up against some peice of metal just so..
Hope that answers your question.
~GoRK
If by "almost identical" you mean "pales by comparison" then I'll take your word for it. In fact, the only features trillian has above gaim are pretty useless - skinning and MDI window capabilities. I guess you might include translucent windows, but who in the fuck uses that useless piece of crap? Until the *background* can be translucent, the *text and borders* solid, and buttons fade toward opaque, window translucency will be nothing more than a stupid hack.
Anyway, that being said, Gaim: 1) suppports more protocols and plugin authors at least have the ability to write support for protocols it doesnt support 2) is scriptable via perl and 3) is monetarily free keeping it out of that shitty grey area where cerullian studios makes money on a product that would not function without the continued work of aol/yahoo/msn, some of whom, at least, lose revenue when their users don't use their own clients.
Anyway, trillian sucks. The SDK sucks, it wouldn't be worth using if someone wrote a Win32 native GUI for GAIM. Too bad GAIM wasn't started on Qt...
Ditto that case. I picked up the steel version nearly three years ago when Antec started making (or OEM branding) them.. Sturdy, roomy, and very easy to work with. I have about 15 of them now, and haven't really found a better substitute.
Note that I have also found that the aluminum versions of this case, although very light, are absolute shit for shipping. They fall apart. Trust your shipping needs to steel!
~GoRK
Actually, the first thing I'm going to point the finger at is the use of a 16 bit WiFi PC Card rather than a CardBus one. That'd probably fix you up right quick. Still, the WinME can't help, but really, CardBus is to PCMCIA as PCI is to ISA (well, not exactly, but from the end user perspective the pitfalls are about the same)
Yeah, if the page is a script that gives out different content based on some parameter, you could easily do this. I would imagine that the module lets you *configure it*.. Gee, imagine being able to change a parameter?!?!
~GoRK
ffmpeg supports encoding and decoding wmv1 and (i think) decoding wmv2)
I suppose this will mean that in order to get LSB 1.2 certified, your software must run on the listed architectures?
Would an application vendor seeking LSB certification be required to release versions of software that worked on the different architectures? IMO, that would be a good thing in an ideal world, but in the real world, it's going to hinder adoption of the LSB because there are going to be companies that it is just not cost effective to develop, maintain, and support their software on multiple architectures. I don't think that's a good idea at all.
~GoRK
While that is true, there are now open source encoders and decoders for those older (3.11) DiVX formats. It is presumable that someone could create decoding hardware for the 3.x format files based upon these known algorithms.
~GoRK
By "roaming" all they really mean is that they are forcing you to pay inflated prices because you are in a captive market when you are on an airplane. How do they propose to block the normal frequencies without active jamming? If they are using active jamming, then why would it be unsafe to use normal phones anyway? Smells like a scam to me. Creating more *reliable* service with a new range of in-air cellular frequencies is one thing, but doing it under the auspices of safety like this is pretty tounge-in-cheek.
All phones look pretty much the same... Something you hold to your ear alongside your face. How do they plan on identifying "normal" phones from the phones that can be used in-flight unless they are in some way actively blocking the other frequencies from use?
Anyway, the only really truly nice thing that can come from this is that whatever technology they develop to shoehorn people into paying inflated prices for cellular service while they sit in an airplane cabin could likely be adopted to things like movie theaters - your phone rings in the movies, for instance, and you owe the theater's private cellular network $5 and $2/min while you blabber. It would make people think twice before they allow their obnoxiousness to annoy everyone else, but still provide for emergency use and whatnot.
While they are at it with adding some cell standards, they should also allow for some sort of device that would force a cell phone that is in-range of the device to vibrate, go silent (if it can't vibrate) - or at least switch to the lowest ring volume. Now *that* would be a nice idea.
~GoRK
I love wireless as much as the next guy (probably more since it has made me quite a bit of money), but if you are talking a 200ft run across the parking lot (probably your own parking lot, right?), then there is a LOT to be said for running a little piece of conduit with a couple pieces of fiber in it. Your workers will likely not be happy running VoIP and the network over consumer grade 802.11b equipment, not to mention the headaches that you will face with QoS or interfacing the IP phones into your existing system.
Unless you plan on replacing your entire phone system, you're probably going to have problems interfacing the IP phones with your existing stuff, and even if you can do that without trouble, you're going to get really stuck when you put it all on that 802.11b link and start having problems. By the time you get enough *good* wireless hardware and routers to handle the link properly, you'll have spent an awful lot of money. Also, as others have mentioned you'll have problems meeting fire and safety codes without land-lines over there anyway, and building-to-building calls would be free with local phone service anyway.
Anyway, my main point is that in the long run, it will not be cost effective to bridge the parking lot with wireless and VoIP versus a buried fiber "extension cord" type setup. If you were a mile away or running across a major street or something, it would be highly cost effective.
If you decide to go for it anyway, here's a couple of suggestions:
1) Don't think VoIP phones. Think VoIP bridges. A couple of channel banks on either end of whatever network link you have should allow you to interface remote phones directly with your PBX. Contact your vendor about it. Hint: Don't mention wireless; say ethernet and fiber instead. Adtran has some inexpensive stuff in this arena.
2) Don't think 802.11b. It's not going to cut it. Get something along the line of the WyLAN or Proxim equipment that's designed to be a point-to-point ethernet bridge. You'll get performance almost an order of magnitude higher than 802.11b with a something like the 12mb proxim system vs 11mb 802.11b. This is a very important consideration. Do not ignore this suggestion.
3) Put real routers at either end of your bridge. Do not rely on crap 802.11b linksys network bridges or something similar. You will need real routers to do the very important job of QoS. If you have voice running over the network, they MUST be able to properly prioritize all traffic. All of the inexpensive consumer wireless products that advertise QoS will likely be unsuitable for your applicaiton. Get some cisco equipment on either end and be happy.
~GoRK
Did it go, like, "Beep boop beep beep?"
Their IDE-RAID is actually software RAID. The SCSI myth can go off the shelf, sure, but don't take the RAID myth down.
The promise FastTrak and Highpoint and a few others are not actually hardware RAID controllers. They are regular controlers with enough firmware to allow BIOS calls to do drive access via software RAID (located in the firmware of the controller), and OS drivers that implement the company's own software RAID implementation at the driver level, thereby doing things like making only one device appear to the OS. Some of the chips have some performance improvements over a purely software RAID solutions, such as the ability to do data comparisons between two drives in a mirror during reads, but that's about it. If you ever boot them into a new install of windows without preloading their "drivers", guess what? Your "RAID" of 4 drives is just 4 drives. The hardware recovery options they have are also pretty damned worthless when it comes to a comparison with real RAID controllers - be they IDE or SCSI.
A good solution to the IDE RAID debacle are the controllers by 3Ware (very fine) or the Adaptec AAA series controllers (also pretty fine). These are real hardware controllers with onboard cache, hardware XOR acceleration for RAID 5 and the whole bit.
Anyway, I'm not really all that taken aback that this webserver is floundering a bit, but seems really responsive when the page request "gets through," so to speak. If it's not running low on physical RAM, it's probably got a lot of processes stuck in D state due to the shit promise controller. A nice RAID controller would probably have everything the disks are thrashing on in a RAM cache at this point.
~GoRK
You forgot at least one step. Pick one to add but not both:
4.5. Just because we're using a mac webserver, doesn't mean we're free from the responsibility of properly tuning our configuration. Anyone can buy a box of any type that's preconfigured to run apache when you first plug it in. Anyway, we tune the heck out of our Apache so that it will stand up to the load we're expecting.
or
7. Wonder what is going wrong when we realize we have no grasp of how our computer or applications actually work.
~GoRK
Yes. I think for the size it could not do that. If an accelerometer were mounted at the top of a very long antenna or mast of some sort to sense the forward/backward tilting, then I believe it could, but I don't think that you're going to get enough data to extrapolate a tilt angle if you mount one on even the very top of the mindstorm brick.
Now, a *very very* sensitive 2d accelerometer with a very very high sample rate might be able to do it, but that's probably a very ineffective method.
~GoRK
I don't know where the reply came from, since it doesn't really apply to my post, but accelerometers will work for what I am talking about - allowing the robot to determine after a fall if it has landed on its front or back (two wheels in contact with the ground) or landed on its side where only one wheel can contact the ground.
It probably has less to do with incompetent idiots cracking the cpu dye because they can't put on a heatsink properly and more to do with the problems of heatsinks falling off during shipping. It's happened to two systems I have sent recently (to the DC area interestingly enough - i guess the FedEx and UPS guys are moving with a bit more haste these days) Trying to explain how to properly install a heatsink to someone over the telephone is *very* difficult, and since most must be installed with some kind of thermal compound, this is a really really large problem.
Large computer manufacturers usually use an external baffle/clip system to hold the heatsinks in place so they won't fall off during shipping.
Anyway, I'm glad for AMD's move to a bolt-down CPU heatsink. I've been missing the luxury of those secure slot processors...
~GoRK
After you spend all that time to get the stickers off of it, you might want to consider that if you want to buy the same DVD cases, you can get them for 10 to 30 cents each...