Well, that's sort of my point. If you just state the facts, simply, then the fact that you know what the hell you are talking about becomes clear.
If you start out with all the "I have $CERTIFICATION and I know what I am talking about", the general response is "Yeah, right!"
If you simply start out saying "Well, things started getting slow, so I pinged the gateway and got some pretty variable timings and about 10% loss", then the 2TS on the line gets the picture that you may actually know which end of the keyboard you type on.
Yes, what people forget is that level 1 service people are usually little better than meatware text ->speech algorithms. If your text->speech algorithm started saying anything other than what you told it to say, you would replace it.
Just lie, play along with the script, and wait to get to the level 2 people. Then, when you get to the level 2 people, DON'T act all cocksure - DON'T try to come across as an "expert". From a L2 tech support person's perspective, a person who is trying to act like an "expert" is usually a moron who knows just little enough to be difficult.
Just report, factually, what you are seeing. Don't try to conjecture, don't offer opinions. Just the facts, m'am - "I do a traceroute and after foo.bar.bax I get no response.", not "the machine after foo.bar.bax is down".
The best, quietest media is a hard disk in another room.
If you are in a studio apartment, then put the server into a large box. Underclock the main CPU. Cover the interior of the large box with foam. Put a big, slow fan in the box, and draw air in from the outside through a plenum that also has foam lining the walls, and makes at least one turn, so that there is no direct line of sight from the fan to the outside. Vent through a large, foam covered hole in the top.
Presto - silent box with as much storage as you want.
(Note - when I say "foam", I mean a low density open cell foam - something porous, like the foam they sell for use in speaker cabinets, NOT insulation panels or Great Stuff)
For the cost of a decent sized NVRAM disk, or Flash, or anything like that, you could create a HUGE RAID-5 array and still be whisper quiet.
An evil trick to play on a telemarketer just crossed my mind - pity I don't get more calls from them.
Set up a computer to record the sound from the phone line, delay it about 250 msec, and play it back on the phone line. Make this switchable, so that you can turn it on at will.
Telemarketer calls. Switch on delay when they speak, off when you speak.
That much delay will make just about anybody stutter like mad.
Go look again at the releases for BeOS. At the time BeOS was released, it could play DVDs, while Linux could not, because BeOS could have a licensed CSS decoder.
And read my post again. Yes, Linux *can* use closed source drivers, but tends not to, because of the problems with linking them against the kernel.
Simple - BeOS is able to use closed source drivers, Linux drivers tend to be open source.
Due to the design of the Centrino chipset, Intel asserts they are unable to release the specs for the chip, or source for drivers for the chip, and maintain FCC type certification. For part 15 certification, the device cannot operate outside specified limits on frequency and power. For a closed source driver, Intel is able to insure that is the case. For open source drivers, the user could alter frequency and power outsiee the limits imposed by the FCC.
This was always the case for BeOS - it could play files, access hardware, and in general do more than a free system could, because Be could sign NDAs and the Free development community could not.
This is obviously the MPAA's new copy protection scheme - if 18 minutes is 3 terabytes, then NOBODY's boing to be able to copy this.
It will work just as well as their previous schemes - i.e. not at all, as people reduce the rez to something meaningful.
/joke
Seriously, this is something I've wondered about for IMAX/Omnimax style theaters - if they could go to a 60 Hz or better refresh rate it would really help on the long pans and flyover sequences, but since the screen is so large (or more precisely since the screen subtends such a large angle of the viewer's vision) you need a boatload of pixels to avoid the "pixels as big as your fist" effect.
And sure, it might at first seem difficult to have the 10 terabytes per 40 minute IMAX movie, but I have.2 terabyte in a single US$220 Firewire drive bay right now - setting up 100 TB of RAID5 for a theater vs. having those huge IMAX movie rolls might not be a bad thing.
Of course you'd need one HELL of a DMD projector to make this work, but.... Moore's law marches on.
Consider how big a flap this outage is causing. Consider how many people feel their whole life has been turned upside down.
Now, ask yourself how many of those people's lives REALLY have been turned upside down.
For a small set of the population, having mobile communications is critical. But that set is MUCH smaller than the set of people who THINK mobile communications are critical. Folks, there are answering machines with remote playback and pay telephones. There is even the idea of WAITING - that this conversation can take place LATER.
I was on a business trip a while back. I was asked by one of our Marketing directors what my cell number was. "I don't have a cell." He was shocked. "I don't need one. When I am not traveling, I can make all the personal calls I want on the local autopatches. Business calls can damn well wait till I am in the office. When I am travelling on vacation, the only calls I need to make are to hotels to book a room, and those are toll free and I can use a payphone at a gas station. When I am traveling on business the company can damn well loan a phone to me."
I'm not a Luddite - quite the contrary, I help design test equipment for cell phone. I know too well what the systems look like. That is one of the reasons I don't have a phone.
For $DEITY's sake folks, unplug once in a while - you will find out that you live quite well without the phone!
Item: IBM has accused SCO of violating copyright. Item: DMCA allows a copyright holder to demand an infringing website be shut down, by the ISP if needed, within 24 hours IIRC.
Scenario: 36 hours before SCO's next conference call, IBM makes a DMCA demand. Thus, during the conference call, SCO's web site is down.
Question: what effect would that have on the investors in SCO?
Believing that giving more money to the government will reduce deficits is like believing that buying an alcoholic another drink will slake his thirst.
The only real way to solve this problem is to put measureable, non-revokable penalties on government officials who overspend - for example, by saying that Congress shall not be paid, nor accrue retirement benefits, during any year in which the government runs a deficit (and a deficit shall be defined simply as "spending more money than you took in", no more funny accounting tricks).
We must be able to run a deficit during times of crisis (think World War II), but there needs to be a strong disincentive to prevent perpetual crisis.
I wish I could believe that Valve was reacting to the negative press about Steam and DRM, and was busily reworking things to remove the dependancy upon Steam in HL2.
I wish I could believe that Valve was working with Icculus.org to make HL2 for Linux.
Unfortunately, I don't. I beleive Valve IS working on Steam, but not to make HL2 less dependant upon it, but rather more so, so that no matter how bad Steam is, you will be FORCED to use it to play HL2.
Oh well, as George Thourogood said, "that don' confront me", as I don't plan on downgrading my OS to play HL2. I'll just wait for Doom3, and play that instead.
This all is coming about because of what I consider to be a design flaw in DNS, to whit:
For any given suffix.foo.bar.baz, there is one and only one possible dataset to answer that query.
So if you are looking up bar.com, there is only one dataset that contains information on.com - thus whosoever controls that dataset controls.com.
Now, what if a server for a given domain, in addition to having a parent had siblings? For example, if you were looking up narf.com, then the queries might look like this: my machine - Hey root, where's narf.com? Root - I don't know, but verisign.com should - ask him. My machine - Hey verisign.com, where's narf.com? Verisign.com - I don't know, maybe alternic.com does. My machine - Hey alternic.com, where's narf.com? alternic.com - narf.com is at 192.168.0.1
In other words, for every zone record there would be a new configuration possible - a list of zero or more siblings. On a negative result, the sibling records would be returned, and the quering name server would consult them. As a result, you could allow joker.com, register.com, verisign.com et. al. to have just their records on their servers, with cross links to the other servers.
Yes, this would increase the number of queries a name server might have to do to resolve a domain, especially in the negative case (domain does not exist). However, just as a name server can and will cache the servers for.com now, a name server could cache the list of name servers for.com now, and could send the queries out in parallel to the servers it knows about, reducing the time.
Under this system, a failure of verisign's server would not black out.com, just the set of domains registered with.com. It would reduce (somewhat) the size of the servers needed to serve the.com domain (as the workload would be spread out among more servers), and would allow for each registrar to maintain their own database without having to go though Verisign.
Extending this to the root servers would allow for things like Alternic to be added - the root servers could say "I don't recognize.foo, but maybe Alternic does".
Yes, it would be possible for Joker.com and Register.com to create records for VerisignEngineersAreWeenies.com, and for those records to disagree. Yes, the set of owners of servers for the.com domain would have to do their homework in registering a domain - there would have to be a clearing house for domain registry.
But were this idea implemented, it would prevent anybody from pulling the kind of unilateral crap that Verisign has.
So, the next time a *BSDer starts bitching about how the GPL is not "free", does this mean we can point out the hypocrisy of criticizing those who provide you with the single most important tool needed to enjoy the environment of your choice?
There are some researchers who beleive they've found the mechanism for this.
When the organism is stressed by lack of food, genes that encode heat shock proteins are activated. HSPs are used when a cell is overtemp, or otherwise stressed, to repair damage to the DNA due to the stress.
The thinking is this: an organism is getting too few calories. The cells start making HSPs due to the stress. The HSPs soak up free radicals, as well as repairing DNA damage. Since the lack of calories is not causing undo damage to the DNA (unlike heat), the net result is more damage due to other environmental effects (radiation, replication errors, toxins) is undone.
In short, the organism's metabolism set to allow it to survive beyond the "famine" to maximize the chances of being able to reproduce once food is available.
The researchers have some good candidate genes for the proteins, and perhaps one day may be able to stimulate the production of these proteins without the need to starve ourselves.
Now, whether the world needs a bunch of long-lived , fat, self-indulgant slobs is another question for which many of the residents of this forum are curiously well-equipped to argue.
OK, this whole thing confuses me, and I am an embedded systems engineer. I've seen this in the trade journals, and I don't get the marketing forces this is supposed to answer.
Consider this: my old laptop that I purchased in 1996 had 3 PCMCIA slots. This was good - I could have my NIC, my SCSI card (for tape backups), and my modem all in place at the same time.
However, any laptop of recent vintage will have at least USB2.0 High speed if not IEEE-1394 (FireWire), so this obviates the need for the SCSI card. It will have built-in Ethernet (at LEAST 10/100 MBits, if not 10/100/1000!), so there goes the Ethernet card. It will have a build in (Win)Modem, removing the need for the modem (at least for Windows users, and very likely for Linux users as well now-a-days).
So what is left for the PCMCIA slots? Flash readers? Built-in, or USB. Video capture? (like you need that in a laptop anyway, but....) Firewire. Video acceleration (MPEG decoding)? Faster CPU. 3D acceleration? Built in.
I can see using PCI-Express (the PROPER name for the new, high-speed serial interface) for the docking station interface - but even then, what do you really need to add to a laptop now?
So what is the point of a PCCARD style interface? OK, I may not be able to get a Firewire tape backup device (or maybe I can - I haven't looked since I don't need one), but if I want to back up a new laptop I can use the network or just dump everything to a Firewire drive.
Now, some may say "Yes, but what about embedded devices". And I can say, as a professional, "What about them?" Either what I am building is a small, simple device, where I would rather build in a USB 2.0 host adaptor, or it is a big, hairy multi-CPU monster that has what it needs built-in. Really, in neither situation would I want to go to the difficulty of adding a PCMCIA-style interface. Been there, done that, and had far too many headaches with people expecting to install J. Random Card and have it work. Sorry, but unless you are using Embedded Windows, you cannot just install the driver disk and go. And if you are using a Windows deriviative, you DON'T WANT users installing their own software (unless you really like watching Customer Service drown).
Again, unless we start seeing laptops with their video on a card, PCMCIA style interfaces are no longer the best engineering decision. Let them die.
Go back and look over this story. There, one before, we had a story on Slashdot about a software vendor not supporting Linux. At that time, I made a very reasonable suggestion - write to the company involved and ask for a Linux port. I also asked folks who had done so to comment in my Journal, so as to have a public record of the number of letters so written so that when the company involved said "We've never had any requests for this" we could trivially disprove the claim.
And what came of it? Nothing. cat/dev/null.
Why did Valve release a Linux server for Halflife? Because the community innundated them with requests for it.
So you want a HalfLife2 for Linux? Innundate Valve with requests! Stop bitching on Slashdot, and write them a physical, paper and toner letter requesting a HalfLife2 client for Linux.
Now, as for the whole "Just suck it up and run Windows" crowd, and the whole "Fuck Windows - Linux or Nothing" crowd: Each of us must make a decision what is more important - running the OS we choose, or playing a game. And you know what? That decision is going to be different for different people - imagine that!
If you are willing to put up with Windows to be able to run Halflife 2, then by all means do so, have fun, and SHUT THE FUCK UP!
If you will 'live free or die', and refuse to run Windows in order to run Halflife 2, then great! Welcome to the fold, accept the consequences of your decision, and SHUT THE FUCK UP (on Slashdot, that is)! Bitching on Slashdot won't change things, writing a letter to Valve just might!
Or if you don't want to write Valve, then help out on the Wine DirectX layer (and yes, I actually DO have contributions in the Wine source tree.)
But whatever your choices are, accept that they come with consequences, and STOP BITCHING ABOUT THEM WHERE IT WILL DO NO GOOD!
When you go to college, you already have to make several purchases (books and such). Why not require students to buy one of the current crop of broadband routers (e.g. Linksys) that provide a layer of firewalling?
Granted, this won't protect the student's computer from local exploits - the worst security exploit of any computer lies between the keyboard and the chair. But for Blaster/Nimda, this would provide another layer of protection - ideally, enough to allow a student to connect their uninfected machine to the network and download the required patches.
In fact, were the routers available in the bookstore, already configured with the appropriate policies by the university, that would greatly simplify the deployment - as the student is buying their books they could also buy a router.
I got some bad news for you, sunshine - not only is Pink not well, but you are listed in a lot more public databases than you might think.
Case in point: I got a spam at my work email, the title of which was:
my.work.address REFINANCE YOUR MORTGAGE ON my home mailing address.
Now, let me explain that at no time is my work address directly linked to my home mailing address, which is a P.O. Box. However, my work email address is of the form firstname.lastname@companyname.example, and I recently bought a house. The fact that I bought this house is a matter of public record, and my name, while hardly unique, is rare enough that a spammer could have cross-referenced it with a list of emails, filtered by the location of the company, and got a hit.
Database mining at work.
However, with respect to the entries in the FCC callsign database - while you must provide the FCC with the location of your primary station (a.k.a. your home for most hams), you do not have to have your home address listed on the database - for example, all you would get were you to look up myh callsign would be my mailing address - the self-same P.O. Box I mentioned before.
The long and the short of it is, if you are worried about it, then get a P.O. Box (or a Mailboxes Etc. box), and use that for all your addresses possible. Then, just like having a spamtrap address, you can decouple information from your physical address. If you are truly paranoid, get 2 mail boxes - one for business, one for "junk". This also has the advantage that should you move within the same area, you needn't fill out change-of-address forms for all your magazines and such.
Oh, BTW - if you subscribe to any magazines, you are probably listed on their databases as well, and guess what - they tend to sell those as well.
It's truly hard to be a blank nowadays.
(Bonus credit for readers - work out what my callsign is - I've given it before....)
Slashdot Mirror
Well, that's sort of my point. If you just state the facts, simply, then the fact that you know what the hell you are talking about becomes clear.
If you start out with all the "I have $CERTIFICATION and I know what I am talking about", the general response is "Yeah, right!"
If you simply start out saying "Well, things started getting slow, so I pinged the gateway and got some pretty variable timings and about 10% loss", then the 2TS on the line gets the picture that you may actually know which end of the keyboard you type on.
Yes, what people forget is that level 1 service people are usually little better than meatware text ->speech algorithms. If your text->speech algorithm started saying anything other than what you told it to say, you would replace it.
Just lie, play along with the script, and wait to get to the level 2 people. Then, when you get to the level 2 people, DON'T act all cocksure - DON'T try to come across as an "expert". From a L2 tech support person's perspective, a person who is trying to act like an "expert" is usually a moron who knows just little enough to be difficult.
Just report, factually, what you are seeing. Don't try to conjecture, don't offer opinions. Just the facts, m'am - "I do a traceroute and after foo.bar.bax I get no response.", not "the machine after foo.bar.bax is down".
The best, quietest media is a hard disk in another room.
If you are in a studio apartment, then put the server into a large box. Underclock the main CPU. Cover the interior of the large box with foam. Put a big, slow fan in the box, and draw air in from the outside through a plenum that also has foam lining the walls, and makes at least one turn, so that there is no direct line of sight from the fan to the outside. Vent through a large, foam covered hole in the top.
Presto - silent box with as much storage as you want.
(Note - when I say "foam", I mean a low density open cell foam - something porous, like the foam they sell for use in speaker cabinets, NOT insulation panels or Great Stuff)
For the cost of a decent sized NVRAM disk, or Flash, or anything like that, you could create a HUGE RAID-5 array and still be whisper quiet.
An evil trick to play on a telemarketer just crossed my mind - pity I don't get more calls from them.
Set up a computer to record the sound from the phone line, delay it about 250 msec, and play it back on the phone line. Make this switchable, so that you can turn it on at will.
Telemarketer calls. Switch on delay when they speak, off when you speak.
That much delay will make just about anybody stutter like mad.
Go look again at the releases for BeOS. At the time BeOS was released, it could play DVDs, while Linux could not, because BeOS could have a licensed CSS decoder.
And read my post again. Yes, Linux *can* use closed source drivers, but tends not to, because of the problems with linking them against the kernel.
Simple - BeOS is able to use closed source drivers, Linux drivers tend to be open source.
Due to the design of the Centrino chipset, Intel asserts they are unable to release the specs for the chip, or source for drivers for the chip, and maintain FCC type certification. For part 15 certification, the device cannot operate outside specified limits on frequency and power. For a closed source driver, Intel is able to insure that is the case. For open source drivers, the user could alter frequency and power outsiee the limits imposed by the FCC.
This was always the case for BeOS - it could play files, access hardware, and in general do more than a free system could, because Be could sign NDAs and the Free development community could not.
This is obviously the MPAA's new copy protection scheme - if 18 minutes is 3 terabytes, then NOBODY's boing to be able to copy this.
/joke
.2 terabyte in a single US$220 Firewire drive bay right now - setting up 100 TB of RAID5 for a theater vs. having those huge IMAX movie rolls might not be a bad thing.
It will work just as well as their previous schemes - i.e. not at all, as people reduce the rez to something meaningful.
Seriously, this is something I've wondered about for IMAX/Omnimax style theaters - if they could go to a 60 Hz or better refresh rate it would really help on the long pans and flyover sequences, but since the screen is so large (or more precisely since the screen subtends such a large angle of the viewer's vision) you need a boatload of pixels to avoid the "pixels as big as your fist" effect.
And sure, it might at first seem difficult to have the 10 terabytes per 40 minute IMAX movie, but I have
Of course you'd need one HELL of a DMD projector to make this work, but.... Moore's law marches on.
Consider how big a flap this outage is causing. Consider how many people feel their whole life has been turned upside down.
Now, ask yourself how many of those people's lives REALLY have been turned upside down.
For a small set of the population, having mobile communications is critical. But that set is MUCH smaller than the set of people who THINK mobile communications are critical. Folks, there are answering machines with remote playback and pay telephones. There is even the idea of WAITING - that this conversation can take place LATER.
I was on a business trip a while back. I was asked by one of our Marketing directors what my cell number was. "I don't have a cell." He was shocked. "I don't need one. When I am not traveling, I can make all the personal calls I want on the local autopatches. Business calls can damn well wait till I am in the office. When I am travelling on vacation, the only calls I need to make are to hotels to book a room, and those are toll free and I can use a payphone at a gas station. When I am traveling on business the company can damn well loan a phone to me."
I'm not a Luddite - quite the contrary, I help design test equipment for cell phone. I know too well what the systems look like. That is one of the reasons I don't have a phone.
For $DEITY's sake folks, unplug once in a while - you will find out that you live quite well without the phone!
OK, but what is the RFC for this physical layer?
Is anybody working on a bridge between TCP over bongos and TCP over avian carriers?
Which has the greater bandwidth?
Wasn't "Wireless USB" pretty much what Bluetooth was supposed to be?
You had FireWire, which was to "move lots of time sensitive freight".
You had USB, which was to "connect everything with wires".
You had BlueTooth, which was "Connect everything without wires".
That's what I thought as well when I saw the Cypress Semi ad for their "Wireless USB" chip.
Good, now they can start sequencing this dog's genome....
Submitted for your approvial:
Item: IBM has accused SCO of violating copyright.
Item: DMCA allows a copyright holder to demand an infringing website be shut down, by the ISP if needed, within 24 hours IIRC.
Scenario: 36 hours before SCO's next conference call, IBM makes a DMCA demand. Thus, during the conference call, SCO's web site is down.
Question: what effect would that have on the investors in SCO?
Really, Slashdot posters should learn to read the posts they are responding to before posting their criticisms.
Had you read my post, you would have seen that I very clearly stated the need to prevent the government from manufacturing crises.
Believing that giving more money to the government will reduce deficits is like believing that buying an alcoholic another drink will slake his thirst.
The only real way to solve this problem is to put measureable, non-revokable penalties on government officials who overspend - for example, by saying that Congress shall not be paid, nor accrue retirement benefits, during any year in which the government runs a deficit (and a deficit shall be defined simply as "spending more money than you took in", no more funny accounting tricks).
We must be able to run a deficit during times of crisis (think World War II), but there needs to be a strong disincentive to prevent perpetual crisis.
I wish I could believe that Valve was reacting to the negative press about Steam and DRM, and was busily reworking things to remove the dependancy upon Steam in HL2.
I wish I could believe that Valve was working with Icculus.org to make HL2 for Linux.
Unfortunately, I don't. I beleive Valve IS working on Steam, but not to make HL2 less dependant upon it, but rather more so, so that no matter how bad Steam is, you will be FORCED to use it to play HL2.
Oh well, as George Thourogood said, "that don' confront me", as I don't plan on downgrading my OS to play HL2. I'll just wait for Doom3, and play that instead.
And are they working on Genuine People Personalities for this couch?
"Welcome to the La-Z-Ass 3000 couch! It is my pleasure to have you sit on me, and I hope your stay will be as comfortable as possible."
Sounds to me like these guys are a mindless bunch of jerks who will be the first up against the wall when the revolution comes.
This all is coming about because of what I consider to be a design flaw in DNS, to whit:
.foo.bar.baz, there is one and only one possible dataset to answer that query.
.com - thus whosoever controls that dataset controls .com.
.com now, a name server could cache the list of name servers for .com now, and could send the queries out in parallel to the servers it knows about, reducing the time.
.com, just the set of domains registered with .com. It would reduce (somewhat) the size of the servers needed to serve the .com domain (as the workload would be spread out among more servers), and would allow for each registrar to maintain their own database without having to go though Verisign.
.foo, but maybe Alternic does".
.com domain would have to do their homework in registering a domain - there would have to be a clearing house for domain registry.
For any given suffix
So if you are looking up bar.com, there is only one dataset that contains information on
Now, what if a server for a given domain, in addition to having a parent had siblings? For example, if you were looking up narf.com, then the queries might look like this:
my machine - Hey root, where's narf.com?
Root - I don't know, but verisign.com should - ask him.
My machine - Hey verisign.com, where's narf.com?
Verisign.com - I don't know, maybe alternic.com does.
My machine - Hey alternic.com, where's narf.com?
alternic.com - narf.com is at 192.168.0.1
In other words, for every zone record there would be a new configuration possible - a list of zero or more siblings. On a negative result, the sibling records would be returned, and the quering name server would consult them. As a result, you could allow joker.com, register.com, verisign.com et. al. to have just their records on their servers, with cross links to the other servers.
Yes, this would increase the number of queries a name server might have to do to resolve a domain, especially in the negative case (domain does not exist). However, just as a name server can and will cache the servers for
Under this system, a failure of verisign's server would not black out
Extending this to the root servers would allow for things like Alternic to be added - the root servers could say "I don't recognize
Yes, it would be possible for Joker.com and Register.com to create records for VerisignEngineersAreWeenies.com, and for those records to disagree. Yes, the set of owners of servers for the
But were this idea implemented, it would prevent anybody from pulling the kind of unilateral crap that Verisign has.
So, the next time a *BSDer starts bitching about how the GPL is not "free", does this mean we can point out the hypocrisy of criticizing those who provide you with the single most important tool needed to enjoy the environment of your choice?
There are some researchers who beleive they've found the mechanism for this.
When the organism is stressed by lack of food, genes that encode heat shock proteins are activated. HSPs are used when a cell is overtemp, or otherwise stressed, to repair damage to the DNA due to the stress.
The thinking is this: an organism is getting too few calories. The cells start making HSPs due to the stress. The HSPs soak up free radicals, as well as repairing DNA damage. Since the lack of calories is not causing undo damage to the DNA (unlike heat), the net result is more damage due to other environmental effects (radiation, replication errors, toxins) is undone.
In short, the organism's metabolism set to allow it to survive beyond the "famine" to maximize the chances of being able to reproduce once food is available.
The researchers have some good candidate genes for the proteins, and perhaps one day may be able to stimulate the production of these proteins without the need to starve ourselves.
Now, whether the world needs a bunch of long-lived , fat, self-indulgant slobs is another question for which many of the residents of this forum are curiously well-equipped to argue.
OK, this whole thing confuses me, and I am an embedded systems engineer. I've seen this in the trade journals, and I don't get the marketing forces this is supposed to answer.
Consider this: my old laptop that I purchased in 1996 had 3 PCMCIA slots. This was good - I could have my NIC, my SCSI card (for tape backups), and my modem all in place at the same time.
However, any laptop of recent vintage will have at least USB2.0 High speed if not IEEE-1394 (FireWire), so this obviates the need for the SCSI card. It will have built-in Ethernet (at LEAST 10/100 MBits, if not 10/100/1000!), so there goes the Ethernet card. It will have a build in (Win)Modem, removing the need for the modem (at least for Windows users, and very likely for Linux users as well now-a-days).
So what is left for the PCMCIA slots? Flash readers? Built-in, or USB. Video capture? (like you need that in a laptop anyway, but....) Firewire. Video acceleration (MPEG decoding)? Faster CPU. 3D acceleration? Built in.
I can see using PCI-Express (the PROPER name for the new, high-speed serial interface) for the docking station interface - but even then, what do you really need to add to a laptop now?
So what is the point of a PCCARD style interface? OK, I may not be able to get a Firewire tape backup device (or maybe I can - I haven't looked since I don't need one), but if I want to back up a new laptop I can use the network or just dump everything to a Firewire drive.
Now, some may say "Yes, but what about embedded devices". And I can say, as a professional, "What about them?" Either what I am building is a small, simple device, where I would rather build in a USB 2.0 host adaptor, or it is a big, hairy multi-CPU monster that has what it needs built-in. Really, in neither situation would I want to go to the difficulty of adding a PCMCIA-style interface. Been there, done that, and had far too many headaches with people expecting to install J. Random Card and have it work. Sorry, but unless you are using Embedded Windows, you cannot just install the driver disk and go. And if you are using a Windows deriviative, you DON'T WANT users installing their own software (unless you really like watching Customer Service drown).
Again, unless we start seeing laptops with their video on a card, PCMCIA style interfaces are no longer the best engineering decision. Let them die.
OK, folks, here we go again.
/dev/null.
Go back and look over this story. There, one before, we had a story on Slashdot about a software vendor not supporting Linux. At that time, I made a very reasonable suggestion - write to the company involved and ask for a Linux port. I also asked folks who had done so to comment in my Journal, so as to have a public record of the number of letters so written so that when the company involved said "We've never had any requests for this" we could trivially disprove the claim.
And what came of it? Nothing. cat
Why did Valve release a Linux server for Halflife? Because the community innundated them with requests for it.
So you want a HalfLife2 for Linux? Innundate Valve with requests! Stop bitching on Slashdot, and write them a physical, paper and toner letter requesting a HalfLife2 client for Linux.
Now, as for the whole "Just suck it up and run Windows" crowd, and the whole "Fuck Windows - Linux or Nothing" crowd: Each of us must make a decision what is more important - running the OS we choose, or playing a game. And you know what? That decision is going to be different for different people - imagine that!
If you are willing to put up with Windows to be able to run Halflife 2, then by all means do so, have fun, and SHUT THE FUCK UP!
If you will 'live free or die', and refuse to run Windows in order to run Halflife 2, then great! Welcome to the fold, accept the consequences of your decision, and SHUT THE FUCK UP (on Slashdot, that is)! Bitching on Slashdot won't change things, writing a letter to Valve just might!
Or if you don't want to write Valve, then help out on the Wine DirectX layer (and yes, I actually DO have contributions in the Wine source tree.)
But whatever your choices are, accept that they come with consequences, and STOP BITCHING ABOUT THEM WHERE IT WILL DO NO GOOD!
OK, here's something to think about:
When you go to college, you already have to make several purchases (books and such). Why not require students to buy one of the current crop of broadband routers (e.g. Linksys) that provide a layer of firewalling?
Granted, this won't protect the student's computer from local exploits - the worst security exploit of any computer lies between the keyboard and the chair. But for Blaster/Nimda, this would provide another layer of protection - ideally, enough to allow a student to connect their uninfected machine to the network and download the required patches.
In fact, were the routers available in the bookstore, already configured with the appropriate policies by the university, that would greatly simplify the deployment - as the student is buying their books they could also buy a router.
Far too many web developers forget what the FIRST T stands for....
You get a cookie - look in your browser's cache....
I got some bad news for you, sunshine - not only is Pink not well, but you are listed in a lot more public databases than you might think.
Case in point: I got a spam at my work email, the title of which was:
my.work.address REFINANCE YOUR MORTGAGE ON my home mailing address.
Now, let me explain that at no time is my work address directly linked to my home mailing address, which is a P.O. Box. However, my work email address is of the form firstname.lastname@companyname.example, and I recently bought a house. The fact that I bought this house is a matter of public record, and my name, while hardly unique, is rare enough that a spammer could have cross-referenced it with a list of emails, filtered by the location of the company, and got a hit.
Database mining at work.
However, with respect to the entries in the FCC callsign database - while you must provide the FCC with the location of your primary station (a.k.a. your home for most hams), you do not have to have your home address listed on the database - for example, all you would get were you to look up myh callsign would be my mailing address - the self-same P.O. Box I mentioned before.
The long and the short of it is, if you are worried about it, then get a P.O. Box (or a Mailboxes Etc. box), and use that for all your addresses possible. Then, just like having a spamtrap address, you can decouple information from your physical address. If you are truly paranoid, get 2 mail boxes - one for business, one for "junk". This also has the advantage that should you move within the same area, you needn't fill out change-of-address forms for all your magazines and such.
Oh, BTW - if you subscribe to any magazines, you are probably listed on their databases as well, and guess what - they tend to sell those as well.
It's truly hard to be a blank nowadays.
(Bonus credit for readers - work out what my callsign is - I've given it before....)