Digital video continues to gain in popularity. There's just one problem... I can barely stand to watch it.

On something such as a DVD, when the scene has more information than the format can store, it falls apart. With VHS, you'd have an image that would look slightly fuzzy or blurred. With digital formats, the distortion looks far worse. Very often, the picture will appear blocky. Besides that, discoloration of objects is far more distracting and generally lowsy looking than the analog equivalents.

There are some tell-tale artifacts that appear on even the most hi-quality digital video. If you watch live sports you've already seen what I mean. Straight lines look broken up into a stair-like pattern, and anyplace there are several small objects, it results in discoloration.

So, everyone colloctively scratches their heads and wonders what the solution could be. Well, I say look to what is old. Perhaps the single most underrated format of all time is the Laser Disc. While DVD has a slightly higher resolution, and the old laser discs were very large, modern technology should have no problem improving the format to surpass DVDs. What's so special about Laser discs, is that they are NOT digital. In fact, they are basically an FM signal recorded to disc. They don't require a large ammount of processing power as DVDs do. That means less electronics, less heat, and less power. What's more, video will just look so much better all around, in any situation. For once, the $10 retail version will look significantly better than the $2 MPEG-4/DivX copy.

Let's compare Ethernet, Firewire & USB shall we?

USB is in every device

Ethernet is in a lot of Hi-End devices (printers, cameras, etc)

Firewire is in digital video cameras, and some external hard drives

USB can go 5 feet

Firewire can go ~25 feet. Further with a repeater

Ethernet can go 100m, infinite with repeaters

USB (2.0) is about half the speed of firewire

Firewire is FAR slower than 100Base-Tx

Ethernet is fast.

USB requires a computer as an intermediary

Firewire does not require any computer

Ethernet needs a computer on each device

__________________________________________________

So, somebody needs to make the next version of firewire with the specs of Ethernet (100 meters, 1000Mbps) while not requring each device to have a CPU & memory, as Ethernet requires. Unfortunately, above all, Firewire needs to be cheaper than USB to take over the device market.

Right now, the device market is completely split. USB has printers, scanners, etc. Firewire is taking over video cameras, hard drives, and has a few niche products that would otherwise use USB (a handful of webcams, printers, scanners, MP3 players, etc.)
Since firewire has not caught on, ethernet is still needed for cameras (Axis), printers, scanners, hard drives (N.A.S.), etc. Which means those devices cost $100 to $200 more than a firewire equivalent would, which would provide adequate distances.

Obviously, even if all current USB devices were firewire, some would still need ethernet functionality, but most would NOT, and they could save $200 a piece.

Perhaps serial SCSI will take the place of firewire in the future, and provide even more performance and flexability. That would be great if it, unlike firewire, could attract device makers.

With the buzz about low-cost PCs (and to prevent having to repeat myself) I decided I would spread some information on Walmart.com's $200 PC, and similiar offerings.

The new $200 PC walmart.com is making available, comes with an 800MHz VIA C3, and that is the root of the problem. Although I knew Cyrix/VIA CPUs have had performance problems in the past (Cyrix is owned by VIA), I was suckered in by the line that said: "offers comparable performance to the 800 MHz Celeron processor". Surely Walmart.com couldn't be lying about such a thing?

I discovered that it was indeed a complete lie. These 800MHz systems, can not even play DivX file at full quality. Many other programs ran slower on the 800MHz Via, than they do on my 750MHz Athlon. So, Walmart.com is falsely advertising the performance of their computers. These 800MHz VIA systems are actually performing comparably to 400MHz AMD/Intel processors.

There are some positive aspects to these systems as well. The power supply had no trouble powering everything I could plug in, all while running much cooler than other power supplies. The CPU, even in 90 degree heat, while having a very small heatsink and fan, remained as cool as room temperature. Every other piece of hardware is of rather high quality as well.

Considering the terrible performance, I would not recomend the $200 system, unless performance is not important to you (perhaps you want to make it a firewall or a file server), or you are unwilling to spend more than $200 for a new computer.

There are other options though. If you look through the "PCs Without Operating Systems", you can find some very good deals. For $300, you can get a 900MHz AMD Duron system, with a higher bus speed, and a winmodem. That is quite a good deal, but I went one better.

I decided to buy the $400, 1400MHz AMD Athlon. In addition to a faster processor, it has an even higher bus speed (266MHz), takes up to 1GB of DDR RAM, and has something I seriously wanted; an AGP slot.

From there, the price curve steepens. The systems that are $100 more, give you only minor improvements. So $400 seems to be the price point for the moment, but a 1.4GHz Athlon with DDR is anything but shabby.

One last piece of concern, is shipping, and sales tax. For each system, you are looking at about $15 for shipping & handling. Tax is a bit more complex.

1. Updates at bottom: currently up to #4

I have been doing serious hard-core printer shopping, and I would like to share my experience with you. This should save many people a lot of time, effort, and money.

Advantages of Laser Printers:

Text looks absolutely perfect, in any font, at any size. It is not jagged, or distorted at all, from font size 2 to 72.

The toner is permanently affixed to the paper. It appears as if the print is part of the paper itself. It will not rub off in any case.

Using the standard of 5% page coverage, you can get black toner for less than 1 cent per-page, and color for about 2 cents per-page, per-color.

Disadvantages of Lasers:

Low-priced laser printers (under $500; black only) almost always have expensive toner (5 cents/page).

You will have to make a large inital investment to get a quality printer.

When low on toner, you will see lines in your prints where it has run out. When you replace a toner cartridge, you are trashing lots of unused toner.

Advantages of SolidInk/Wax Printers:

Black solidink can be had for as little as 2 cents per page (from Xerox), and possibly lower from third parties.

Text is nearly as good as a laser printer

SolidInk does not run out in heavily utilized areas first, as does laser toner.

Unlike laser prints, solidink prints are bright and glossy.

Disadvantages of SolidInk Printers:

The layers of solidink on the paper easilly scratches off, just like wax!!!

The solidink is layered on the paper. This gives a raised look, and varying textures across the print.

Certain areas, and certain colors will be more glossy than others. An unusual look.

Inital printer costs are nearly as high as color laser printers.

Advantages of Inkjet Printers:

Very low inital costs.

Wide range of printers, (6-color, 4-color, wide-format, etc).

Higher resolutions available than laser or solidink.

Supplies (ink, heads) are more readily available.

Ink can be nearly as permanent as laser toner

Disadvantages of Inkjet Printers:

None print text nearly as good as laser/solidink.

Some printers' inks will bleed on the paper, producing fuzzy-looking prints.

Almost always, higher per-page costs (ink, heads) than laser or solidink.

Your computer's processor is heavily loading during the printing process, doing all the processing for the printer.

Slow. The fastest inkjet printers are far slower than the slowest laser or solidink printers, bar none!

Editorial:

SolidInk/Wax:
I would avoid solidink/wax printers due simply to how easilly the print can be damaged. I must conceed that the prints do look very nice. What is needed is some sort of finishing process that will secure the wax to the paper.

Laser:
I love the results of a laser. Unfortunately, they are very expensive, and not available at the highest resolutions that inkjets are. Laser is certainly the best where large volumes need to be printed, and if you can afford the inital investment. It pays off in the end though.

Inkjets:
Inkjets can produce some impressive quality prints. If you look hard enough, you may find high-quality printers, with very low per-page costs. This is the best option if you can not afford to invest in a laser printer, or need to print only a small number of documents.

Recomended Printer

I would recomend the $150 Epson Stylus C80 (or the newer C82). Black ink is less than $0.0115 per-page. Each color of ink is less than $0.0238 per-page (that rivals the best color laser printers). The C80 makes very hi-quality prints, with instant-drying, durable ink.

I welcome any comments or questions.
________________________________________

Update 1: Oct 3, 2002
Just bought myself an Epson C82. Apparently, it has a slightly higher resolution, but the big selling point is the new ink. 70 year life on plain paper, less seeping, and just generally better stuff. Strangely enough, when you go to purchase ink for the C82, it shows the exact same cartridges that go with the C80. Hey, if Epson themselves thinks the old ink is good enough, I'll get it cheap frem 123inkjets.com, and not pay 3 times as much for this new, "better" ink.

END of Update 1
________________________________________

Update 2: Oct 4, 2002
Well, the C82 is up and working. I must say one thing.... Gimp-print SUCKS! That project completely ignores the Unix philosophy of doing one thing as simply as possible. All a print driver needs to do is take postscript from stdin, then output formatted data to stdout. Instead, they introduce a myriad of dependencies on other large software packages (ghostscript, cups, gimp).

Besides that, I do like the C82 so far. The picture output looks better than any other printer I've dealt with, and the text is close to laser printer quality (if you maintain a distance of at least 12 inches from the paper, you wont be able to see the tiny imperfections even if you attempt to). The top-loader is very solid (there will not be any jamming) and quite convient. Very quick and easy to add paper, or for double-sided printing. Far different than older top-loading mechanisms that had so many problems.

Additionally, I decided to stress-test the inks. First, the ink does dry almost instantly, and causes just a minimum of wrinkling of the paper. With the paper completely soaked in water, the ink does not come off. Even rubbing it, the ink stays in place, and in perfect color, until the paper itself rubs off. Even concentrated cleaning fluid only slightly discolored the ink after soaking for several minutes. On the other hand, absolutely nothing took the toner off of my laser printer printouts. So, if your prints are in your janitor's line of fire, look into laminating them, putting them behind glass, or using a laser printer.

So, better graphics than color laser printers, with text quite good, and very low per-page costs. All without needing to fork out at least $6,000 for a hi-end color laser printer (which I had seriously been contemplating for some time). Even if I have to replace this C82 40 times before the color laser might have needed it's first servicing, I would still come out ahead, and have better quality photos all that time.

END of Update 2
________________________________________

Update 3: Oct 5, 2002
Today I learned everything there is to know about the gimp-print drivers. There's about 6 different resolutions... I'll mainly be using 720, since it looks quite good, and there is little noticable difference between 720 and higher resolutions. The hi-quality settings (hq, hq2, uni) don't produce documents that are even slightly better than their standard quality settings (sw). I happen to think that my sw print looks nominally better than the hq2 printout of the same photo.

Speed is an issue that is seriously iritating me. The advertised speed is 22ppm. While I had no expectations of getting that kind of performance, I certainly would think that, at the second lowest resolution setting, I could print text a little bit faster than one page every four minutes. Hell, my 1994 laser printer can print a full 600x600dpi page every 5 seconds, no matter how complex it is, and that is FASTER than it's 4ppm rated speed. All that, and the laser printer does not tie my processor in knots all the while it is printing. Well, that is another pro/con for me to add above.

END of Update 3
________________________________________

Update 4: Dec 1, 2002
Besides being very slow, the printer also has a bad tendency to develop a build-up of ink, and splatter large (1/8th inch) drops of black ink on the paper. This problem didn't start until I had been using it for quite sme time, about half the black ink cartridge, so I can't yet tell if color will have the same problems. Perhaps it will go away when I change to a new cartridge, perhaps not.

It does go away for a short time after I (push the button to) clean the print heads. That supposedly wastes ink. More importantly, the problem resurfaces quikly, so I need to clean the heads about every 10-20 pages or else the prints will be ruined.

I haven't yet changed the black ink cartridge, so I hope doing that will eliminate the problem.

END of Update 4
________________________________________

The whole methodology that is being taught to programmers needs to change.

Programmers' insistence on saving a small portion of their time by making a program grossly ineffecient, or dependant on a great many other libraries, does not save money. If anything, that cost will be more than paid for by those that spend hours trying to get different programs working together. If that application is used by a large number of people, then the hours spent (to make up for the time the programmer saved) by end-users/admins is huge.

I've been there. I've done the work of trying to have two different versions of GLIB/GTK installed because one program didn't work with the newest version, and other apps didin't work with older versions. And that's just something basic... The same situation in a GNOME/KDE environment is hell (and it does happen quite often).

Another thing I hear a lot of is: Use the right tool for the job. This is meant to justify wasting resources, depending on tons of other libraries, and all for the same minor improvement in the speed with which a program is created.

'Use the right tool for the job' only applies to internal use situations, and sometimes not even then. If you're going to make something for use by a large number of others, avoid using anything which requires additional programs/libraries to be installed.

Hey, maybe Perl is the best tool for the job of one part of a program, Python better for another portion, Ruby better for another part, Java for another part, and a bash-specific shell script is best for an entirely different part. So now users are required to install all those, just to use a single program..

Stallman and his FSF (Free Software Foundation) posse are saying how great the GPL is, and at the same time screaming for standardization in everything. What they all fail to realize is that the GPL and standards are mutually exclusive.

First of all, we have the GPL. It is explicitly written to forbid commercial software developers from using GPL'ed programs in their own propritary software. Stallman has dreams of a world where all software is GPL'ed, but that is just not the world we live in. To get ANYTHING accepted widely, it MUST be under a license that allows commercial software developers to adopt it. Like it or not, the majority of products out there are closed-source, and will always refuse to adopt software (no matter how good it may be) if it forces them to relase their own source code. In fact, if it restricts their use of the software in ANY WAY, it will never be adopted.

If you don't believe me, just look at any standards we have. None were relased under the GPl. Most were under public-domain, or the BSD license. A good example is the TCP/IP stack. In almost every operating system, the TCP/IP implimentation was taken from 4.4 BSD-lite. Meaning that, if no open source TCP/IP stacks were availabe, the internet as we know it may not exist today. It would instead be flooded with dozens of different, propritary protocols.

Another example: Kerberos was adopted by Microsoft... Not OpenLDAP, or any of a dozen other options available under the GPL. That single fact results in Kerberos becomming widespread, rather than it's competitors. If Kerberos wasn't under a reasonable license, we would not have the interoperability between Windows and Unix. I'd say open standards are a much better solution than attempting to reverse engineer propritary protocols.

So, I say, let commercial vendors use your software. It's not the best software that becomes standard, rather the best license wins. It in no way hurts anyone that a propritary vendor uses open sourced software. It even helps out by promoting standards, interoperability, etc.

Additionally, the very way people are learning to program today, is limiting the usefulness of software, even if it is released under a reasonable license. Every programmer today is taught to program grossly ineffeciently. Waste CPU cycles madly, there will always be a faster processor!

Take a good look at KDE & Mozilla. Your computer will never be fast enough to run KDE or Mozilla. The programmers writing KDE and Mozilla will continually add CPU-consuming features, to the point that you need to upgrade your computer every 6 months. What's more, these CPU-consuming features are tiny details such as bug-fixes, not such major features that someone might be willing to spend thousands of dollars for new hardware.

I don't propose that programmers spend great deals of time and effort to get a program going just a little faster... However, you are not going to get anyone to impliment any feature if it is consuming vast ammounts of CPU power or Memory for no good reason. That's among the reaons you will not see Mozilla embedded in many devices. The increased harware requirements would negate the software cost (and that's assuming upgrading the hardware is an option)!

Meanwhile, we see the diametrically opposed OpenSSH. It's included by default on many platforms (Linux, *BSD, Solaris, et al), and embedded in many hardware devices as well. Now that should be proof enough that what I'm saying is true. Build good software, release it under a good license (BSD, MIT, public-domain) and everyone, including you, may benefit a great deal, in the end.

The software you save, may be your own!

Tailgating is a horendous thing. It is a major cause of fatalities on the roads, and it is getting more and more common.

For some reason, people believe it is their right to go as fast as they want to on the road. This same mindset, that they are somehow being denied their birthright is what causes accidents, aggressive driving, and (as the morons in the media call it) "Road Rage".

Now, people are welcome to put their own lives in jeopardy, but unfortunately, all too often, innocent bystanders are killed by this mindset, and behaviour resulting from it. So, we need to stop it from happening.

There are two ways to make people better drivers. The first, most important, necessecary, and universal solution, is to impose stricter penalties. The fact is, we have idiots on the roads who have been in accidents, or stopped by police again and again, but the courts NEVER impose a decent penalty.

Just think about it. Someone who has had several major accidents, and has been found criminally at fault, will almost never have their license suspended. It's like taking murders, and attempted murders, trying them, then giving them a slap on the wrist, and handing them their knife and gun as we throw them out of court and back into society.

Want to stop drunk driving? Give it a minimum sentence of 1 year in jail, and revoke their license for 10 years. You will see the number of drunk drivers quickly drop to zero. Same for unsafe driving. First time, revoke their license for a year. Second time caught driving unsafely, mandatory permanent revocation of a license.

Now, those standards will never become law because of our severely left-wing government. Not that I don't agree with the left on some issues, but for the most part, they are nothing but people completely removed from reality. If you think I am lying, just look a Texas. All the stories we heard from our left-ist representitives that allowing guns to be carried would result in old-west style daily shootouts have not only been provent completely wrong, but have been shown to be the exact opposite. In fact, crime is now less common than ever, in Texas.

Since we can not expect legislation to save us, we have to take matters into our own hands. That is not as bad as it sounds.

Think about what tailgaters are doing. They decrease their following distance, in an attempt to make you nervous, so that you will speed up, or change lanes. The answer we hear from left-ist authorities is to pull over so they can get past us. Of course, giving in to someone's intimidations will just cause them to do it even more. Obviously you SHOULD NOT pull over, and you should not maintain your speed and position (for your own safety). How about speeding up? For one thing, speeding is illegial, and, just like pulling over, you give the aggressive driver what they wanted.

The solution is the only option we have left... SLOW DOWN. The tailgater has reduced his following distance to one that is unsafe at your current speed, so you should take your foot off the accelerator until you are going so slow that your tailgater's following distance is safe. What's more, this will accomplish the exact opposite of what they wanted, and will result in people no longer tailgating.

What is the worst that could happen in this senario? The tailgater does not slow down, and you are involved in a collision with a net speed of a quarter of a mile per hour... In that case, you write down the license plate, and pull over. He is at fault for tailgating you. If he does not pull over, you have his license plate number, a description of the car and the person, so you simply file hit-and-run charges with the police and highway patrol.

One more thing that needs to be mentioned. If your tailgater decides to change lanes after you have slowed down, get back up to your original speed. The danger and threat is gone, so you can get on with your life. Also, the tailgater may be trying to pass you, so speeding up will again prevent his tailgaiting from being prosperous.

If you do what I have outlined here, you will be far safer on the roads than ever before. If enough people pick up this habit, tailgating, and aggressive driving will become a thing of the past. Nothing but a distant bad memory.

And remeber... Did your billions of tax dollars result in these instructions being made by professions who write up DMV books on driving? NO! The left-wing government is far too detached from reality to accomplish anything. It took a nobody like myself, who is not getting paid by anyone to do this. I am just here, spitting in the face of government officials, and political correctness, giving you the truth that no one else wants you to know.

The internet is at a dangerous level. Specifically, web technologies like HTML can leak important and useful private information to anyone who wants it. The only thing that keeps us from getting absolutely overwelmed with SPAM, is SPAMers' lack of intelligence or imagination.

Lets assume that a SPAMer has a server with an address of http://www.spam_mail.com .
The first thing I do is set up an automated program to generate e-mail addresses. These will be created by putting each word found in the dictionary in front of common e-mail domains (such as @hotmail.com). So the first e-mail may be sent to aardvark@hotmail.com , and we go through the alphabet from there. In the body of the email is an <IMG tag. The <IMG SRC= tag will contain the server's address, followed by the path to an image and again followed by the e-mail address the message was sent to.

So to put it very simply, when aardvark@hotmail.com reads this message, the image will be requested from the server and his e-mail address will show up in the SPAMer's server logs. e.g. The server logs would look something like:

At 2:30am 203.17.12.191 requested page /images/spam.jpg?aardvark@hotmail.com

And the most sinister part is that aardvark@hotmail.com wouldn't know that he just told a SPAMer that he exists and reads his e-mail. Now the SPAMer would add him to a huge spam list of KNOWN live email addresses.

People don't have to be stupid enough to click on an unsubscribe link, or post their e-mail address on the internet. They just have to be dumb enough not to impliment the e-mail filtering method I've outlined, (which very few people do at this point).

Perhaps just as sinister; this entire process could be automated with simple common tools freely available today.

The necessary tools for this job? Basic scripting/programming skills, a web server, a hostname, and any internet connection.

By Ryan Cooley Tue Jun 25, 2002

See Addenium #1 at the bottom

While you may skip right to the second section, I suggest that you read my "Motivations" as well.

Motivations: Before today, I could sleep at nights, knowing that my computers were completely impervious to attacks. It may sound arrogant, but the only network-accessible service running as Root was OpenSSH. Everything else runs as an unprivlidged user, and is behind and OpenBSD firewall, with additional safeguards. Now, finding out that OpenSSH may very well have a remote Root exploit, I drew upon my years of knowledge, and decided it was time to look into the additional forms of security that I had previously dismissed. Specifically, I was looking for a piece of software that detects if a service acting unusually, and proceeds to kill and restart the software.

Now, through a search on Google, I did not find the software, but I did find an incredible number of sites selling software to protect users from e-mail borne viruses. That is what prompted me to finally write this document, and that is only one of the problems that this method solves.

How to Stop Unwanted E-Mail: Blocking unwanted e-mail is no secret; it doesn't take skill, talent, or magical abilities. All you need to do, is think up a passphrase, and require that passphrase to appear in the subject line of all e-mail you recieve. To keep things simple and easy for everyone, I recomend a single word or phrase, with no punctuation, spaces, or specialy characters. So, a good passphrase could be: yellowsnow. So, then go to your e-mail client's preferences, and setup a new filter that says, if yellowsnow is not in the subject line, deliver to Trash. Now, we are half-done.

The next step is to let everyone know about your filter. So, send them an e-mail, or post a note on a message board of your choice, specify your e-mail as normal, but include a small note afterwards that says:

The word yellowsnow MUST appear in the subject of all e-mails.

Viola, now every piece of spam, and e-mail viruses will be sent directly to the Trash. Not to mention that you will no longer find mis-addressed e-mail sent to you.

Mailing Lists: Unfortunately, you will find all of the mailing lists that you subscribe to (Daily Dilbert, Joke of the Day, MSNBC), end up in your trash as well. This is a simple problem to solve. You go back to your filter preferences, and simply add a new rule to allow those mailing lists through. Of course these rules should preceed the rule that filters based on lack of a passphrase. So, you simply Pass any mail From (whatever e-mail address the list is sent from) to your Inbox. Or, perhaps you'd like to create a new folder to hold all of your mailing lists. Simply specify the name of that folder rather than the Inbox.

Caution: The one thing that can screw up this system is if everyone chooses nearly identical passphrases, or specifies them on public sites in an identical manner. For instance, if everyone uses my example sentence, changing nothing but the phrase yellowsnow to something else, spammers will quickly be able to pickup on that system. Of course, even in that case, you will get only a tiny fraction of the spam you likely recieve now.

Suggestions: To add someone (that uses a passphrase)to your address book, you could simply put the passphrase in the Comments field, so you have it available should you forget it.

When sending e-mail to an automated service, or a help-desk, include your passphrase in the subject line. That way, most replies will have your passphrase in the subject, and special instructions aren't needed for one-time contacts. However, some services will not include the original subject line in the reply, and you'll just be forced to check your Trash folder around the time you expect a reply. You could also set up a filter that treats e-mail from that address as a mailing list, and remove the filter rule when it is no longer needed.

Avoid adding filter rules that bypass this passphrase method, when possible. If one person (who's e-mail bypasses this passphrase check) should be infected with a virus, you have no protection against recieving that same virus from that person.

Conclusion: Now you can cut through the crap. No more million dollar virus solutions, no more wasted time with spam (although you still waste a bit of bandwidth). No more complex setups, or highly questional methods to filter spam which require special accounts (running your own mail server, a POP3-based account). Now with a little setup of a (possibly free) e-mail account, you now have a mailbox more spam and virus-free than the best professional solutions could offer. I hope not too many people loose their jobs or go out of business because of this!

Addenium 1 It seems that Scott Adams, the comedian behind Dilbert, is now using this method to block spam. No doubt having an AOL account just serves to compound this problem. Here is the relevant exerpt from the monthly DNRC e-mail newsletter:

scottadams@aol.com.
IMPORTANT: Put "Dilbert" at the end
of your subject line so my spam filter
won't bounce it back.

This brings us to another possibility in spam filtering. Most filters will allow you to require a certain string to be at the beginning or the end of the subject-line. At the beginning doesn't work out well at all, since all replies, or forwarded messages will have "RE: " automatically prepended to the subject. The only problem with sticking it at the end, is that some e-mail clients will crop the subject, meaning replies from SOME people may no longer contain the passphrase. It is less likely that a word, at the end of the subject, will be removed, but it is still a possibility.