That said, I have found a huge oversight on Microsoft’s part, supporting old printers.  I have an older HP Laserjet 4L in my garage attached to my Linux server.  It is shared to the network using CUPS.  In Windows XP they had drivers included for that printer, so I just set up a new printer, pointed it at the right network address, and selected the driver and was up and running.  Piece of cake!  Not so with Windows 7.

I tried to set up the printer on Windows 7 and I noticed the list of available printers is very small, and my printer was nowhere to be found.  No big deal I thought, I’ll go find the Generic Post Script driver and use that, since CUPS and my printer work just fine that way.  Again, nowhere to be found.  Whiskey Tango Foxtrot?  From what I read it appears Vista suffers from the same shortcoming.  I can understand they need to let go of legacy hardware at some point.  Interfaces change, and the wider range of hardware that needs to be supported just holds back advancement and introduces more opportunities for bugs to show up.

The problem here is I don’t even need to directly interface to the printer.  All I need is a postscript file generated and sent across the wire, where the print server will take care of the rest of it.  There is no reason for the generic post script driver to not be included with the OS since it is non-hardware-specific.  I’m communicating using Internet Printing Protocol, which Windows 7 knows how to do.  I don’t need any fancy proprietary USB driver that should never have been supported in the first place.

With the big effort being made by people across the world to be “green” this is clearly a step in the wrong direction.  If I was most any consumer I’d throw the perfectly good printer in the trash and go buy a new one.  Besides the environmental impact, Microsoft is expecting me to not only shell out $300 for the SuperDuperUltimate edition of the OS, they’re now expecting me to spend another $200-300 for a new “compatible” printer.  Upwards of $500 and another hunk of landfill space being taken up is the wrong answer!

What I did was a simple and (I think) elegant solution.  It does involve a couple of extra steps now, but I think the advantages far outweigh the drawbacks, and actually gives me some benefit over direct printing.

1. I created a directory on my Linux box acting as the CUPS server and shared it using SMB.
2. On my Windows 7 machine I connected a drive letter to it.
3. I installed the basic free edition of the CutePDF print driver, so now when I hit print it gets saved as a PDF document rather than being sent to an actual printer
4. When I want to print something, I “print” it using whatever program I normally use, then drop the generated file on the mapped drive.
5. cron runs a script (listed at bottom of post) runs every 5 minutes checking for the presence of PDF documents.  If they exist it sends them to the printer and moves them to a “printed” directory.

To me the extra step of saving the file and moving it to the print queue is a good thing.  If I want to print a page for my records, it gives me the opportunity to just keep it electronically while having the exact paper copy later if I need it.  I’ve also gone to print webpages and have it be completely screwed up when it comes out, even though the “print preview” looked good.  This way I have the opportunity to see exactly what is going to come out of the printer before it is ever on the queue.

The biggest disadvantage of this setup is it won’t work with any printing services that require direct contact with the printer, such as printing postage from the USPS website or some of those online coupon printing programs.  Luckily for those I can still use my wife’s PC.  As I said, these small disadvantages are trivial to me considering I don’t print very much, but when I need to print I can.

Hopefully this will help someone else out who wants to move to a newer Microsoft OS but doesn’t want to throw out some perfectly good hardware in the process.  It took me probably a half hour total of setup time (the print sharing and samba sharing was already set up) to get everything figured out and the script debugged.  You’ll need to extend your time if you need to install additional components that I already had setup.

Here is the bash script I use, triggered by cron every 5 minutes during normal PC usage times:

#!/bin/bash
cd ~/PrintQueue/

# Space removing portion of script lifted from #http://www.linuxconfig.org/Bash_scripting_Tutorial

DIR=”.”
# Controlling a loop with bash read command by redirecting #STDOUT as
# a STDIN to while loop
# find will not truncate filenames containing spaces

find $DIR -type f | while read file; do

# using POSIX class [:space:] to find space in the filename

if [[ "$file" = *[[:space:]]* ]]; then

# substitute space with “_” character and consequently rename #the file
mv “$file” `echo $file | tr ‘ ‘ ‘_’`
fi;

# end of while loop
done

#sleep 5

for f in $( ls ~/PrintQueue/*.pdf ); do
echo “$f”;
lpr “$f”
mv “$f” ~/PrintQueue/Printed/
done

The last 2 years I’ve volunteered to be a speaker.  Both years I’ve given an instruction on the importance of and a crash-course on revision control.  While I feel it is a subject all programmers should be more than familiar with, I wanted to do something a little different this year.

I like to stick to beginner level subjects.  While I could get into serial port and socket programming in Delphi and Java, I feel Code Camp already has a good selection of subjects for the intermediate to advanced programmers.  Since Code Camp targets all audiences, I want to give people a chance to pick up something completely new to them without falling asleep because the material is far above their experience levels.  At the same time I like to choose something I will learn from while I prepare my demonstration.   As an example, the first time I taught revision control I did it along side Jim Munn.  I knew some of the Subversion basics but branching and merging were something I hadn’t done yet.  Jim showed the attendees (including me) that part and I focused on the integration with Trac.  The following year I explored it a little more in depth and taught the whole kit and caboodle.

Here are some of the subjects I considered  for my presentation:

• Test Driven Development in Java
• Beginner’s Home Automation using Misterhouse
• AGI programming for Asterisk
• WordPress widget/plugin development
• Dojo web programming

All of those subjects are topics I’ve dabbled in on some level.  I’ve started with some basic Dojo application development for example, but I’ve worked extensively with Asterisk.  I could even combine some of the above topics, such as using a Dojo interface to a Java based Asterisk AGI program in order to interface with a home automation system that makes a WordPress blog entry when it is complete!  The fun never ends.

In the end I decided to give a beginner level presentation on developing for the Xbox 360 console.  I’m already familiar with the language required due to some projects at work, so this will make for a fun project.  It will also be something a little different and I’m hoping will help attract some of the younger audience to Code Camp that may not have attended otherwise thinking it would be all about business programming topics.

All I had to do was think back and remember trying to play games online in other free networks (such as Delta Force 2 on the PC,) and I’m glad they charge a small fee.  It adds just enough deterrent to keep people from being complete jerks.  When people don’t have to pay, they have no incentive to play nicely.  They can register a free account, then wreak havoc by cheating, shooting team members, using vulgar language, and just general bad behavior.  They are then free to switch back to their other account and play without anyone knowing any better.

With the Xbox Live, you have to be a paying subscriber to play online, and it is per identity, called a gamertag.  Therefore if you want to jump online for a bit just to upset people, you’d have to register a new gamertag and pay to upgrade it to a gold account.  They have free 30 day trials, but they limit that to 3 per console (life time) so it can’t be abused more than a couple of times at the most.  Of course they could use their single gamertag they’re paying for, but there is a decent reputation reporting system so they wouldn’t be able to play with people any more.

My only remaining beef with it is that you can’t register more than one gamertag per Live account.  Since we only have one console, my wife and I will never be able to use our accounts simultaneously.  It seems like quite a waste for both of us to have to pay the yearly fee in order to build our own friends list and statistics.  I wish they would do like Everquest and World of Warcraft and allow a limited number of gamertags under the same account.  That way a family could have one membership and cover everybody.  The way it sits a large family would pay more per year than the purchase price of the console if all of the kids had their own accounts as well.

I’m now anxiously awaiting the Gold subscription keys to arrive that came with some accessories I ordered.  With as much entertainment value as we’ll get out of them, they are worth their weight… in gold.

I’ve also added the Google Analytics plugin so I can see how my blog is being used.  The primary motivation behind this is so I can become more familiar with Analytics.

After switching I always liked many things better, such as the color display, the greater availability of software, and greater compatibility with my Office applications.  Sure there were color displays available on the new Palm models, and I could have paid for software to solve the Office compatibility, but I actually liked my new device.  The one thing I never could find a replacement for was the Eat Watch application.  This was a companion program to the Hacker’s Diet weight loss program I was following.  There were some applications that came close, but none of them were quite what I wanted, and therefore I was unwilling to pay the $50+ pricetag to purchase them.

I happened to stumble upon a site the other day called the Physics Diet, which is essentially a 3rd party online version of EatWatch, inspired by the original program.  I was getting all set to start using it, and was browsing the user forums to see what others had to say about it.  One of the posts pointed to an online version of the original Eat Watch program recently released by the original author!  I was literally beginning the planning phases of porting over the original Palm code to Windows Mobile when I made this discovery, but was very glad to see it ready to use on the web!

I’m still considering building a Windows Mobile version, but now I’m thinking it may be better to create a client to the website that will allow offline entry, rather than writing the whole entire program.  For now I’m going to enjoy using the online version through the mobile web browser on my phone/PDA.  Thank you John Walker!

Starting Domain domain1, please wait.
Log redirected to <glassfish-home>/domains/domain1/logs/server.log.
Incorrect master password
CLI156 Could not start the domain domain1.

I changed the password back to default, scoured the web looking for a fix.  After a bit of hair-pulling, I found the file <glassfish-home>/domains/domain1/master-password was owned by root-root.  I ran the password update command using sudo, but I forgot to force it to my glassfish user.  When Glassfish tried to start, it found it could not read the file containing the password it needed to access its keystore and failed.  A quick chown command to change the owner back to the proper setting (the non-privileged glassfish user) and the server fired up like a champ.

Hopefully this will help the fellow Java geeks from losing any more hair than necessary.

The premise behind this product is that during a seizure, you may be able to catch a PC already booted, in which case the encryption key is already loaded and the data is available.  How do you get it back to the lab without shutting down the system? Enter HotPlug.  To make it simple, you attach the HotPlug device to the computer’s power cables (several methods available, from simple plug-in down to wire stripping and piercing) and this device seamlessly transfers power from the wall socket to your UPS.  You can then take it in a powered up state back to the lab for forensic analysis.  They also have a mouse jiggler device to prevent the screen saver from locking as well, assuming they executed the seizure while the target person was present and using the computer.

Thinking along these lines got me contemplating how to protect oneself against such a device.  While I do not condone working to defeat a lawful seizure by warrant-producing members of law enforcement, I feel these attacks will become more and more common by more shady people, such as private investigators, corporate spys, and identity thiefs.  Stealing a running PC with personal info such as bank account numbers, personal contacts, and online logins could become very much more valuable than the PC itself.  The days of coming home and finding a missing PC with nothing else touched is getting near.  With that in mind, I came up with some interesting ways to defeat the HotPlug:

1. Have your system require a password within say 2 minutes of installing an unknown USB device.  This would prevent the mouse jiggler from doing its job.  It should not prompt for the password, but you would specifically have to go run the utility and enter the key.  That way the attacker would not know he needed the password until it was too late.  If the password is not entered in time, lock it down.  By locking it down I mean wiping the key from memory (to prevent freezing the ram and reading the key) then powering the system off.
2. Use some sort of hardware disconnect to trigger a lock down.  If the LAN cable, USB hub, joystick, or any hardware device the system can detect is disconnected, lock it down.  This would be especially useful if there is no way to move the system without unplugging something.  As an example, don’t use a short cable to attach your machine directly to your network switch/router, otherwise they could leave the router on the HotPlug device too and defeat that security measure.  Leave your router in another room or closet and use in-wall wiring to connect them.  Having multiple hardware devices that must stay connected also takes the job from a one-person in-and-out job to a several person job to move all of the devices without disconnecting them.  Think along the lines of PC, monitor, flat bed scanner, laser printer, external drive, and joystick.  How easy would it be to move that computer with all of those devices remaining powered on and connected?
3. Use a network based device to trigger a lockdown.  For example, have your security software watch to make sure a specific IP or MAC address is available.  This could be inside your network, or for even more security an outside source.  If you have control of the outside source, make sure it only accepts requests from your IP.  If the system is moved and re-attached to a different network (such as a cell phone network) in order to circumvent your security the system would still lock down.  A drawback to this is in the event of a network outtage your system could lock down when you don’t want it to.
4. Use a bluetooth device hidden somewhere else in range of your PC, such as an old cell phone.  Plug it into its own power source separate from your PC (again, preferably a different room) and do not have a battery installed.  This way even if they find and take the device with them, it will still power down and lose connection.  When your PC loses contact with the device, lock it down.
5. GPS presents several protection methods.  Several laptops now-a-days have internal GPS devices.  Program your security system to watch the location and if it moves more than say 1000 yards (to take GPS drift into account) lock it down.  This could be paired with several of the methods above too.  Bluetooth, disconnect or (if they find the GPS device) location change will lock it down.  A USB GPS receiver attached via a long-distance USB cable that can not be moved without disconnecting it would also provide double-protection (if they manage to find a way to move the device while maintaining its connection to the PC, the location change would still lock it down later.

I can think of several varients of the above using different technology, such as cellular data cards (if it detects a tower outside of the normally available towers) WiFi network detection (if the available AP list changes more than X percent over X amount of time) and so on.  In a way these methods of protection are modifications to the old dead man switch idea made popular in several movie plots, only these ideas may be a bit more convenient for the user to have in place.  What ideas can you come up with to protect yourself in the event that your computer is stolen?

When I first installed my equipment, I opted to use wireless mode. That way I could move the sensors around a bit until I found the best location. I also wasn’t sure if I was going to keep the console in the house or place it in the garage along side my server. Since I am happy with the sensor location, and will likely keep the console in the house for the foreseeable future, I went ahead and ran the wire between the sensor and console on Sunday afternoon.

I’ve also been in contact with the author of the WUHU software I use in order to share my updates with Weather Underground (WU). I noticed that any time the software can’t contact WU that my data will have gaps and/or big lines in it. After talking with the author he has given me a test version that will store the update in memory and upload the batch as soon as the connection is restored. That way if a router locks up, a line goes down, or WU has errors, the data will still end up on the site so that people who find it useful will still be able to get to it.