Follow Slashdot stories on Twitter


Forgot your password?
DEAL: For $25 - Add A Second Phone Number To Your Smartphone for life! Use promo code SLASHDOT25. Also, Slashdot's Facebook page has a chat bot now. Message it for stories and more. Check out the new SourceForge HTML5 Internet speed test! ×
User Journal

Journal Journal: Bacteriophages: Another reason to love open source

This month's Wired Magazine contains an article about phages. Phages are Soviet discovered virii which effectively target and destroy antibiotic-resistant bacteria (as well as other strains). Western pharmaceutical companies and beuracracy are to blame for phages not being widely available here. Despite not having gained US acceptance, they are available over the counter in some European contries. Some people infected with drug resistant bacteria are even travelling overseas to get access to phages which haven't yet received FDA approval. Where there's a will, there's a way.

Part of what makes Phages interesting is comparing how they work to how our more common antibiotics work. According to the article: How Ravenous Soviet Viruses Will Save the World antibiotic molecules work by replacing stuctural emzymes in bacterial cell walls with themselves. Think of a mason replacing cement bricks in a wall with paper-mache bricks. Eventually the walls become liquid-permeable, and fluid rupures the cell walls killing the bacteria.

In contrast, a phage works by attaching itself to a bacterial cell, then injecting DNA into the host cell. Think of a mosquito here. The injected DNA replicates, and directs the host cell to manufacture new phages (inside the bacteria cell host). Eventually the phage clones burst through the cell wall, killing the host cell. Think of the movie Alien here.

The part of the article which really grabbed my imagination was when they talked about all of the money and time it takes to create an antibiotic. Unfortunately, these antibitocs are good only until evolution produces antibiotic resistant strains, then all the investment is effectively gone.

Phages aren't subject to obsolesence. You just need to figure out which phage like to eat the specific bacteria strain you are fighting. It just keeps working. Oh, and thanks to mother nature, between our rivers and oceans, there's a phage waiting to be found for (theoretically) every bacteria out there, and they evolve naturally alongside their beloved bacteria hosts. Yes, the means of locating the phages is reasonable, and much cheaper than developing a new antibiotic.

So, like a true geek, I naturally started thinking about this relates to technology. Over the years, Microsoft has conditioned us to believe that their software is the one antibiotic miracle drug to solve your problems. Like the US based pharmaceutical industry, they downplay the value of open source just like soviet developed bacteriophage methjods.

In the end it would seem that antibiotics are burning themselves out of usefulness, and we'll be back at mother nature's door as she continues to naturally evolve phages in parallel with bacteria. Likewise, business will eventually realize the futility of locking themselves into only seeing one door when there are two very nicely built doors in front of them. That's step one. Step two happens when business realizes that what's behind open source door #2 is a naturally evolving phage. One that evolves faster then closed source is capable of in the long run. Not only does open source appeal to me for political (perhaps religious?) and aesthetic reasons, but after reading this Wired article I am also drawn to open source because it seems to be more in harmony with its envronment.

User Journal

Journal Journal: How does the Windows world tolerate it? 1

As part of my Windows duality I wanted to get my system in Windows as close as possible to my Linux environment. The final hurdle in my path was the CD Burner. Since I bought that burner in my monogomous Linux days I hadn't managed to archive the driver CD that came with it. Surely Windows has a generic CD burner package by now - these things are everywhere.

Surely not.

I searched everywhere for a free tool (read: FREE, not download an eval that needs to be licensed). I found nothing. It looks like there are two primary CD burner packages out there people use, but both cost $$. Not unreasonable amounts, but more than I want to pay given that I have great free software on my other partition that does it for free.

I really had to stop and wonder... For something as basic as CD writing, how can the windows world tolerate all of the nickel-and-diming that goes on? While Windows itself had become less abrasive to me over the years, the neighborhood had remained the same. How sad that so many go without enlightenment.

User Journal

Journal Journal: Going home and keeping an open mind 1

When I first started using computers I was an Apple man. I instantly fell in love with the Apple ][ our school had in the library, and began dabbling in BASIC. Staying true to to my relationship, I maintained my fascination with Apples right through the Macintosh period of my life. All the while ignoring that PCs existed. When my Dad bought his first 486 with Windows 3.1 for his business my mind opened. This machine could do a lot of cool things... And it had a command line (DOS) which brought me back to the Apple ][ days I had not realized that I missed. From that point on I had been converted as there was definitely more under the hood of a PC than the Macs, and oh, did I love to tinker.

I continued using Windows faithfully through my sophomore year of college when I had yet another awakening. A friend of mine had managed to get an early 1.x Linux kernel working on their machine. Now *this* was cool I though to myself. I stayed on course with Windows however as I could get my UNIX fix from the Ultrix machines on campus and didn't really see the need for another OS in my life.

Sometime later that year a few of my friends had repartitioned their hard drives and loaded up Linux. Now its around the Junior year of computer science that things start getting hairy, and that's when I started feeling the need for a programming environment in my dorm instead of the lab. Suddenly I had a "killer app" that justified an OS adoption. That killer app turned out to be PPP.

From that point in ~1993 on I fell deeper into my Linux relationship each day. There was literally something new to discover each day (and 10 year slater this hasn't changed at all!). It took me, and possibly Linux as well, another 2 years to become monogomous. My Linux partition was full, and the Windows partition was home only to a few remaining games which I was able to bid farewell without significant remourse.

For 10 years I remained faithful to Linux. I used it for everything, and did so happily until now. I volunteer as a system architect for the local Boy Scouts of America council where two Linux servers are gainfully employed in the service of ~ 30 Windows 2000 workstations. My dream is to replace their Windows apps with LAMP based web apps once step at a time. First step is to convert their Access apps to hit PostGreSQL via ODBC drivers and replace the local tables.

After fighting it for some time I finally gave in and repartitioned my home workstation's drives. The office had retired a machine and freed up a Win2k license for me so I could begin reaquainting myself with my old friend. In some ways it was like going home - so much was familiar to the Win95 I last knew - yet so much had changed.

My Linux environment was GNOME. Without getting into a flame war I will say that I feel that KDE feels too much like a carnival to me - too many flashing lights. GNOME feels mostly unobtrusive, and that's what it's all about for me.

My first eye opener came during the OS loading. I have a KT400 chipset and a new Radeon vid card, so it's not shocking that the Win2k CDs didnt' recognize everything. The surprise was that without additional drivers the OS could only function in 800x600 with 16 colors. But before I could clue in the OS I needed to get my AGP recognized. In the end, before my OS could make use of all my hardware I needed to not only load my OS, but 5 different driver disks. Most of these drivers wanted to reboot after loading.

This is in contrast to my experience with RH Linux (since the first RH release...). With RH I have every driver I need to run my system optimally when the OS comes up. The only thing I'm missing is a working sensors installation for my ASUS board, but I'll overlook that.

Next, I wanted to see how the apps had changed. Amazingly, there was almost nothing of use on the system... The OS, and a few applets like Wordpad were there. Again, in contrast to Linux where you have almost every conceivable application on the system when it loads. Exploring was a little boring.

What I *did* notice was that I'm no longer religiously aggressive towards my desktop environment. I used to worship at the temple of WindowMaker regularly, and during those years I grew to hate the Windows UI. Perhaps using GNOME had helped with the transition. Whatever the reason I found that it didn't feel abrasive to me.

I loaded up a few apps... Putty for an SSH client, OpenOffice (because all my docs are in that format), and Mozilla. Now it was beginning to feel like home. In fact, there wasn't a lot of difference. Putty feels just like an Xterm, except I had grown fond of Gnome-terminal's tabs, and wished I could find them here. OO and Moz really dont' behave with any difference, so there was no loss here.

Eventually I had loaded up all of my disks with things like GPS software, digital camera software, my astronomy program, etc. At the end of that CD swapping freenzy I had another impression... Linux still hasn't figured out how to make software installs easy. Understand that I'm somewhat geeky, and have been with UNIX for a long time now, so it's not that I can't get software installed. It's more that I don't want to blow time installing software, or locating that missing library that make complains about. It was nice to click a few buttons and have something that works. I wish the configuration and tweaking in Windows was as accessible as it is in Linux - no doubt I'm not in sync with that design. But Linux could gain a lot of users if it were as easy to get going as Windows.

Even with RPM it often requires a lot of tweaking in the Linux world. A prime example is the UPS monitor I've messed with. I have a USB HID interface to my UPS. To get rolling I have to weed through drivers, READMEs, etc. What I *want* is a nice simple probe so I don't need to think about it. I want to spend my time hacking, not getting my UPS recognized.

Other than the convenience of having access to the platform most software is written for again, I still have to say I prefer Linux. It looks better, runs faster, and does more. But I have to say that I have a new confidence in Linux' ability to break into the desktop space with projects like Sun's MadHatter. It's already surpassed Windows usability, and aesthetics - now its just about finding the killer app. Hopefully in the meantime Linux can make some leaps and bounds in the "final polishing" of its apps and give RPM some steroids to make installations more convenient. Until then, I'll keep testing the waters and playing both sides of the fence.

User Journal

Journal Journal: Frustration with Linux 2.6.0 test 1

It was with eager anticipation that I downloaded the 2.6 test 1 kernel from I couldn't wait to see how a preemtible kernel would "feel" on my machine, and I was also psyched to have XFS (the filesystem - not the font server) available without needing the SGI patches. Unfortunately, what I found was what I continue to feel holds linux back - inaccessibility.

In my mind a test-1 kernel is much like a release candidate. Something we would expect to see some minor bugs in, and perhaps a bit less polishing than the final product, but overall funtionally stable. Maybe I expect too much.

My goal was to have something as close to the RedHat kernel as possible. They've done a lot of work on picking options, so who am I to challenge them? I only muck with the kernel when I need a feature, so the only change I wanted to make was compiling in XFS for some benchmarks on a new server I'm building.

When playing with 2.4.21+SGI_XFS I was able to meet my goal by copying the RH config file into place, then selecting XFS in the menuconfig, and resaving. A quick couple of command lines and my Athlon would make quick work of the job. Then came the 2.6 darkness.

It toook me about 2 days (in between other things) to go through the config, build, die cycles. Each time I'd unselect an offending module (bzImage went fine - modules kept blowing up), then begin anew. Finally, I made it through no less than 20 iterations and finished with no errors. I did a make install, then make modules_install thinking I was in the clear. After a panic because the kernel was unable to mount the root filesystem which 2.4.21 finds ok, I am on the verge of giving up.

I'd love to think that the problem is some RH quirk which could easily be solved, but then it would amaze me if RH had designed a system so as to be completely incompatible with the stock kernels. I'm sure someone can hardly resist starting a religous war over that comment.

I'm no kernel expert - not by a long shot. But I do have a decent grasp on Linux, and a strong one on Solaris - I'm no UNIX newbie. So why on Earth is this kernel (and a bunch of the 2.5 kernels I tried) almost impossible to use?

Here's a fraction of the ones I ran into - got tired of recording them all.

make[2]: *** [drivers/media/video/zr36120.o] Error 1
make[2]: *** [drivers/scsi/cpqfcTSinit.o] Error 1
make[2]: *** [drivers/scsi/pci2000.o] Error 1
make[2]: *** [drivers/scsi/dpt_i2o.o] Error 1
make[2]: *** [drivers/video/pm3fb.o] Error 1
make[2]: *** [drivers/video/pm2fb.o] Error 1
make[2]: *** [drivers/video/pm3fb.o] Error 1

Slashdot Top Deals

"If you can, help others. If you can't, at least don't hurt others." -- the Dalai Lama