Think about all the drugs you'll be taking over the next 50 years. Many of those are being developed by the application of population genetics to study populations of people and common mutations in people in order to find causal links between people and disease.
All the statistics for these methodologies rest on assumptions we've made about the underlying models, many of which are based on evolution.
I'm currently looking at very similar (conserved) regions in mice, chimps, and humans, and pinpointing interesting places in the genome that we'll want to study further. If there was no evolution, then why would we be so similar to chimps? Why would we have so many regions of similarity to mice?
There are 3,000,000,000 bases in our genome. If every genome is essentially random, then the level of conservation we see (99% similarity in chimps!) would not be there.
I think this evidence makes it obvious that we shared a common ancestor with chimp. And that's how we leverage evolution to produce drugs.
Please note that while population genetics has been around for a long time, the horsepower to do research (via experimental process) has been in a moore's law type situation for the last 10 years, so we're just getting into the range of doing large scale, comprehensive analysis of the human genome.
The next 10 years are going to be pretty damn cool.
the cover story of the January 2005 issue of Business.
Any "science" discovery published outside of peer review journals isn't worth reading. If I see info about this in a peer reviewed journal, I'll get the following:
a) The data, or references to it. b) The *real* methodology behind generating the data c) The analysis methods used to come to their conclusion, along with important statistics like confidence invervals around there data.
If this is a REAL discovery, then the scientists would publish in a true journal, instead of publishing somewhere else and risk getting 'scooped', when someone else replicates their tests and writes it up the correct way....*sigh*...Back to searching for signals of selection in the genome....
I've got a much better chance at self-actualization if I'm financially viable.
See: Maslow's hierarchy of needs for further questions.
See: Long term planning for reasons why not to live *just* for today. Also check out Bush's enviornmental / foreign / business policies for someone who thinks we'll be hit by an asteroid tomorrow.
I don't much doubt you on the infesability of DNA computers. As a former molecular biologist (turned informatics guy), I have an idea of what primers cost, and how much effort it would take a lab to run any sort of analysis.
Even if it took a cluster a week to solve a problem, that would still be "free", as the cluster is running, and we don't really count electricity bills in our lab costs. Primers cost money, money takes PO's, and PO's have to be explained to the boss.
I can solve any problem using the lab clusters (and hey, MIT has a fairly nice set of machines for me to use any time I want...:) On top of that, after I write the code, there's no additional effort - what if the person running the lab work screws it up? After working in a lab for 5 years, I can say - it does happen, and you'd hate to waste the time and effort repeating the process (and if you get results you wouldn't assume are right, you'll repeat the process anyway, just to make sure...)
Try buying green beans, and roasting them yourself. You do know that the roasted beans flavor will really degrade over a two week period, and after two weeks they will not be worth much, right? Well, they'll be worth something if you aren't tasting your brew - but then why transport them in the first place?
So, if you're bringing roasted beans along, then um...why?
On the other hand, if you're bringing green beans along, they'll last for up to a year if stored properly. If you are in the market for green beans, then you can buy them at web sites in the states for $4-$5 a pound for high quality coffee.
And roasting your own is a great experience, and has a lot of geeky joy to it.
How much does it cost to generate those billions of oligo combinations?
I forget what the current cost of an oligo for an order is, but it might be on the order of $1 per 20bp. Or maybe I have that an order of magnitude too cheap.
Why 20bp? That's a normal standard for 'unique in the human genome' length for a primer. I'd guess when you're in the 'billions' of sequences, you'd probably go at least that specific. Good luck checking them all to make sure they aren't cross-reactive, btw.
So, you're talking about massive amounts of money to get a 'real' computer up and off the ground. The one advantage is that you can solve the same problem over and over - or, if you can reuse subsets of a massive oligo order, you might be better off.
But, between price, time, and design issues, I can't imagine someone doing this in the near future. It's just not a feasable from the reagent side of things.
Unless of course you consider that the only problems that are worth solving on DNA computers are ones that are massively parallelizable.
On the other hand, synthesising all those oligonucleotides is still pretty expensive. Ordering 20 oligos in the order you want them might cost you $.50 or a $1 (if I remember my numbers right...) Or is it about $1 a base? I haven't looked at the lab side of things for prices in a while.
Needless to say, if you want to order a TON of different oligos to to some massive operation, it's going to cost you time and lots of money to synthesise what you want to run. And what if you screw up your 'code', and you have to run it again with a different set of oligos? Ouch. I guess you'd have to 'debug' pretty heavily with simulations before you do the real thing, which um...might be as good as just doing it on a computer.
I have a CD-book that has all the episodes of MST3K. Why do I think to keep all the episodes? Unlike some people who have gotten into the show more recently, I was trading episodes by VHS TAPE and snail mail in the mid 90's, before this new-fandled P2P stuff came along.
Back then, it was a royal pain in the ass to get any new episodes (unless you caught it on a repeat on TV), and there *was* value. Now, I preserve them in case they fall out of favor, so I know there's a complete set somewhere in the world, and if anyone requests an episode, I can fufill that request.
Of course, knowing I can watch any episode I want is a bonus when I feel like watching one. None of that dissapointing "Yeah, I've got a lot of epsiodes, but not the one you want to watch" when a MSTie friend is over.
If all you watch and listen to is present-day hyped media crap that has 10,000 copies on Kazza, then sure, you can download it again. But some of us like things that are a bit more rare, and you can't just 'redownload' them in a suitable timespan.
It's amusing that right now I'm investigating intronic DNA, and looking for signals of selection. A few percent of the genome is conserved in non-gene regions between humans and mice (for example.) Why would the DNA be conserved (against a backround mutation rate), unless it was important.
I can't think of many scientists who think about "junk" DNA anymore...but if I ever get my research finished and published, then I'll add one more nail to the coffin.
Of course, being a fellow Boston citizen, you probably live in an older house. And you're aware that those houses (unless you've rebuilt them from scratch) have very old plumbing systems. These plumbing systems greatly increase the amount of garbage you find in your drinking water. One of the apartments I was leasing a few years ago had a significant amount of lead in the water comming out of the pipes.
If I was working with your friend, I might drink the water at his job, but once it had traveled to my house and been contaiminated, I'd want to avoid it without some kind of secondary purification system.
That said, investing in a Britta or Pur system renders the water safe and palatable enough for my tastes.
No, no, I'm lying. I encode with ogg, set at quality 6. That's not bad, but it still isn't CD quality.
If you have a *good* stereo (no, your computer speakers, or a headphone pluged into your soundcard does not count), you'll hear artifacts if you actually have the real source. In addition, mp3's at moderate quality always sound "flat" to me.
I'll wait until someone offers lossless quality downloads. Until then, I'm far better off buying used CD's...at $3-$5 a CD, it's a far better value.
What's it cost to keep making room for all the crap we generate? Is that included in your costs?
My brother is an enviornmental engineer, and he's moving a freakin' river to make more room for a waste disposal site. They aren't moving the river 5 feet for a few yards, either. This is a major redirection, to clear out more space, to throw more crap.
That's funny. I thought P&T were pretty freakin' biased. They had an idea they wanted to present, and they present it in an over-the-top manner, where it's clear that they aren't presenting the other side of the argument.
For example: They had the chiropractors (etc) are all scammers. So, they interview a very small number, and show you ONE crazy guy who's ripping people off. But they don't present many people who are not acting the same way.
Armed with a little critical thinking, you'll start to see the holes, and misrepresentations in what they present. They don't lie, per se, as much as misrepresent. I find their show to be so irritating because of this that I can't watch anymore (and I saw all of season 1).
Technology is growing by leaps and bounds for taking these mesurements, yes. It used to take a week to get a few thousand data points, at $1.00 or more per data point. A few years later, we can do millions of points a week at a tenth of the cost.
There's still a massive amount of infastructure demanded of this system. And because we can get closer to getting more data, we WANT MORE. More data = better statistics to prove that you're right, covered the problem space, etc. We're still orders of magnitude from doing whole genome scans on people at an affordable cost. Maybe in 10-20 years at the current rate.
Of course, all the neat-o technologies like expression and proteomics are a lot less advanced, and that needs to come up to speed as well.
It's a problem of getting enough biological 'compute' (which is prohibitively expensive), and picking the right ways to analyse the data afterwards.
As for people getting this technology: You could do PCR (dna amplification) in your kitchen. Who does? And that's simple. Who's going to get the massive amount of equipment, and UNDERSTAND how to use all of it to get to where you're going?
I don't see the average joe hacking the genome. Well, I do, but that's in scifi books, written by people who don't have a clue how complex the genome is.
I imagine, that if the technology was out and as available to use as a web browser, that an appropriate 'sandbox' would also be created for the same reason.
You can throw your arms up in the air about some horrible future 100 years from now where people could create tailor fraken-bacteria, but it seems a BIT premature to worry about it...
Agreed, completely. I just wanted to point out to all the people talking about bladerunner, et al. that they were making a *large* jump.
I like bateria. They make nice cheese and champaign. My younger brother works in enviornmental engineering, and hopefully he'll work on some more bacterial models for bioremediation (where this may be useful.)
It's interesting to note that this isn't a new technology. It's more of a catalog of ideas, of things we already knew. It's just put in database form now. And, yes, he isolated some sequence in tubes. But creating long oligonuclotides is getting easier and easier - why not use synthesis to create these short sequence runs?
Before you all get carried away with this, a few things to note:
This is a bacterial genome. What is currently being produced is isolated sets of parts of the genome that have been cataloged as having specific functions in a bacteria. These 'blocks' could be put together, if you knew how to regulate all of them, and you were smart enough to add all the neccesary components for replication.
This sort of information is already known for some bacteria. There is a very small amount of DNA in bacterial genomes, and it's easy to sequence. On top of that, it's easier to figure out exactly what a particular bit of sequence does, so this is just creating a one stop shop to look up particular coding sequences.
What this *isn't* is a eukaryotic genome. You aren't going to be putting together complex organisms this way in our lifetime. We don't even know what the VAST majority of the genome does. Do you remember the phrase 'junk dna'? We're now figuring out that the 'junk' actually has function, and there's even been a case where a mutation in intronic DNA has been shown to cause disease. Life is much more complicated in organisms larger than bacteria, and it's going to take the rest of our lives to reverse engineer complex life, much less begin to design it from scratch.
So, the take home message: It's cool, and it may be useful for bacteria. We're not going to grow organisms, people, tissue, organs, etc with this idea.
There's a difference with biotech vs. real life: In many genetic experiments, I have the 'algorithm' written to solve the problem (given various degrees of statistical confidence.) The problems are that it takes a long time to perform all the experiments, and it costs a LOT of money. Say measuring a data point costs you $0.10 to do. In order to cover a problem space with *real* statistical accuracy, you'd like to do 500,000 data points for each of 10,000 people. That's a LOT of money.
While you might be able to 'design' something to grow, getting the appropriate systems and reagents is going to cost you a TON of money. Most people have no idea how much it costs to do genetic research, because of the enzyme/reagent costs involved, as well as the instrument costs (think $1,000,000 to set up a lab to do some low-moderate work.) So, I don't think your average nutjob is going to have access or money to do the level of tinkering neccesary.
On top of that, I don't care what any of you programmers say. I write code for a living (bioinformatics), and the biology half of my job is by FAR more complex than any of the software systems we'll ever need to write. I don't think 'your average nutjob' is going to be able to understand how incredibly complex any interesting system (read as: does something BAD, or something at all) is. A biological system is NOT a perl script. Why do you think it's taking forever to reverse-engineer life?
That's funny you should say Intel is keeping AMD alive. Who has the faster processors? Who's now copying who's instruction set? Who's later on the 64 bit architecture?
Someone better start holding AMD under a little harder, before they lose the whole market.
When you believe in a religion, you're just delaying your reward (the afterlife.) So, you are refusing to do things you want to do now (screw your neighbor's wife, kill the guy who annoys you) so you will go to later. It doesn't matter if it's heaven or nirvana, there's always a carrot at the end of the stick, and religion teaches you that that dangling carrot will taste much better than the forbidden fruit.
Not that anyone could really tell you how that carrot tastes. That's why there's faith.:)
Yeah. EJBs take 5-7 files to build. And all but one are built by and 1/2 decent IDE. All you have to do is write the implementation class, and the other classes + XML are automagically generated. And if you don't like that, look into XDoclet.
EJBs aren't hard at all to use. But use 'em when appropriate - when you need connection pooling, transactions, declarative security. Use them as a facade for plain ol' java classes that do the other heavy lifting. Get the benefits without putting much work in.
That's what EJBs are for - to utilize services provided by the container, not a catchall place to put *all* your work. A little design goes a long way.
Re:Arguments against postmodern deconstructionists
on
The Golden Ratio
·
· Score: 1
I guess you didn't read New Scientist for the week of jan 31-feb 6th. The main article is all about how scientists are probing how people have different sets of perception, and why.
Of course, it's New Scientist, so the article is shit (they reference someone who 'sequenced' 40 people's whole genomes...that's a misunderstanding), but there's work being done, and there may be some cool papers to read on the subject...
If it was equivalent to a Geforce 2MX, then wouldn't it be named a Geforce 2MX?:)
It's "the successor to the Geforce 2MX" line, but still not as powerful as a Geforce 3.
Upon further reading, I notice that it doesn't even have directX 8 support. Before, I just remembered that the card sucked. Now I know *just how much* it sucks...
Slashdot Mirror
Think about all the drugs you'll be taking over the next 50 years. Many of those are being developed by the application of population genetics to study populations of people and common mutations in people in order to find causal links between people and disease.
All the statistics for these methodologies rest on assumptions we've made about the underlying models, many of which are based on evolution.
I'm currently looking at very similar (conserved) regions in mice, chimps, and humans, and pinpointing interesting places in the genome that we'll want to study further. If there was no evolution, then why would we be so similar to chimps? Why would we have so many regions of similarity to mice?
There are 3,000,000,000 bases in our genome. If every genome is essentially random, then the level of conservation we see (99% similarity in chimps!) would not be there.
I think this evidence makes it obvious that we shared a common ancestor with chimp. And that's how we leverage evolution to produce drugs.
Please note that while population genetics has been around for a long time, the horsepower to do research (via experimental process) has been in a moore's law type situation for the last 10 years, so we're just getting into the range of doing large scale, comprehensive analysis of the human genome.
The next 10 years are going to be pretty damn cool.
the cover story of the January 2005 issue of Business.
...*sigh*...Back to searching for signals of selection in the genome....
Any "science" discovery published outside of peer review journals isn't worth reading. If I see info about this in a peer reviewed journal, I'll get the following:
a) The data, or references to it.
b) The *real* methodology behind generating the data
c) The analysis methods used to come to their conclusion, along with important statistics like confidence invervals around there data.
If this is a REAL discovery, then the scientists would publish in a true journal, instead of publishing somewhere else and risk getting 'scooped', when someone else replicates their tests and writes it up the correct way.
So, um...are people donating computer time to IBM's grid to help out a pharma? Or is this being done to aid particular researchers?
Who am I making money for if I donate my time?
I've got a much better chance at self-actualization if I'm financially viable.
See: Maslow's hierarchy of needs for further questions.
See: Long term planning for reasons why not to live *just* for today. Also check out Bush's enviornmental / foreign / business policies for someone who thinks we'll be hit by an asteroid tomorrow.
www.sweetmarias.com
Nuff said.
I don't much doubt you on the infesability of DNA computers. As a former molecular biologist (turned informatics guy), I have an idea of what primers cost, and how much effort it would take a lab to run any sort of analysis.
Even if it took a cluster a week to solve a problem, that would still be "free", as the cluster is running, and we don't really count electricity bills in our lab costs. Primers cost money, money takes PO's, and PO's have to be explained to the boss.
I can solve any problem using the lab clusters (and hey, MIT has a fairly nice set of machines for me to use any time I want...:) On top of that, after I write the code, there's no additional effort - what if the person running the lab work screws it up? After working in a lab for 5 years, I can say - it does happen, and you'd hate to waste the time and effort repeating the process (and if you get results you wouldn't assume are right, you'll repeat the process anyway, just to make sure...)
Try buying green beans, and roasting them yourself. You do know that the roasted beans flavor will really degrade over a two week period, and after two weeks they will not be worth much, right? Well, they'll be worth something if you aren't tasting your brew - but then why transport them in the first place?
So, if you're bringing roasted beans along, then um...why?
On the other hand, if you're bringing green beans along, they'll last for up to a year if stored properly. If you are in the market for green beans, then you can buy them at web sites in the states for $4-$5 a pound for high quality coffee.
And roasting your own is a great experience, and has a lot of geeky joy to it.
How much does it cost to generate those billions of oligo combinations?
I forget what the current cost of an oligo for an order is, but it might be on the order of $1 per 20bp. Or maybe I have that an order of magnitude too cheap.
Why 20bp? That's a normal standard for 'unique in the human genome' length for a primer. I'd guess when you're in the 'billions' of sequences, you'd probably go at least that specific. Good luck checking them all to make sure they aren't cross-reactive, btw.
So, you're talking about massive amounts of money to get a 'real' computer up and off the ground. The one advantage is that you can solve the same problem over and over - or, if you can reuse subsets of a massive oligo order, you might be better off.
But, between price, time, and design issues, I can't imagine someone doing this in the near future. It's just not a feasable from the reagent side of things.
You mean the stuff that degrades naturally, and is hard-as-hell to work with without totally screwing up? :)
After doing DNA-RNA extractions and experiments for a while, I know which one I'd rather work with.
-Jim
Unless of course you consider that the only problems that are worth solving on DNA computers are ones that are massively parallelizable.
On the other hand, synthesising all those oligonucleotides is still pretty expensive. Ordering 20 oligos in the order you want them might cost you $.50 or a $1 (if I remember my numbers right...) Or is it about $1 a base? I haven't looked at the lab side of things for prices in a while.
Needless to say, if you want to order a TON of different oligos to to some massive operation, it's going to cost you time and lots of money to synthesise what you want to run. And what if you screw up your 'code', and you have to run it again with a different set of oligos? Ouch. I guess you'd have to 'debug' pretty heavily with simulations before you do the real thing, which um...might be as good as just doing it on a computer.
Ouch.
-Jim
I have a CD-book that has all the episodes of MST3K. Why do I think to keep all the episodes? Unlike some people who have gotten into the show more recently, I was trading episodes by VHS TAPE and snail mail in the mid 90's, before this new-fandled P2P stuff came along.
Back then, it was a royal pain in the ass to get any new episodes (unless you caught it on a repeat on TV), and there *was* value. Now, I preserve them in case they fall out of favor, so I know there's a complete set somewhere in the world, and if anyone requests an episode, I can fufill that request.
Of course, knowing I can watch any episode I want is a bonus when I feel like watching one. None of that dissapointing "Yeah, I've got a lot of epsiodes, but not the one you want to watch" when a MSTie friend is over.
If all you watch and listen to is present-day hyped media crap that has 10,000 copies on Kazza, then sure, you can download it again. But some of us like things that are a bit more rare, and you can't just 'redownload' them in a suitable timespan.
It's amusing that right now I'm investigating intronic DNA, and looking for signals of selection. A few percent of the genome is conserved in non-gene regions between humans and mice (for example.) Why would the DNA be conserved (against a backround mutation rate), unless it was important.
I can't think of many scientists who think about "junk" DNA anymore...but if I ever get my research finished and published, then I'll add one more nail to the coffin.
Of course, being a fellow Boston citizen, you probably live in an older house. And you're aware that those houses (unless you've rebuilt them from scratch) have very old plumbing systems. These plumbing systems greatly increase the amount of garbage you find in your drinking water. One of the apartments I was leasing a few years ago had a significant amount of lead in the water comming out of the pipes.
If I was working with your friend, I might drink the water at his job, but once it had traveled to my house and been contaiminated, I'd want to avoid it without some kind of secondary purification system.
That said, investing in a Britta or Pur system renders the water safe and palatable enough for my tastes.
I encode at 256 or 320 VBR.
No, no, I'm lying. I encode with ogg, set at quality 6. That's not bad, but it still isn't CD quality.
If you have a *good* stereo (no, your computer speakers, or a headphone pluged into your soundcard does not count), you'll hear artifacts if you actually have the real source. In addition, mp3's at moderate quality always sound "flat" to me.
I'll wait until someone offers lossless quality downloads. Until then, I'm far better off buying used CD's...at $3-$5 a CD, it's a far better value.
What's it cost to keep making room for all the crap we generate? Is that included in your costs?
My brother is an enviornmental engineer, and he's moving a freakin' river to make more room for a waste disposal site. They aren't moving the river 5 feet for a few yards, either. This is a major redirection, to clear out more space, to throw more crap.
I bet that costs a few bucks.
That's funny. I thought P&T were pretty freakin' biased. They had an idea they wanted to present, and they present it in an over-the-top manner, where it's clear that they aren't presenting the other side of the argument.
For example: They had the chiropractors (etc) are all scammers. So, they interview a very small number, and show you ONE crazy guy who's ripping people off. But they don't present many people who are not acting the same way.
Armed with a little critical thinking, you'll start to see the holes, and misrepresentations in what they present. They don't lie, per se, as much as misrepresent. I find their show to be so irritating because of this that I can't watch anymore (and I saw all of season 1).
Technology is growing by leaps and bounds for taking these mesurements, yes. It used to take a week to get a few thousand data points, at $1.00 or more per data point. A few years later, we can do millions of points a week at a tenth of the cost.
There's still a massive amount of infastructure demanded of this system. And because we can get closer to getting more data, we WANT MORE. More data = better statistics to prove that you're right, covered the problem space, etc. We're still orders of magnitude from doing whole genome scans on people at an affordable cost. Maybe in 10-20 years at the current rate.
Of course, all the neat-o technologies like expression and proteomics are a lot less advanced, and that needs to come up to speed as well.
It's a problem of getting enough biological 'compute' (which is prohibitively expensive), and picking the right ways to analyse the data afterwards.
As for people getting this technology: You could do PCR (dna amplification) in your kitchen. Who does? And that's simple. Who's going to get the massive amount of equipment, and UNDERSTAND how to use all of it to get to where you're going?
I don't see the average joe hacking the genome. Well, I do, but that's in scifi books, written by people who don't have a clue how complex the genome is.
I imagine, that if the technology was out and as available to use as a web browser, that an appropriate 'sandbox' would also be created for the same reason.
You can throw your arms up in the air about some horrible future 100 years from now where people could create tailor fraken-bacteria, but it seems a BIT premature to worry about it...
Agreed, completely. I just wanted to point out to all the people talking about bladerunner, et al. that they were making a *large* jump.
I like bateria. They make nice cheese and champaign. My younger brother works in enviornmental engineering, and hopefully he'll work on some more bacterial models for bioremediation (where this may be useful.)
It's interesting to note that this isn't a new technology. It's more of a catalog of ideas, of things we already knew. It's just put in database form now. And, yes, he isolated some sequence in tubes. But creating long oligonuclotides is getting easier and easier - why not use synthesis to create these short sequence runs?
Before you all get carried away with this, a few things to note:
This is a bacterial genome. What is currently being produced is isolated sets of parts of the genome that have been cataloged as having specific functions in a bacteria. These 'blocks' could be put together, if you knew how to regulate all of them, and you were smart enough to add all the neccesary components for replication.
This sort of information is already known for some bacteria. There is a very small amount of DNA in bacterial genomes, and it's easy to sequence. On top of that, it's easier to figure out exactly what a particular bit of sequence does, so this is just creating a one stop shop to look up particular coding sequences.
What this *isn't* is a eukaryotic genome. You aren't going to be putting together complex organisms this way in our lifetime. We don't even know what the VAST majority of the genome does. Do you remember the phrase 'junk dna'? We're now figuring out that the 'junk' actually has function, and there's even been a case where a mutation in intronic DNA has been shown to cause disease. Life is much more complicated in organisms larger than bacteria, and it's going to take the rest of our lives to reverse engineer complex life, much less begin to design it from scratch.
So, the take home message: It's cool, and it may be useful for bacteria. We're not going to grow organisms, people, tissue, organs, etc with this idea.
There's a difference with biotech vs. real life: In many genetic experiments, I have the 'algorithm' written to solve the problem (given various degrees of statistical confidence.) The problems are that it takes a long time to perform all the experiments, and it costs a LOT of money. Say measuring a data point costs you $0.10 to do. In order to cover a problem space with *real* statistical accuracy, you'd like to do 500,000 data points for each of 10,000 people. That's a LOT of money.
While you might be able to 'design' something to grow, getting the appropriate systems and reagents is going to cost you a TON of money. Most people have no idea how much it costs to do genetic research, because of the enzyme/reagent costs involved, as well as the instrument costs (think $1,000,000 to set up a lab to do some low-moderate work.) So, I don't think your average nutjob is going to have access or money to do the level of tinkering neccesary.
On top of that, I don't care what any of you programmers say. I write code for a living (bioinformatics), and the biology half of my job is by FAR more complex than any of the software systems we'll ever need to write. I don't think 'your average nutjob' is going to be able to understand how incredibly complex any interesting system (read as: does something BAD, or something at all) is. A biological system is NOT a perl script. Why do you think it's taking forever to reverse-engineer life?
That's funny you should say Intel is keeping AMD alive. Who has the faster processors? Who's now copying who's instruction set? Who's later on the 64 bit architecture?
Someone better start holding AMD under a little harder, before they lose the whole market.
When you believe in a religion, you're just delaying your reward (the afterlife.) So, you are refusing to do things you want to do now (screw your neighbor's wife, kill the guy who annoys you) so you will go to later. It doesn't matter if it's heaven or nirvana, there's always a carrot at the end of the stick, and religion teaches you that that dangling carrot will taste much better than the forbidden fruit.
:)
Not that anyone could really tell you how that carrot tastes. That's why there's faith.
Yeah. EJBs take 5-7 files to build. And all but one are built by and 1/2 decent IDE. All you have to do is write the implementation class, and the other classes + XML are automagically generated. And if you don't like that, look into XDoclet.
EJBs aren't hard at all to use. But use 'em when appropriate - when you need connection pooling, transactions, declarative security. Use them as a facade for plain ol' java classes that do the other heavy lifting. Get the benefits without putting much work in.
That's what EJBs are for - to utilize services provided by the container, not a catchall place to put *all* your work. A little design goes a long way.
I guess you didn't read New Scientist for the week of jan 31-feb 6th. The main article is all about how scientists are probing how people have different sets of perception, and why.
Of course, it's New Scientist, so the article is shit (they reference someone who 'sequenced' 40 people's whole genomes...that's a misunderstanding), but there's work being done, and there may be some cool papers to read on the subject...
If it was equivalent to a Geforce 2MX, then wouldn't it be named a Geforce 2MX? :)
It's "the successor to the Geforce 2MX" line, but still not as powerful as a Geforce 3.
Upon further reading, I notice that it doesn't even have directX 8 support. Before, I just remembered that the card sucked. Now I know *just how much* it sucks...