On the subject of management in general, may I also recommend:
First, Break All the Rules. The advice in the book is backed by 25 years of actual research into correlation between management and business outcomes of productivity, profitability, retention, and customer satisfaction. And it just makes a lot of sense.
So a more accurate thing to say might be, "As far as I can tell, no one can be sure they know the truth; they only have their best guess as to what the truth might be." A statement which, as far as I can tell, is true, but not nearly so catchy.:-)
No, it doesn't bother me. It actually quite amuses me when people make these kinds of statements, completely oblivious to the contradiction it contains. I was actually trying to go for the sort of old-west, "Well, is that a fact?" tone of voice, but I obviously failed to convey that.:-)
Those 60 seconds you lose over 10 minutes aren't seconds that can be put into good use, because they are not in one nice 60 second block.
I guess it depends on the exact situation. If the 60 seconds is divded up into sub-10ms chunks, no one will notice; and if the cached images take a little longer to load, but I can start reading the article anyway, it won't matter either. But, for example, if my wife is scanning through her professional photo files with Adobe Bridge, and she's consistently waiting 2-3 seconds every time she navigates a directory for it to load up all of the thumbnails, then she will take 60 seconds longer to finish her task in the end.
And of course, it's perfectly possible that the block cache is actually making bad decisions. And it's even more possible that Firefox has sprung a memory leak, and that it's got a 800MB core, which is the reason why it got swapped out in the first place.:-)
The question, though, is how is the reduction in disk cache size resulting from having no virtual memory to speak of affecting your runtime? Rather than seeing it all at once, like when you swap back in Firefox, are you taking longer to navigate directories because it has to read them in every single time? And when you're using firefox, does it take longer to check its disk cache? Are you saving 2 seconds when you switch applications by losing 60 seconds over the course of 10 minutes as you're actually using an individual application?
Saving the 60 seconds (perhaps at the expense of the 2 seconds) is exactly what the block cache is trying to do for you. Whether it's succeeding or not, or how well, is a different question.:-)
How about this: The difference between men and women makes women in a field valuable. Our field is suffering as a result of having such an imbalance of men and women, and not just because women are nice to to talk to and look at. They think differently and have different values and experiences, and bring a different perspective to problem-solving and team dynamics. 50/50 may not be acheivable, but 70/30 would be much better than 95/5.
(BTW, I think this would also go for traditionally female-dominated fields, like nursing, teaching, and social work: men bring something different to teaching or social work than women, and the field benefits from diversity.)
Not to mention that if the percentage of women in a field is not due to simple preference of the field content but due to some other reason, it's an unjust situation that needs to be rectified.
Yeah, management, entrepreneurship, business development, business strategy... all of these are completely different activities requiring different skill sets. What I was talking about wasn't engineers going into management, so much as engineers using their analytical thinking to optimize business practices, finding the best place / the best way to make money in a particular competitive landscape, making strategic alliances for growth, &c.
There's a link my brother sent me awhile back with a guy with an interesting theory about why there are more men than women in advanced sciences:
Because a PhD in science is a really poor career choice, and women are usually smart enough to avoid it, while men are much more likely to be sucked into it.
The fact is that if you have the brains and discipline to get a PhD in science and become a prof at MIT or Harvard, you could probably make a *ton* more money, and quicker, and have much more job security, going into medicine, law, or business. Or you could be slave-labour grad student for 7 years, then serfdom post-doc for 6 years, slave as an untenured professor for 6 years and then be fired... er sorry, "denied tenure" and be looking for a brand new job or an entirely new career at the age of 35.
But back to the question: one really good career path is to study engineering for undergrad because of the analytics and problem-solving skills it gives you, and then go into business or law. A lot of engineers make awesome businessmen because of the way they've been trained to think by their engineering education.
The fact is that you stand to benefit benefit from that kid, and especially from his education. When he grows up he'll contribute to the economy in which you are a part; he may even help invent the artificial heart or the drug that allows you to live to the age of 130. And when you're too old to work, that kid will be generating whatever money you get from your retirement fund / social security / medicare / whatever, not to mention helping you get into bed, giving you a bath, or whatever else you're too feeble to do. It's only right that you contribute to rearing him.
Paine's main point seems to be, "The four accounts of the resurrection are irreconcilably inconsistent. We may therefore conclude that the Gospel narratives are not reliable sources of information."
But what he actually describes seems very far from irreconcilably inconsistent to me. All I've heard from anyone who actually has interviewed independent witnesses to an event has testified that the details are often contradictory. I've seen it in my own family, when my aunts and uncles remember a specific event in their childhood differently (although they were all there). I've even experienced it myself, when my wife and I both remember an event differently, even one of immense importance to both of us, like the first time we talked about being in love.
Not to mention the different ways we tell a story even when we agree on the details. A story is meant to have a purpose and a point; you arrange how you tell it and what details you include to make the flow.
I challenge you to find four different accounts of such a momentous event (say, the shooting of Martin Luther King Jr) by independent witnesses and not find similar differences. In fact, I would venture to say that in a real court of law, if witnesses' testimonies did not differ in these kinds of details, the opposition would charge that the witnesses had gotten together beforehand to "get their story straight".
The fact is that skeptics to the Gospels want to have it both ways. In passages where Matthew, Mark, and Luke are the same, they say, "These are so similar that they must have been copied from the same document [which textual critics call Q]. Therefore they are not independent confirmations." And when they differ, they say "These differing details are irreconcilable contradictions; we can't trust these accounts." So both similarity and difference undermine the reliability in their eyes. One wonders what evidence would promote the reliability in their eyes.
The fact remains that eleven of the twelve disciples were killed for insisting that Jesus rose from the dead. If they did not have, in their minds, irrefutable evidence that Jesus really had risen from the dead, there's no way they would have gone through that.
FWIW, creationism could still be exactly true and it still would never be science. It makes no predictions, and is not falsifiable.
That's not true. I attended a talk by a geologist as part of a course on Biblical exegesis and teaching here in the UK. The geologist described the layers of sediment clearly visible along one of the coastlines, and how "young-earth" creationists say that these were created by the world-wide flood described in Genesis. Then he asked us, "Now, if these layers were created in 40 days by the flood, what would you expect them to look like -- orderly, or all jumbled up?" Everyone said, "All jumbled up." He then proceeded to describe how the layers are not in fact jumbled, but change in type and makeup as they go upwards; and also how different layers contain different kinds of fossils, going from very simple to more complex.
This resembles, at a popular level, exactly what you describe: ask what testable results a given hypothesis would predict, and then see whether we see that result or not. Whether they like it or not, "young-earth" hypotheses explaining certain phenomena do make falsifiable predictions about what else we might find. The reason they're often viewed as quacks is the refusal to acknowledge those predictions and the subsequent falsifications of it.
BTB, the geologist himself was a Christian, but firmly believes that evolution happened. His main goal in talking to us was to convince us not to dismiss evolution out-of-hand, and that we would do a lot better to avoid evolution altogether, and instead focus on Jesus' resurrection, an argument where we're on a lot firmer ground.
That's a complex piece of accounting, and it's not a job for an IT worker to be asked to do. If you're making the IT worker spend time to justify their job financially, you're not being a very efficient company.
I think this is the most insigtful comment so far. Why the heck should they rely on the assessment of an IT guy? That's not his job. He doesn't had any training, expertise, or experience. Are they going to fire him if he puts the value too low? What do they do when they find out he was wrong then?
Any reasonably sane person in this situation would find out how much his salary is, and then make up a bunch of plausible numbers that is at least 2x that. (Given that he's been asked already, saying "You really need to ask a business productivity expert about that" probably isn't going to get him very far unless he has a particularly good relationship with his superiors.)
Unless he's 40, there's still plenty of time for that.
Seriously, I dated two girls in high school,and then didn't date anyone until 27, when I met the love of my life. Now I'm really happy. The not-dating wasn't really on purpose, but it sure saved a lot of hassle and broken hearted-ness; and it meant that I had a solid 10 years to be really free to do my own thing. It probably also saved me from a highly sub-optimal decision: I'm not really that picky, and if I'd been dating someone reasonably OK at 25, I might have "settled" for that, gotten married, and missed out on something much better.
Actually, something like "make -j" in the base directory of a linux kernel source tree is essentially a fork bomb as well. Did that once by accident; never again.:-)
The "work on an open-source project" suggestion has been mentioned several times, and it seems an obvious one. But does it matter when the rubber hits the road? Can I hear some concrete stories of people (you or someone you know) who either:
Was hired as a result, in part, of their work on open-source projects, or
Hired someone based, in part, on their work in open-source projects?
"if an unauthorized wireless mouse can bring down a plane, we need the entire fleet of such badly defective planes grounded and fixed yesterday".
No joke. If this really is the case, then any terrorist organization could just pop five guys with wireless mice (or, God forbid, laptops searching for wifi signal), and crash the plane. A whole new wave of terrorism is on its way.
I don't think one needs to be a civil libertarian to see "collateral damage" all over the argument in the Lori Drew case. TOS are a simple contract, and the last thing we need is for people to be jailed for violating a (typically) over-zealous contract that nobody reads.
The article talks about Facebook & Myspace suing in civil court spammers who violate the TOS, purely as a breach of contract. No extension there. It's probably a good idea for them to do this, but the fact is that those kinds of spammers don't really affect my Facebook use much -- I have a threshold of friending someone that's more than "random person I have no connection with who asked to be my friend".
But spammers who break into a legitimate user's account to post stuff on their friends' walls -- that just makes me angry. And it is (or should be) 100% criminal, without needing to change or extend any laws.
The thing that's different with Facebook / MySpace spam is that often, they're definitely stealing someone's password to post stuff as them. That is (or should be) squarely in the realm of "breaking into a computer system". Since there is massive evidence of repeated criminal behavior (one for each account they break) and they leave behind a money trail (where does the money for the adwords on xxx.blogspot.com go?), it should be possible to bring a criminal case, have them extradited, and throw them in federal prison.
Someone creating their own account and posting spam is one thing (although as the article points out, even that is covered by the TOS). But someone breaking into someone else's account is a completely different thing.
I used to think this too; but the fact is that what the government is really selling is the enjoyment of dreaming about being rich. People buy the lottery because they find it rewarding: not monetarily, but just in the idea of possibly getting something, or possibly becoming rich. Given that there's a huge number of people out there (a "market") who are willing to "buy" this feeling, wouldn't you rather the government "sell" it (and at least do something potentially useful with the money) than have hucksters and charlatans pocket the money?
For me it works the opposite way: I feel so stupid after paying $5 for what is essentially now just a worthless piece of cardboard, even if the odds were in my favor I might not play.:-)
Also, what are they actually planning on doing with the patent? The rules of the game now include software patents, so a lot of the computer companies I know of are looking for "defensive patents" to fight back with if sued, not to pick a fight. RedHat and Microsoft are good examples of people that have a software patent portfolio but never (yet) sued anyone. (Although MS has threatened -- but in FUD, threats are more effective than lawsuits, as they don't cost anything and you can't be proven wrong in a court of law.)
130 years ago in the US we had factory workers that started working as children, and were forced to work 12+ hour days with no health care and no job security. I say "forced" because the alternative was basically to starve. If your "decision" is based on working in a sweatshop or starving to death, it's not really a decision at all.
What makes it really sting, and people cry "unfair", is that the work they do is much more valuable to the people paying them. The US unions have basically proved that -- if it's a choice between not making cars or paying $25/hr plus all the perks, they'll pay $25/hr. (Just like if it's a choice between starving and working for $2/hr and no perks, a lot of workers will work for $2/hr.) Now maybe $25/hr is not sustainable for a given job -- but if unions didn't think management was always out to screw them over, they might believe them when they said they can only afford $20/hr.
Now, as it happens, a lot of people who work in factories in East Asia have good working conditions and a healthy living wage, so both sides are happy. But you can't just assume, based on the fact that people are coming to the factory, that they're happy with it, or that they're being paid a fair wage.
It might also be a bit of chicken-and-egg problem. I occasionally have ideas for OSS projects that I think would be useful. I've also dabbled enough in UI to know that I don't know jack. So, given that I don't know any UI gurus, how do I take the article's advice that the UI should be designed, or at least sketched-out first, before any coding is done?
After all (as far as I understand it) most independent open-source projects start with one guy "scratching an itch" that happens to scratch some other people's itches, who start adding stuff until it grows. By the time an actual expert looks at it, won't it most likely be so entrenched in badness that it will require a complete re-write?
Maybe what we really need is a UI-designer friendly way of making mockups capable of attracting coders. Then instead of coders "scratching an itch", the UI people could "scratch an itch", and then attract coders to fill in the stuff behind it.
Back in 1997, I started to do this kind of thing in MS Word, and I did define everything. By the time I realized what a piece of crap Word was at the time, I was too far into my semester-long project to switch over. I got into the habit of pressing Ctrl-S compulsively every sentence (since it crashed at least once every editing session), and learned that sometimes you have to remove and add the same diagram 2-3 times before it will put it in a rational place.
Granted, that was over 10 years ago now, but I wish I'd known about LaTeX at the time. Doing my PhD thesis in LaTeX was a much better experience, even with the problems importing graphics.
Slashdot Mirror
Aren't you legally required to accept cash? I.e., "This note is legal tender for all debts, public and private"?
On the subject of management in general, may I also recommend: First, Break All the Rules. The advice in the book is backed by 25 years of actual research into correlation between management and business outcomes of productivity, profitability, retention, and customer satisfaction. And it just makes a lot of sense.
So a more accurate thing to say might be, "As far as I can tell, no one can be sure they know the truth; they only have their best guess as to what the truth might be." A statement which, as far as I can tell, is true, but not nearly so catchy. :-)
Heh -- so you're saying Nietzsche was trolling?
No, it doesn't bother me. It actually quite amuses me when people make these kinds of statements, completely oblivious to the contradiction it contains. I was actually trying to go for the sort of old-west, "Well, is that a fact?" tone of voice, but I obviously failed to convey that. :-)
Is that a fact?
I guess it depends on the exact situation. If the 60 seconds is divded up into sub-10ms chunks, no one will notice; and if the cached images take a little longer to load, but I can start reading the article anyway, it won't matter either. But, for example, if my wife is scanning through her professional photo files with Adobe Bridge, and she's consistently waiting 2-3 seconds every time she navigates a directory for it to load up all of the thumbnails, then she will take 60 seconds longer to finish her task in the end.
And of course, it's perfectly possible that the block cache is actually making bad decisions. And it's even more possible that Firefox has sprung a memory leak, and that it's got a 800MB core, which is the reason why it got swapped out in the first place. :-)
The question, though, is how is the reduction in disk cache size resulting from having no virtual memory to speak of affecting your runtime? Rather than seeing it all at once, like when you swap back in Firefox, are you taking longer to navigate directories because it has to read them in every single time? And when you're using firefox, does it take longer to check its disk cache? Are you saving 2 seconds when you switch applications by losing 60 seconds over the course of 10 minutes as you're actually using an individual application?
Saving the 60 seconds (perhaps at the expense of the 2 seconds) is exactly what the block cache is trying to do for you. Whether it's succeeding or not, or how well, is a different question. :-)
How about this: The difference between men and women makes women in a field valuable. Our field is suffering as a result of having such an imbalance of men and women, and not just because women are nice to to talk to and look at. They think differently and have different values and experiences, and bring a different perspective to problem-solving and team dynamics. 50/50 may not be acheivable, but 70/30 would be much better than 95/5.
(BTW, I think this would also go for traditionally female-dominated fields, like nursing, teaching, and social work: men bring something different to teaching or social work than women, and the field benefits from diversity.)
Not to mention that if the percentage of women in a field is not due to simple preference of the field content but due to some other reason, it's an unjust situation that needs to be rectified.
Yeah, management, entrepreneurship, business development, business strategy... all of these are completely different activities requiring different skill sets. What I was talking about wasn't engineers going into management, so much as engineers using their analytical thinking to optimize business practices, finding the best place / the best way to make money in a particular competitive landscape, making strategic alliances for growth, &c.
There's a link my brother sent me awhile back with a guy with an interesting theory about why there are more men than women in advanced sciences:
Because a PhD in science is a really poor career choice, and women are usually smart enough to avoid it, while men are much more likely to be sucked into it.
The fact is that if you have the brains and discipline to get a PhD in science and become a prof at MIT or Harvard, you could probably make a *ton* more money, and quicker, and have much more job security, going into medicine, law, or business. Or you could be slave-labour grad student for 7 years, then serfdom post-doc for 6 years, slave as an untenured professor for 6 years and then be fired... er sorry, "denied tenure" and be looking for a brand new job or an entirely new career at the age of 35.
But back to the question: one really good career path is to study engineering for undergrad because of the analytics and problem-solving skills it gives you, and then go into business or law. A lot of engineers make awesome businessmen because of the way they've been trained to think by their engineering education.
The fact is that you stand to benefit benefit from that kid, and especially from his education. When he grows up he'll contribute to the economy in which you are a part; he may even help invent the artificial heart or the drug that allows you to live to the age of 130. And when you're too old to work, that kid will be generating whatever money you get from your retirement fund / social security / medicare / whatever, not to mention helping you get into bed, giving you a bath, or whatever else you're too feeble to do. It's only right that you contribute to rearing him.
Paine's main point seems to be, "The four accounts of the resurrection are irreconcilably inconsistent. We may therefore conclude that the Gospel narratives are not reliable sources of information."
But what he actually describes seems very far from irreconcilably inconsistent to me. All I've heard from anyone who actually has interviewed independent witnesses to an event has testified that the details are often contradictory. I've seen it in my own family, when my aunts and uncles remember a specific event in their childhood differently (although they were all there). I've even experienced it myself, when my wife and I both remember an event differently, even one of immense importance to both of us, like the first time we talked about being in love.
Not to mention the different ways we tell a story even when we agree on the details. A story is meant to have a purpose and a point; you arrange how you tell it and what details you include to make the flow.
I challenge you to find four different accounts of such a momentous event (say, the shooting of Martin Luther King Jr) by independent witnesses and not find similar differences. In fact, I would venture to say that in a real court of law, if witnesses' testimonies did not differ in these kinds of details, the opposition would charge that the witnesses had gotten together beforehand to "get their story straight".
The fact is that skeptics to the Gospels want to have it both ways. In passages where Matthew, Mark, and Luke are the same, they say, "These are so similar that they must have been copied from the same document [which textual critics call Q]. Therefore they are not independent confirmations." And when they differ, they say "These differing details are irreconcilable contradictions; we can't trust these accounts." So both similarity and difference undermine the reliability in their eyes. One wonders what evidence would promote the reliability in their eyes.
The fact remains that eleven of the twelve disciples were killed for insisting that Jesus rose from the dead. If they did not have, in their minds, irrefutable evidence that Jesus really had risen from the dead, there's no way they would have gone through that.
That's not true. I attended a talk by a geologist as part of a course on Biblical exegesis and teaching here in the UK. The geologist described the layers of sediment clearly visible along one of the coastlines, and how "young-earth" creationists say that these were created by the world-wide flood described in Genesis. Then he asked us, "Now, if these layers were created in 40 days by the flood, what would you expect them to look like -- orderly, or all jumbled up?" Everyone said, "All jumbled up." He then proceeded to describe how the layers are not in fact jumbled, but change in type and makeup as they go upwards; and also how different layers contain different kinds of fossils, going from very simple to more complex.
This resembles, at a popular level, exactly what you describe: ask what testable results a given hypothesis would predict, and then see whether we see that result or not. Whether they like it or not, "young-earth" hypotheses explaining certain phenomena do make falsifiable predictions about what else we might find. The reason they're often viewed as quacks is the refusal to acknowledge those predictions and the subsequent falsifications of it.
BTB, the geologist himself was a Christian, but firmly believes that evolution happened. His main goal in talking to us was to convince us not to dismiss evolution out-of-hand, and that we would do a lot better to avoid evolution altogether, and instead focus on Jesus' resurrection, an argument where we're on a lot firmer ground.
I think this is the most insigtful comment so far. Why the heck should they rely on the assessment of an IT guy? That's not his job. He doesn't had any training, expertise, or experience. Are they going to fire him if he puts the value too low? What do they do when they find out he was wrong then?
Any reasonably sane person in this situation would find out how much his salary is, and then make up a bunch of plausible numbers that is at least 2x that. (Given that he's been asked already, saying "You really need to ask a business productivity expert about that" probably isn't going to get him very far unless he has a particularly good relationship with his superiors.)
Unless he's 40, there's still plenty of time for that.
Seriously, I dated two girls in high school,and then didn't date anyone until 27, when I met the love of my life. Now I'm really happy. The not-dating wasn't really on purpose, but it sure saved a lot of hassle and broken hearted-ness; and it meant that I had a solid 10 years to be really free to do my own thing. It probably also saved me from a highly sub-optimal decision: I'm not really that picky, and if I'd been dating someone reasonably OK at 25, I might have "settled" for that, gotten married, and missed out on something much better.
Actually, something like "make -j" in the base directory of a linux kernel source tree is essentially a fork bomb as well. Did that once by accident; never again. :-)
No joke. If this really is the case, then any terrorist organization could just pop five guys with wireless mice (or, God forbid, laptops searching for wifi signal), and crash the plane. A whole new wave of terrorism is on its way.
I don't think one needs to be a civil libertarian to see "collateral damage" all over the argument in the Lori Drew case. TOS are a simple contract, and the last thing we need is for people to be jailed for violating a (typically) over-zealous contract that nobody reads.
The article talks about Facebook & Myspace suing in civil court spammers who violate the TOS, purely as a breach of contract. No extension there. It's probably a good idea for them to do this, but the fact is that those kinds of spammers don't really affect my Facebook use much -- I have a threshold of friending someone that's more than "random person I have no connection with who asked to be my friend".
But spammers who break into a legitimate user's account to post stuff on their friends' walls -- that just makes me angry. And it is (or should be) 100% criminal, without needing to change or extend any laws.
The thing that's different with Facebook / MySpace spam is that often, they're definitely stealing someone's password to post stuff as them. That is (or should be) squarely in the realm of "breaking into a computer system". Since there is massive evidence of repeated criminal behavior (one for each account they break) and they leave behind a money trail (where does the money for the adwords on xxx.blogspot.com go?), it should be possible to bring a criminal case, have them extradited, and throw them in federal prison.
Someone creating their own account and posting spam is one thing (although as the article points out, even that is covered by the TOS). But someone breaking into someone else's account is a completely different thing.
I used to think this too; but the fact is that what the government is really selling is the enjoyment of dreaming about being rich. People buy the lottery because they find it rewarding: not monetarily, but just in the idea of possibly getting something, or possibly becoming rich. Given that there's a huge number of people out there (a "market") who are willing to "buy" this feeling, wouldn't you rather the government "sell" it (and at least do something potentially useful with the money) than have hucksters and charlatans pocket the money?
For me it works the opposite way: I feel so stupid after paying $5 for what is essentially now just a worthless piece of cardboard, even if the odds were in my favor I might not play. :-)
Also, what are they actually planning on doing with the patent? The rules of the game now include software patents, so a lot of the computer companies I know of are looking for "defensive patents" to fight back with if sued, not to pick a fight. RedHat and Microsoft are good examples of people that have a software patent portfolio but never (yet) sued anyone. (Although MS has threatened -- but in FUD, threats are more effective than lawsuits, as they don't cost anything and you can't be proven wrong in a court of law.)
130 years ago in the US we had factory workers that started working as children, and were forced to work 12+ hour days with no health care and no job security. I say "forced" because the alternative was basically to starve. If your "decision" is based on working in a sweatshop or starving to death, it's not really a decision at all.
What makes it really sting, and people cry "unfair", is that the work they do is much more valuable to the people paying them. The US unions have basically proved that -- if it's a choice between not making cars or paying $25/hr plus all the perks, they'll pay $25/hr. (Just like if it's a choice between starving and working for $2/hr and no perks, a lot of workers will work for $2/hr.) Now maybe $25/hr is not sustainable for a given job -- but if unions didn't think management was always out to screw them over, they might believe them when they said they can only afford $20/hr.
Now, as it happens, a lot of people who work in factories in East Asia have good working conditions and a healthy living wage, so both sides are happy. But you can't just assume, based on the fact that people are coming to the factory, that they're happy with it, or that they're being paid a fair wage.
It might also be a bit of chicken-and-egg problem. I occasionally have ideas for OSS projects that I think would be useful. I've also dabbled enough in UI to know that I don't know jack. So, given that I don't know any UI gurus, how do I take the article's advice that the UI should be designed, or at least sketched-out first, before any coding is done?
After all (as far as I understand it) most independent open-source projects start with one guy "scratching an itch" that happens to scratch some other people's itches, who start adding stuff until it grows. By the time an actual expert looks at it, won't it most likely be so entrenched in badness that it will require a complete re-write?
Maybe what we really need is a UI-designer friendly way of making mockups capable of attracting coders. Then instead of coders "scratching an itch", the UI people could "scratch an itch", and then attract coders to fill in the stuff behind it.
Back in 1997, I started to do this kind of thing in MS Word, and I did define everything. By the time I realized what a piece of crap Word was at the time, I was too far into my semester-long project to switch over. I got into the habit of pressing Ctrl-S compulsively every sentence (since it crashed at least once every editing session), and learned that sometimes you have to remove and add the same diagram 2-3 times before it will put it in a rational place.
Granted, that was over 10 years ago now, but I wish I'd known about LaTeX at the time. Doing my PhD thesis in LaTeX was a much better experience, even with the problems importing graphics.