I really do feel the RSI when coding in C++ compared to Python.
On the other hand, better keyboards don't just improved typing speeds, but also reduce strain. For some people this is the difference between being able to use a computer in their work and barely being able to use one at all.
What would be really interesting is if they migrate to a setup where you plug a real guitar in, and then enjoy a game like Guitar Hero. The game would be harder, but much more rewarding. Likewise for bass guitar, drums, or other instruments. That would be the logical next step.
I've also found the command complexity of git to be a bit over the top, in particular compared to mercurial, which in its core can be understood quickly and is very easy to use. Does it need to be that way?
There are several third-party front-ends to git; has can anyone comment of the comparative usability of those front-ends?
I don't agree. Anyone who expect a generation educated by the state to be independent of the state deserves what they get. State-run education needs to stop. That's the real cancer thats been eating our core heritage.
Before the government ran public schools, wasn't it only the church(es)? I prefer the secular version please.
The standard human response for someone into whose outstretched hand you place an opportunity to get out of his rut is to reach out the other hand, palm up.
And equally, the standard human response for poverty and desperation is crime and violence. I'd rather give people a chance.
Maybe he just wants to live long enough to see it begin. Then again, it changes the ball game and seems a shame to die when it seems likely that one day people won't have to (for example, through digitising themselves).
It's certainly part of the job description to teach users, but only as one of many tools in your arsenal. If you can reduced repeated support requests by teaching something to a user, I'm sure you'd do it. Even better, if you can somehow encourage an exploratory problem solving mindset in some of your users, this could reduce even more requests in the long run, at least for the simple problems.
Then again, maybe I'm just being idealistic.
You could equally put a button on the bottom right labelled "Tell me again", and imagine people who click to dismiss over and over and just have it pop up again, only to finally read the button =)
Maybe add a button on the bottom left: "Explain how to replace disk".
1. Your arms are supposed to be relaxed and at your sides.
2. Your wrists should be relaxed, up in the air, and in a fairly horizontal position. The should not be resting on anything.
3. There should be no tension anywhere.
The irony is that people with RSI get immediate pain when these postural aspects are bad, so they usually fix them without much training. This doesn't mean that their RSI goes away, or is mystically cured, but it gets them by.
Granted that desktops are too high for natural typing, but seriously, do try to kill the bad habits. Typing requires no less skill, discipline or practice than playing a musical instrument well. Advocating palm rests and ergonomic keyboards is like recommending beanbag chairs for people who have back problems and habitually slouch, or blowjobs for people with overly-stressed lifestyles.
I think you're making a straw man out of my comments.
There's some irony in suggesting that postural aspects are crucial and equipment is useless, when posture is largely in response to your equipment. I'm not just talking keyboard now, but also chair, desk, monitor, etc. All these things need to be calibrated for your workspace to aid rather than hinder you.
Secondly, my RSI flared up long before I got the Kinesis keyboard, and getting the keyboard did help a lot (disclaimer: after about a week; muscle memory needed retraining). Whilst no new keyboard by itself is a sufficient solution, you'd be naive to think that all keyboards are the same.
The keyboard I use has four advantages:
Encourages better posture
Offloads commonly used keys to the thumbs, which are strong (think space, backspace, delete, ctrl, alt, etc...).
Curves lower and upper rows around home keys. This reduces the overall distance your fingers move.
Reduces typing errors, since it keys are arranged in a grid, instead of the strange diagonals your fingers move in on a normal keyboard
There is a huge market for ergonomic keyboards which the article completely avoids. It is these keyboards, rather than the ones which they present, which offer substantial differences in the typing experience.
I can think of two good examples off the bat. The Kinesis contoured keyboard is what I use, which includes palm rests and vastly different shape which reduces the distance your fingers travel, and takes a lot of strain off your arms. It's also programmable, which is a life-saver for devs. It comes PS/2 or USB, querty or dvorak switchable (i.e. in-keyboard switchable, with dual letter cues), and is solidly constructed.
The SafeType keyboard instead has two vertical parts of a split keyboard, so that your arms have a similar shape when typing as they would holding a large ball by the sides. This is supposed to reduce strain in your arms, by removing some torsion and keeping them in a more natural position. Some keys devs might need (arrow keys) are in a regular, central part of the keyboard. If you need these constantly, it could limit the benefit of changing your hand posture. Still, vim users (for example) wouldn't be affected. Ideally, you'd also get the Evoluent Vertical Mouse to match, so your mousing is also done with your hand in a handshake position.
As someone who's had RSI for a long time, I can tell you that none of the keyboards reviewed in the article are much better, or even significantly different from one another, compared with the difference with real ergonomic keyboards.
Also, sometimes you get bad luck with reviewers. You might get reviewers who think the entire premise behind your research is crap, or ones who don't like your group for whatever reason.
Try submitting to a conference for blind review where you get to rebut your reviewers comments before they make an accept or reject decision. It's a much more pleasant experience, and really improves the whole process.
Often reviewers will fundamentally misunderstand a paper's contribution, if it's from an area very different from their own. But in that case, you get to set them straight *and* redraft your paper a little to make it clearer. If their criticisms weren't fundamental flaws, there's a good chance you'll then be accepted.
However, you are correct that merely increasing the quantity of papers (which is all the current rules do) will cause the quality to suffer. The total thought put in to N papers over a period of time t cannot exceed the total amount of thought the brain can output over time t.
Mod parent up.
I've heard people talk about an MPU (Minimum Publishable Unit). Essentially you can only do so much work in the year. But, you don't get credit on your work directly, instead it's indirect through publications. If you can correctly determine your MPU, you're maximising the credit you get by submitting new papers with minimally acceptable incremental improvement over past papers.
Does this increase in papers increase the actual contribution? No. It's just a natural response to the way academic success is measured. I haven't heard anyone come up with a better way yet though.
For some papers they collaborate on, the effort involved will be to essentially write two paragraphs summarizing the relevance of their specialty area to the main thrust of the paper. For others, they have guided a student through a research problem, and then their contribution to the drafting process can be huge. They can be the ones to say, "It doesn't leave our lab unless it's of a certain standard." They then help students make that standard by helping in the drafting.
Finally, there are some papers which they work on which will tackle a really significant, high-level problem head on, in which they will first-author some work.
With all these, they don't squeeze out papers every two weeks real time, but when an important conference comes around, don't be surprised to see 5 or 6 papers with their name on it.
To what extent this algorithm can be considered "quicksort" is debatable.
You can also perform quicksort on a linked list in C. The algorithm isn't different, just the underlying data structure. The haskell implementation above operates on something akin to a linked list.
It sorts in O(n log n) time for randomized input, and O(n^2) worst case. It uses the method of separating values around a pivot. These things roughly define quicksort, not a particular C imperative algorithm, or particular underlying data structures such as arrays.
We're not talking about subjective value-feelings here; we're talking about intentional manipulation by a sleak advertising campaign that turns people into drones who really do believe that there is something magical in a Mac that other computers don't have.
Let's play the devil's advocate. Suppose you had two products of otherwise equal value. If one is marketed effectively, is it possible that people not only perceive more value in that product, but actually derive more value from it too?
Can you be gamed by marketing, know you're being gamed, and go along with it, because you know you will be more satisfied?
Retirement is different, because it's something you have much of your life to plan for. You might not be able to decide you'll never retire, but you can decide to build up enough savings. Hmm, unless that level of savings turns out to have funny effects on the economy, that would be interesting.
In Australia, we have forced saving through a superannuation system. Systems like that put a huge amount of money into the share market, which is probably the only funny economic effect you could expect. Check long-term share graphs and see if you can see where super investment kicked in. Many people also effectively save by buying a home, which then increases in value over their lifetime.
Aside from these examples though, some people live literally hand-to-mouth their whole lives, due to lack of opportunity or simply bad decisions. I don't want them to die for that, but I don't want them rewarded either. The pension's a good middle ground. Luxury it ain't.
The direct analogy between these services doesn't quite work.
For example, the article claims 50% usage by 3% of the population when everyone has unlimited plans. Does Medicare really face the same cost/benefit distribution? Any evidence you have is welcome.
Medicare is not unlimited in Australia, or elsewhere. It's served on a needs basis, which precludes life-style drugs, cosmetic surgery and alternative therapies with unproven medical benefit. It also may not cover the newest and best treatments for rare conditions.
Starting to sound more like a sound economic investment?
The article argues that we're already there, that people use these drugs and that countries are already partially competing in the pharma scene. It's just that only the newest and most experimental drugs are getting through drug testing.
If we allowed anyone to take anything, I'm sure it would raise the bar, but I doubt it would change other aspects of the competitors. They're already freaks, physically and mentally tuned way off from normal.
I don't quite understand your complaint about the way LaTeX is structured wrt packages. It's pretty much the same thing you see with Firefox where you have a core program with lots of useful plug-ins for added functionality, and as such it's the same argument as it has.
Except that Firefox allows you to search for and install plugins automatically and easily from a central repository. These plugins can also update themselves from newer versions. It seems most people avoid the difficulty of Tex plugins by using a texlive distribution. A better package manager for tex plugins would be very helpful.
For the main extensions I've needed, in particular, cjk extensions, they've always been a real pain. Every new Linux distro has required different steps to get them working; the details of the header annotations required in your document to get them working has changed; and packages in Ubuntu haven't installed the required fonts to get them working, another minefield.
I've started using xelatex instead, which avoids unicode and font issues by using truetype fonts. It's far from unproblematic though. Many existing packages for latex fail to work or work in an incomplete manner when used in combination with xelatex. There's no warning of conflicts, just cryptic failure. There's a lot of room for improvement.
I think you severely underestimate the task of the "slick sales guys". I'm sure that every IT organization has its own fine-tuning in terms of the sales/engineering balance, but they're probably not far off as it is.
Even if your product is the best in the market, it's not worth anything if you can't convince people of that. The world is full of great ideas and products which went out of business to an inferior competitor who marketed better.
Furthermore, once your product is above a certain level in cost/impact, customers demand a level of professionalism in sales. You should be able to convince them that your product is the right choice, and then overcome emotional barriers to making that choice.
That said, the best sales people (in my experience) are those who elicit your real needs, tell you hard facts, and point you in the right direction, even if that doesn't involve buying their product this time. Sounds a lot like requirements engineering, doesn't it. These guys get your respect, and the long term return business, since you'll go back to them every time.
Exactly. The technology they're suggesting is not that useful. Let's think of a better idea.
Suppose instead you use cameras with a full field of view, that don't need to swivel at all and always can record everything. Aside from recording a crime, can we do more?
If you still have these microphones, you could can use them to pinpoint where on a hi-res camera feed the noise came from. If you can identify the type of sound, you could use them in some sort of alert system which escalates warnings to a real person.
None of these fixes the quiet garotting scenario, since there's no sound. Instead, you have AI looking at physical cues and body language for suspicious behaviour. Even then, we're just talking about trying to get there in time to apprehend the culprit; nothing will save the victim.
Slashdot Mirror
I really do feel the RSI when coding in C++ compared to Python.
On the other hand, better keyboards don't just improved typing speeds, but also reduce strain. For some people this is the difference between being able to use a computer in their work and barely being able to use one at all.
What would be really interesting is if they migrate to a setup where you plug a real guitar in, and then enjoy a game like Guitar Hero. The game would be harder, but much more rewarding. Likewise for bass guitar, drums, or other instruments. That would be the logical next step.
I've also found the command complexity of git to be a bit over the top, in particular compared to mercurial, which in its core can be understood quickly and is very easy to use. Does it need to be that way?
There are several third-party front-ends to git; has can anyone comment of the comparative usability of those front-ends?
I don't agree. Anyone who expect a generation educated by the state to be independent of the state deserves what they get. State-run education needs to stop. That's the real cancer thats been eating our core heritage.
Before the government ran public schools, wasn't it only the church(es)? I prefer the secular version please.
The standard human response for someone into whose outstretched hand you place an opportunity to get out of his rut is to reach out the other hand, palm up.
And equally, the standard human response for poverty and desperation is crime and violence. I'd rather give people a chance.
Maybe he just wants to live long enough to see it begin. Then again, it changes the ball game and seems a shame to die when it seems likely that one day people won't have to (for example, through digitising themselves).
Exactly. People only seem to use video calls on their computers, when they're stationary, for example when skype calling one-another.
It's certainly part of the job description to teach users, but only as one of many tools in your arsenal. If you can reduced repeated support requests by teaching something to a user, I'm sure you'd do it. Even better, if you can somehow encourage an exploratory problem solving mindset in some of your users, this could reduce even more requests in the long run, at least for the simple problems. Then again, maybe I'm just being idealistic.
You could equally put a button on the bottom right labelled "Tell me again", and imagine people who click to dismiss over and over and just have it pop up again, only to finally read the button =) Maybe add a button on the bottom left: "Explain how to replace disk".
1. Your arms are supposed to be relaxed and at your sides. 2. Your wrists should be relaxed, up in the air, and in a fairly horizontal position. The should not be resting on anything. 3. There should be no tension anywhere.
The irony is that people with RSI get immediate pain when these postural aspects are bad, so they usually fix them without much training. This doesn't mean that their RSI goes away, or is mystically cured, but it gets them by.
Granted that desktops are too high for natural typing, but seriously, do try to kill the bad habits. Typing requires no less skill, discipline or practice than playing a musical instrument well. Advocating palm rests and ergonomic keyboards is like recommending beanbag chairs for people who have back problems and habitually slouch, or blowjobs for people with overly-stressed lifestyles.
I think you're making a straw man out of my comments.
There's some irony in suggesting that postural aspects are crucial and equipment is useless, when posture is largely in response to your equipment. I'm not just talking keyboard now, but also chair, desk, monitor, etc. All these things need to be calibrated for your workspace to aid rather than hinder you.
Secondly, my RSI flared up long before I got the Kinesis keyboard, and getting the keyboard did help a lot (disclaimer: after about a week; muscle memory needed retraining). Whilst no new keyboard by itself is a sufficient solution, you'd be naive to think that all keyboards are the same.
The keyboard I use has four advantages:
Tell me again how none of this works?
There is a huge market for ergonomic keyboards which the article completely avoids. It is these keyboards, rather than the ones which they present, which offer substantial differences in the typing experience.
I can think of two good examples off the bat. The Kinesis contoured keyboard is what I use, which includes palm rests and vastly different shape which reduces the distance your fingers travel, and takes a lot of strain off your arms. It's also programmable, which is a life-saver for devs. It comes PS/2 or USB, querty or dvorak switchable (i.e. in-keyboard switchable, with dual letter cues), and is solidly constructed.
The SafeType keyboard instead has two vertical parts of a split keyboard, so that your arms have a similar shape when typing as they would holding a large ball by the sides. This is supposed to reduce strain in your arms, by removing some torsion and keeping them in a more natural position. Some keys devs might need (arrow keys) are in a regular, central part of the keyboard. If you need these constantly, it could limit the benefit of changing your hand posture. Still, vim users (for example) wouldn't be affected. Ideally, you'd also get the Evoluent Vertical Mouse to match, so your mousing is also done with your hand in a handshake position.
As someone who's had RSI for a long time, I can tell you that none of the keyboards reviewed in the article are much better, or even significantly different from one another, compared with the difference with real ergonomic keyboards.
Also, sometimes you get bad luck with reviewers. You might get reviewers who think the entire premise behind your research is crap, or ones who don't like your group for whatever reason.
Try submitting to a conference for blind review where you get to rebut your reviewers comments before they make an accept or reject decision. It's a much more pleasant experience, and really improves the whole process.
Often reviewers will fundamentally misunderstand a paper's contribution, if it's from an area very different from their own. But in that case, you get to set them straight *and* redraft your paper a little to make it clearer. If their criticisms weren't fundamental flaws, there's a good chance you'll then be accepted.
However, you are correct that merely increasing the quantity of papers (which is all the current rules do) will cause the quality to suffer. The total thought put in to N papers over a period of time t cannot exceed the total amount of thought the brain can output over time t.
Mod parent up.
I've heard people talk about an MPU (Minimum Publishable Unit). Essentially you can only do so much work in the year. But, you don't get credit on your work directly, instead it's indirect through publications. If you can correctly determine your MPU, you're maximising the credit you get by submitting new papers with minimally acceptable incremental improvement over past papers.
Does this increase in papers increase the actual contribution? No. It's just a natural response to the way academic success is measured. I haven't heard anyone come up with a better way yet though.
They won't be first author on all these papers.
For some papers they collaborate on, the effort involved will be to essentially write two paragraphs summarizing the relevance of their specialty area to the main thrust of the paper. For others, they have guided a student through a research problem, and then their contribution to the drafting process can be huge. They can be the ones to say, "It doesn't leave our lab unless it's of a certain standard." They then help students make that standard by helping in the drafting.
Finally, there are some papers which they work on which will tackle a really significant, high-level problem head on, in which they will first-author some work.
With all these, they don't squeeze out papers every two weeks real time, but when an important conference comes around, don't be surprised to see 5 or 6 papers with their name on it.
To what extent this algorithm can be considered "quicksort" is debatable.
You can also perform quicksort on a linked list in C. The algorithm isn't different, just the underlying data structure. The haskell implementation above operates on something akin to a linked list.
It sorts in O(n log n) time for randomized input, and O(n^2) worst case. It uses the method of separating values around a pivot. These things roughly define quicksort, not a particular C imperative algorithm, or particular underlying data structures such as arrays.
And he clearly states that before he even goes into Djangos documentation and concept of MVC.
He ignores Django's own take on their differences with traditional MVC, and prefers to just bash a straw man.
We're not talking about subjective value-feelings here; we're talking about intentional manipulation by a sleak advertising campaign that turns people into drones who really do believe that there is something magical in a Mac that other computers don't have.
Let's play the devil's advocate. Suppose you had two products of otherwise equal value. If one is marketed effectively, is it possible that people not only perceive more value in that product, but actually derive more value from it too?
Can you be gamed by marketing, know you're being gamed, and go along with it, because you know you will be more satisfied?
In my world it's called Delicious Library.
Mod parent up. I was wondering when someone would make the connection. This could be like a portable version for shopping.
Retirement is different, because it's something you have much of your life to plan for. You might not be able to decide you'll never retire, but you can decide to build up enough savings. Hmm, unless that level of savings turns out to have funny effects on the economy, that would be interesting.
In Australia, we have forced saving through a superannuation system. Systems like that put a huge amount of money into the share market, which is probably the only funny economic effect you could expect. Check long-term share graphs and see if you can see where super investment kicked in. Many people also effectively save by buying a home, which then increases in value over their lifetime.
Aside from these examples though, some people live literally hand-to-mouth their whole lives, due to lack of opportunity or simply bad decisions. I don't want them to die for that, but I don't want them rewarded either. The pension's a good middle ground. Luxury it ain't.
The direct analogy between these services doesn't quite work.
For example, the article claims 50% usage by 3% of the population when everyone has unlimited plans. Does Medicare really face the same cost/benefit distribution? Any evidence you have is welcome.
Medicare is not unlimited in Australia, or elsewhere. It's served on a needs basis, which precludes life-style drugs, cosmetic surgery and alternative therapies with unproven medical benefit. It also may not cover the newest and best treatments for rare conditions.
Starting to sound more like a sound economic investment?
I recommend brewing beer. Then at least you'll be reducing their living costs.
The article argues that we're already there, that people use these drugs and that countries are already partially competing in the pharma scene. It's just that only the newest and most experimental drugs are getting through drug testing.
If we allowed anyone to take anything, I'm sure it would raise the bar, but I doubt it would change other aspects of the competitors. They're already freaks, physically and mentally tuned way off from normal.
I don't quite understand your complaint about the way LaTeX is structured wrt packages. It's pretty much the same thing you see with Firefox where you have a core program with lots of useful plug-ins for added functionality, and as such it's the same argument as it has.
Except that Firefox allows you to search for and install plugins automatically and easily from a central repository. These plugins can also update themselves from newer versions. It seems most people avoid the difficulty of Tex plugins by using a texlive distribution. A better package manager for tex plugins would be very helpful.
For the main extensions I've needed, in particular, cjk extensions, they've always been a real pain. Every new Linux distro has required different steps to get them working; the details of the header annotations required in your document to get them working has changed; and packages in Ubuntu haven't installed the required fonts to get them working, another minefield.
I've started using xelatex instead, which avoids unicode and font issues by using truetype fonts. It's far from unproblematic though. Many existing packages for latex fail to work or work in an incomplete manner when used in combination with xelatex. There's no warning of conflicts, just cryptic failure. There's a lot of room for improvement.
I think you severely underestimate the task of the "slick sales guys". I'm sure that every IT organization has its own fine-tuning in terms of the sales/engineering balance, but they're probably not far off as it is.
Even if your product is the best in the market, it's not worth anything if you can't convince people of that. The world is full of great ideas and products which went out of business to an inferior competitor who marketed better.
Furthermore, once your product is above a certain level in cost/impact, customers demand a level of professionalism in sales. You should be able to convince them that your product is the right choice, and then overcome emotional barriers to making that choice.
That said, the best sales people (in my experience) are those who elicit your real needs, tell you hard facts, and point you in the right direction, even if that doesn't involve buying their product this time. Sounds a lot like requirements engineering, doesn't it. These guys get your respect, and the long term return business, since you'll go back to them every time.
Exactly. The technology they're suggesting is not that useful. Let's think of a better idea.
Suppose instead you use cameras with a full field of view, that don't need to swivel at all and always can record everything. Aside from recording a crime, can we do more?
If you still have these microphones, you could can use them to pinpoint where on a hi-res camera feed the noise came from. If you can identify the type of sound, you could use them in some sort of alert system which escalates warnings to a real person.
None of these fixes the quiet garotting scenario, since there's no sound. Instead, you have AI looking at physical cues and body language for suspicious behaviour. Even then, we're just talking about trying to get there in time to apprehend the culprit; nothing will save the victim.