You know it really isn't that hard to document code as it's being written. Seriously. Just put four or five lines at the top of the function saying what it does, what it takes, and what it returns. A single line before each "paragraph" of code does wonders.
Putting it off later not only makes the documentation harder (stale code is hard to read), but all but ensures that it will never happen.
That's interesting at shdh30 I was talking with this one guy who pointed another guy across the room. Apparently the other guy was fine engineer, but constantly complained about how the code "wasn't clean enough" and the design "wasn't right" and all the other standard moans of engineers, but to the point that he would constantly get in fights with the other engineers. (There's a certain level of mess and incompleteness/wrongness you have deal with. Nothing is perfect, not even your own work. Eventually, you have to grow up and learn to deal.) So the guy quit to do consulting on his own where he only takes projects that he can write all by himself.
The thing that weirded me out was the fact that the guy I was talking to genuinely admired antisocial engineer. My comment was, "Huh. I wouldn't hire him. There's six guys in this room that that could do his job without having me put up with shit."
Well, in any scientific collaboration consisting of more then four people, there's most likely someone senior and crotchety who's stuck in his ways doesn't want to completely change the way he works. You'd also have to build a consensus that svn+latex was the best available solution, and that might not be so easy. I've used svn+latex. It sucked, partly because svn sucks. (Git is a lot better.)
If you have four people all trying to edit the same document, you've got a crap document. The average paper is 8 to 12 pages including all figures and citations. Journal articles are like 20. One person writes it, the other people review it and send changes. If a big section needs to be written. It's sent to the main author who rewrites it if needed. If you don't do this, you get an article that looks like it was written by four people.
Word is a piece of crap when it comes to equations and citations. Also, you're constantly stuck screwing around with the formatting which really is something that the editor should mess with. Yes, there's the "style" menu in Word, but no one ever uses it, and ensuring that styles are consistently used throughout the doc is quite error prone.
Well the problem with #3 is that each potential planet object has to get larger the further from the star it is before it can clear its orbit. If you put Earth out at the orbit of Pluto, it wouldn't be able to clear its path either. Strictly speaking, Earth hasn't it cleared its orbit either.
The most damning thing I found about the IAU's vote wasn't the criteria. It all sounds reasonable, albeit with the problem of requiring "planets" to increase in mass with orbital distance. It's the fact that The "planets" of our solar system aren't actually defined by the criteria but rather a footnote! As seen in the video of the vote, the IAU made clear that this planet definition is only applied for our solar system. One astronmer even asked, "Well then, why don't we just vote on the footnote then?", and the answer back from the chair was essentially, "Well because."
And don't even get me started with the requirement that the definition not include "too many" objects. Essentially, they just wanted eight.
There's potential value. Facebook has gobs of potential value, but they've never shown they know what to do with it. Even if assume that they're breaking even, they're leaving tons of money on the table. They know all about each user's demographics, likes and dislikes, usage rates, and the same stuff about all his/her friends and the strength of connections across the social graph, and yet FB is still serving gay dating ads to straight men, misspelled "Test your IQ" scams, and Russian mail order bride sites. These are spam ads!
They don't know what they're doing, and some grownup needs to oust Mark Zuck and install someone that actually knows how to make money. The fact that FB hasn't shown they know how to extract revenue is hurting them. Deals are falling apart because FB wants to pay in stock, and there's growing consensus that FB's privately held stock isn't worth nearly what FB says it is.
If I was making a deal with FB, I'd insist on cash.
Provide for people who still have old equipment and cannot reasonably afford replacement equipment, provide a way to buy a device that gives them basic functionality without requiring a subscription service. Subsidize this out of funds paid for by industry.
Thus the $40 government coupon for a convertor box.
Indoor rabbit ears and UHF doesn't propagate as well as the longer wave lengths.
I grew up with rabbit ears. I used rabbit exclusively for over 20 years. It's quite a bit more than 20 miles. The biggest factor is strength of the transmitter. You don't know what you're talking about.
There isn't much good on TV these days anyway. Even the PBS stuff is getting kind of shitty - they take an hour to show and explain things that can be done in 10 minutes. Also, some of the nature shows are so depressing:
Waaa! Waaa! Mommy, learning about how the world actually works makes me feel sad. Make it all go away so I can live in ignorant bliss, and then complain when people point out that the real world much more messy than my Ayn Rand books would have me believe.
The observations aren't in question. It's the CONCLUSIONS that are debatable. Jumping to conclusions is NOT IN THE PROVINCE OF SCIENCE, but rather it is a tactic of politicians, and grant chasers.
Ignoring obvious conclusions drawn from both ample repeatable evidence and supported theories to provide an illusion of doubt IS NOT THE PROVINCE OF SCIENCE EITHER, but rather is the tactic of politicians and vested interests in the status quo.
The spec does not support 1 amp. If you want to talk about manufacturers going off on their own to extend the spec in a proprietary fashion, I think you lose the usefulness of the standard USB interconnect. A good example is the Macbook Air cdrom, which works with nothing except the usb on the Macbook Air.
The only thing you need is multiple usb hubs routed to one connector. You can do this with a Y-splitter. I own two, and carry one with me incase I need two connect a usb external drive to my 2005 powerbook g4. Works great.
As for the macbook air superdrive, you're still wrong. The reason why the macbook air superdrive doesn't work is because there's a firmware check on the ide to usb controller board. Power has nothing to do with it, but branding does. For $9, you can the replace the board and it will work just fine in everything.
I'd like to point out that NASA is merely loaning the classroom space. This does not imply agreement or support from all corners of said administration.
I never thought it was an endorsement by NASA, but good point. NASA Ames will rent out big wings of the the building at Moffett as long as your checks clear.
What educational value beyond the current educational technology (e.g. books and mentoring) does VR actually give you? And keep in mind it has not only provide some value, but a value that offsets the increased cost of the VR environment to begin with? Wouldn't in the past 20 years we'd actually see these immersive VR environments in at least some form? The only thing close has been flight simulators, and they've been around for at least 60 years.
Games? Tried that. No body wanted to purchase the hardware when a home console was almost just as good.
Personally, I feel the label "religion" is a bit inappropriate whenever log-log plots are a crucial part of the pitch. Feel free to disagree.
Okay, pseudoscience
Bonus counter point to your log-log plot remark: Scientology has 1950s scifi styled lie detectors as a crucial part of their religion. And the Catholic church one time published long treaties on just how many angels could physically dance on the head of a pin, so you see the trappings of science, don't make it science.
Not really. It's pseudoscience. It's as John Horgan so succulently called it, "The Rapture for nerds," or as Mitch Kapor devastate said, "[It's] Intelligent design for the IQ 140 people."
Kurtzweil takes makes some safe predictions, then makes the same almost believable prediction that's been made for the past 20 years -- that full immersion VR is just around the corner (a dead concept, since there really isn't any compelling reason for it; and let's face it, it's the 1980s version of the atomic-powered-flying-cars prediction.), then applies exponential growth to predict that by 2050 everyone will have their minds uploaded into the computer that used to be called the planet Earth, and then will transport themselves at superluminal velocites to form nanofog bodies throughout the universe, and everywhere Humans++ go, the matter will transmuted into a giant computer. Seriously. We're never going to die. And conveniently -- most conveniently -- this immortality is going to come around just in time to save Kurzweil and Aubrey de Grey's own lives. (Funny how everyone that has predicted that actual immortality will be here Real Soon Now(TM) always makes the prediction that it will save their own lives.)
Kurzweil's assumption of exponential advancement of general technology (Let's assume for the sake of argument that historical advancement is true, as per his slide.) will continue ad infinum, or at least the 50 years he says it will take to convert the entire universe into one big computer is preposterous on prima facia. It's as if Kurzweil never heard the story about why assuming exponential growth forever is bad. As the fortune file recounts:
It is reported that in 1977 there were 37 Elvis impersonators in the world. In 1993 there were 48,000. At this rate, by the year 2010 one out of every three people in the world will be an Elvis impersonator.
It's all predictions that don't even make sense. I don't understand why anyone, let alone anyone even halfway scientifically minded, would take him seriously.
Most of our desktops in my lab are named after local cities and towns, except for one which is named "<job designation of user1", and our servers, which I am not kidding are named: < acronym-of-lab > < acronym-of-lab >1 < acronym-of-lab >-1
Yeah, not only do the servers have names that differ from by a single character, but the names suffer from verbal name collisions, leading to comments like: "Did you mean 'one' or 'dash one'?"
Someone really ought to take away naming rights from my advisor. Worst Naming Scheme Ever.
You had me right up until you said "more competition, cheaper prices." Sorry dude. You've got four major established players, and you'd naively think this would mean vigorous competition, but in fact there's littleevidenceforthisidea, and more troubling, there's evidence for collusion. (Shock! Gasp! A free Market perfect information perfect competition, doesn't and can never exist in actuality? Sob! Genuflecting at the Alter of Free Markets has done nothing! Cartels exist? Cry! Suicide!)
Of course gconf isn't really a solution now is it? It's just a bunch undocumented key-value pairs. Telling users to stop what they're doing, waste a few hours in forums and reading source code to learn that org.gnome.gkrabble.xyzzy should be set to 17 instead of 0 in order to enable scrollwheel events is totally useful.
GNOME's UX sucks because they don't know what they're doing. Havoc read a Sun study back in 1999 that said, "Having five clocks with names such as 'Clock', 'Another Clock' is confusing. As is having options like 'Enable CABAC Decoding'," and thought the take away was "Remove all options". Now it's justified as being useful for users that have never used a computer before, but it's 2009. Those people were hard to find 10 years ago when that study was performed, it's damn near impossible to find them today.
Couple this with a tendency to copy all the worst aspects of Windows and Mac (GTk File Chooser, I'm talking about you!) you have a complete clusterfuck of a project.
Slashdot Mirror
Exactly. I'm suggesting that he's solving the wrong problem.
You know it really isn't that hard to document code as it's being written. Seriously. Just put four or five lines at the top of the function saying what it does, what it takes, and what it returns. A single line before each "paragraph" of code does wonders.
Putting it off later not only makes the documentation harder (stale code is hard to read), but all but ensures that it will never happen.
Honestly, it takes 5 seconds.
That's interesting at shdh30 I was talking with this one guy who pointed another guy across the room. Apparently the other guy was fine engineer, but constantly complained about how the code "wasn't clean enough" and the design "wasn't right" and all the other standard moans of engineers, but to the point that he would constantly get in fights with the other engineers. (There's a certain level of mess and incompleteness/wrongness you have deal with. Nothing is perfect, not even your own work. Eventually, you have to grow up and learn to deal.) So the guy quit to do consulting on his own where he only takes projects that he can write all by himself.
The thing that weirded me out was the fact that the guy I was talking to genuinely admired antisocial engineer. My comment was, "Huh. I wouldn't hire him. There's six guys in this room that that could do his job without having me put up with shit."
And how do you exactly propose to convert this wiki into a PDF?
If you have four people all trying to edit the same document, you've got a crap document. The average paper is 8 to 12 pages including all figures and citations. Journal articles are like 20. One person writes it, the other people review it and send changes. If a big section needs to be written. It's sent to the main author who rewrites it if needed. If you don't do this, you get an article that looks like it was written by four people.
Word is a piece of crap when it comes to equations and citations. Also, you're constantly stuck screwing around with the formatting which really is something that the editor should mess with. Yes, there's the "style" menu in Word, but no one ever uses it, and ensuring that styles are consistently used throughout the doc is quite error prone.
Judging by the fact that my world history book in 1994 ended with the Camp David Accords, no.
Well the problem with #3 is that each potential planet object has to get larger the further from the star it is before it can clear its orbit. If you put Earth out at the orbit of Pluto, it wouldn't be able to clear its path either. Strictly speaking, Earth hasn't it cleared its orbit either.
The most damning thing I found about the IAU's vote wasn't the criteria. It all sounds reasonable, albeit with the problem of requiring "planets" to increase in mass with orbital distance. It's the fact that The "planets" of our solar system aren't actually defined by the criteria but rather a footnote! As seen in the video of the vote, the IAU made clear that this planet definition is only applied for our solar system. One astronmer even asked, "Well then, why don't we just vote on the footnote then?", and the answer back from the chair was essentially, "Well because."
And don't even get me started with the requirement that the definition not include "too many" objects. Essentially, they just wanted eight.
It's not exactly the best definition.
There's potential value. Facebook has gobs of potential value, but they've never shown they know what to do with it. Even if assume that they're breaking even, they're leaving tons of money on the table. They know all about each user's demographics, likes and dislikes, usage rates, and the same stuff about all his/her friends and the strength of connections across the social graph, and yet FB is still serving gay dating ads to straight men, misspelled "Test your IQ" scams, and Russian mail order bride sites. These are spam ads!
They don't know what they're doing, and some grownup needs to oust Mark Zuck and install someone that actually knows how to make money. The fact that FB hasn't shown they know how to extract revenue is hurting them. Deals are falling apart because FB wants to pay in stock, and there's growing consensus that FB's privately held stock isn't worth nearly what FB says it is.
If I was making a deal with FB, I'd insist on cash.
Thus the $40 government coupon for a convertor box.
The DTV conversion. The spectrum giveaway worked perfectly.
I grew up with rabbit ears. I used rabbit exclusively for over 20 years. It's quite a bit more than 20 miles. The biggest factor is strength of the transmitter. You don't know what you're talking about.
Waaa! Waaa! Mommy, learning about how the world actually works makes me feel sad. Make it all go away so I can live in ignorant bliss, and then complain when people point out that the real world much more messy than my Ayn Rand books would have me believe.
You're right. True scientists know that there's no such thing as evidence supported conclusions .
"I believe that nicotine is not addictive."
Like cigarettes, the debate has been over for years. (skip to page 38 for smoking gun)
Ignoring obvious conclusions drawn from both ample repeatable evidence and supported theories to provide an illusion of doubt IS NOT THE PROVINCE OF SCIENCE EITHER, but rather is the tactic of politicians and vested interests in the status quo.
The only thing you need is multiple usb hubs routed to one connector. You can do this with a Y-splitter. I own two, and carry one with me incase I need two connect a usb external drive to my 2005 powerbook g4. Works great.
As for the macbook air superdrive, you're still wrong. The reason why the macbook air superdrive doesn't work is because there's a firmware check on the ide to usb controller board. Power has nothing to do with it, but branding does. For $9, you can the replace the board and it will work just fine in everything.
And they're all completely unenforceable.
Now in Washington...
I never thought it was an endorsement by NASA, but good point. NASA Ames will rent out big wings of the the building at Moffett as long as your checks clear.
What educational value beyond the current educational technology (e.g. books and mentoring) does VR actually give you? And keep in mind it has not only provide some value, but a value that offsets the increased cost of the VR environment to begin with? Wouldn't in the past 20 years we'd actually see these immersive VR environments in at least some form? The only thing close has been flight simulators, and they've been around for at least 60 years.
Games? Tried that. No body wanted to purchase the hardware when a home console was almost just as good.
It's not a religion. It's pseudoscience.
Okay, pseudoscience
Bonus counter point to your log-log plot remark: Scientology has 1950s scifi styled lie detectors as a crucial part of their religion. And the Catholic church one time published long treaties on just how many angels could physically dance on the head of a pin, so you see the trappings of science, don't make it science.
Not really. It's pseudoscience. It's as John Horgan so succulently called it, "The Rapture for nerds," or as Mitch Kapor devastate said, "[It's] Intelligent design for the IQ 140 people."
Kurtzweil takes makes some safe predictions, then makes the same almost believable prediction that's been made for the past 20 years -- that full immersion VR is just around the corner (a dead concept, since there really isn't any compelling reason for it; and let's face it, it's the 1980s version of the atomic-powered-flying-cars prediction.), then applies exponential growth to predict that by 2050 everyone will have their minds uploaded into the computer that used to be called the planet Earth, and then will transport themselves at superluminal velocites to form nanofog bodies throughout the universe, and everywhere Humans++ go, the matter will transmuted into a giant computer. Seriously. We're never going to die. And conveniently -- most conveniently -- this immortality is going to come around just in time to save Kurzweil and Aubrey de Grey's own lives. (Funny how everyone that has predicted that actual immortality will be here Real Soon Now(TM) always makes the prediction that it will save their own lives.)
Kurzweil's assumption of exponential advancement of general technology (Let's assume for the sake of argument that historical advancement is true, as per his slide.) will continue ad infinum, or at least the 50 years he says it will take to convert the entire universe into one big computer is preposterous on prima facia. It's as if Kurzweil never heard the story about why assuming exponential growth forever is bad. As the fortune file recounts:
It's all predictions that don't even make sense. I don't understand why anyone, let alone anyone even halfway scientifically minded, would take him seriously.
Most of our desktops in my lab are named after local cities and towns, except for one which is named "<job designation of user1", and our servers, which I am not kidding are named:
< acronym-of-lab >
< acronym-of-lab >1
< acronym-of-lab >-1
Yeah, not only do the servers have names that differ from by a single character, but the names suffer from verbal name collisions, leading to comments like: "Did you mean 'one' or 'dash one'?"
Someone really ought to take away naming rights from my advisor. Worst Naming Scheme Ever.
You had me right up until you said "more competition, cheaper prices." Sorry dude. You've got four major established players, and you'd naively think this would mean vigorous competition, but in fact there's little evidence for this idea, and more troubling, there's evidence for collusion. (Shock! Gasp! A free Market perfect information perfect competition, doesn't and can never exist in actuality? Sob! Genuflecting at the Alter of Free Markets has done nothing! Cartels exist? Cry! Suicide!)
Of course gconf isn't really a solution now is it? It's just a bunch undocumented key-value pairs. Telling users to stop what they're doing, waste a few hours in forums and reading source code to learn that org.gnome.gkrabble.xyzzy should be set to 17 instead of 0 in order to enable scrollwheel events is totally useful.
GNOME's UX sucks because they don't know what they're doing. Havoc read a Sun study back in 1999 that said, "Having five clocks with names such as 'Clock', 'Another Clock' is confusing. As is having options like 'Enable CABAC Decoding'," and thought the take away was "Remove all options". Now it's justified as being useful for users that have never used a computer before, but it's 2009. Those people were hard to find 10 years ago when that study was performed, it's damn near impossible to find them today.
Couple this with a tendency to copy all the worst aspects of Windows and Mac (GTk File Chooser, I'm talking about you!) you have a complete clusterfuck of a project.
And Yes, I am a UX expert.
Apparently not.
Signed, 1000 Dead Palestinians, and 6 dead Israelis (3 killed by friendly fire)