It is *free* and *available* for 98% of the world. The last 2% includes linux, bsd, plan 9, etc.
Note how you listed more desktop OSes in your "last 2%" than support Silverlight for those "98%".
Oh, and does it support iPhone? Android? There are large chunks of the world, significantly more than 2%, for which their phone is their computer. That's not an exaggeration, either -- people who actually do have a phone, but no desktop, laptop, etc.
You know what? You made a concious choice to use an OS that is not only in the minority, but is miniscule in use compared to win and osx. You knew that
What does this have to do with the ideals of free information and education?
so stop whining when a company makes a product
And herein lies the problem. I wasn't whining about Netflix, I just don't use it.
But to buy the rights to these lectures and to put them up in a form which requires a fucking corporate product in order to watch them, when it would have been exactly as much effort to slap a Creative Commons license on them and put them up in a form that everyone can use, is just insulting.
To tell that last 2% to shut up and stop whining because you deliberately made a conscious choice to support 98% instead of 100% is asinine.
And before you go off your horse again, I use FreeBSD most of the time
That's exactly as relevant as saying you have a gay friend and expecting it to excuse a homophobic rant. I don't care if your OS is made of Richard Stallman's tears, your argument is invalid.
it is their perogative, not mine.
They have every right to be assholes, yes. And I have every right to call them on it.
And you have every right to read my entire fucking post, and take some extra time to tell me to "stop whining". It just makes you kind of a dousche.
If you don't live decently close to the equator solar panels on your roof are nothing but an expensive status symbol.
Actually, small amounts of wind and solar can and do pay for themselves. It doesn't make you independent -- you still need the grid to compensate for a cloudy, calm day -- but when you generate an excess of power, your meter runs backwards and the power company pays you. Even if this doesn't happen, it's still reducing your power bill significantly.
Also, while I agree with your overall point,
a coal miner here a gas worker there and every now and then someone dies installing panels on their roof.
If we weren't dependent on coal and gas, maybe these could be avoided, but if people weren't installing those panels on your roof, surely they'd be doing something very similar. Maybe installing a satellite dish on someone else's roof.
It also makes your argument sound very, very weak when you complain that people react irrationally to one big, scary anecdote, but you haven't actually provided any statistics for how many coal, gas, or solar workers die every year, nor do you seem to be considering wind -- or even sources like hydro, which would make your point much stronger.
What does FOSS have for a web framework that is a viable alternative to Silverlight or Flash?
What about HTML5 isn't viable?
More than that, how about not making it streaming-only? While I'd prefer a free codec, I can play pretty much anything in mplayer or VLC, if you give me a download URL (or a torrent). And these are things I'd want to keep around.
...that only works for Flash, and you all hate Flash too.
Well, it's tricky. In theory, I like Silverlight better than Flash, because Moonlight seems to be much more stable and complete than Gnash. But in practice, there actually is a native Flash player for Linux, and the nspluginwrapper crap isn't really worse than Flash in a 32-bit browser, which is all you get on Windows anyway -- whereas both Moonlight and Gnash only work on a ridiculously small subset of the Silverlight and Flash content out there.
Add to this the fact that the DRM in Silverlight does not work on Moonlight, so while this particular site might work, Netflix, for example, will not. So even if Moonlight was flawless, you'd still have content that requires the official Silverlight.
And if that wasn't enough, with the few videos I've watched, Moonlight didn't do anti-aliasing. I think Silverlight did, but I'm not sure. Flash does, and you better believe mplayer does.
having competing (albeit commercial) frameworks to choose from is a Good Thing[TM] IMHO.
Nope. Having multiple competing implementations is a Good Thing. Having multiple competing standards is a problem, especially when several of them are proprietary. I have no problem that IE exists, so long as we can develop to web standards and, with minimal hackery, have our websites work on all major browsers, including IE. I did have a problem when IE was the defacto standard.
Where's the FOSS alternative, and which major site's require me to use it for the best experience?
Erm, since when did we judge standards based on which ones we're forced to use? WTF makes you think that's a good criterion?
By that logic, the fact that so many apps force you to use Windows means Windows should be the standard, and people should stop bashing it, and nobody should complain if these Feynman lectures -- or, for that matter, our tax forms -- are Windows-only. (Right now, they're Flash-only, which is an improvement, but still retarded.)
Free download, but only available for Windows and OS X. If you're on Linux, there's Mono, but that tends to lag behind -- I usually have to get some bleeding-edge version whenever I actually need some Silverlight content. And contrary to popular belief, neither Windows nor OS X is "free".
What's insulting about this, especially to Feynman's legacy, is that there's a very simple right way to do this: HTML5. And that actually is behind a free download -- Chrome, Firefox, etc, assuming you don't already have a browser capable of playing it. Or, for that matter, multiple technologies at once, if you're afraid of the codec issue -- put it in, say, H.264, then you should be able to develop Flash and Silverlight shims for browsers which don't support H.264 in HTML5.
Well, cool, but how useful is it, really? Of the factors you mentioned, I don't know of anyone using 5, 10, 15, or 30 -- that is, no one talks about needing exactly twelve minutes, or four minutes, etc. That leaves us with intervals of 5 minutes, 10, 15, 20, 30, and 60.
Of those, the problematic ones are 5, 10, and 20, and I contend that the 5 and 10 minutes are less about dividing the hour into one sixth or one twelfth, and more about multiplying the minute by the very decimal values of five or ten. I could make the same argument about twenty, but we actually _do_ see people wanting to divide the hour into thirds -- though here, 3.3 and 6.6 (or 33 and 66) would be close enough. After all, if we want to go by number of factors alone, 60 only buys us two more factors than 100.
60 seconds per minute seems even less useful, aside from the symmetry with sixty minutes per hour. Problem with that argument is that there aren't 60 hours in a day, there are 24. So, again, it seems entirely archaic and arbitrary, and if we're going with archaic and arbitrary, why not go with a convention that's intuitively easy, well understood, and cuts across all standard units? We have the language already -- instead of hours, we could use decidays. Instead of minutes, millidays, and so on. Definitely easier to go that way and come up with anything meaningful than to start with the current SI unit of a second and go the other way -- a day is then just over 86 kiloseconds? That's not terribly useful.
I agree that stubborn pig-headedness is why we don't change, but that doesn't change the fact that I set every temperature source I check to Celsius, and I kept having to check Fahrenheit. I'd basically have to force myself to change by not looking at Fahrenheit, but just going outside.
The kilowatt hour isn't a unit of power. It's a unit of energy. Watts are power. The Kilowatt Hour is a bastardization of the Joule (energy), and a Watt is a Joule per second.
And I can't think of a single reason to prefer horsepower to watts, unless you just like the idea of comparing your car to a horse. For everything else, watts make a lot more sense. Even for energy, much as I prefer Joules for science, everything electrical is already in watts, so it's a lot easier to figure out how many Kilowatt Hours a given appliance will use per month than how many Joules, or how many... horsepower-seconds?
Temperature - Keep Fahrenheit. Celsius is good for Science, but I much prefer a 75 F degree afternoon to a 24 C degree one...
I don't know, a lot of those intermediate ranges are useless. Can you really tell the difference between 40 F and 42 F? But the difference between 22 C and 24 C is a bit more significant.
That said, I never was able to properly train myself to use Celsius. I gave up and just use Fahrenheit, for now.
Sometimes it's a matter of scale -- for instance, units based on electron-volts are useful when you want to talk about the energy in a single photon or electron, or things on that scale, while the Joule is a much more accessible unit when you're at the scale where metric units make sense -- can't get much simpler than a kilogram meter per second. It's not just a matter of a nano-Joule vs a Joule -- it's a matter of the electron-Volt being based on how elementary particles actually behave, while the Joule is based on fairly arbitrary (but convenient) metric units.
Sometimes it's a matter of who is using the unit, and what they're using it for. I often talk shit about the Kilowatt Hour -- Watts are Joules per second (energy per time) and the Kilowatt Hour is Kilowatts per Hour (power per time, where power is energy per time) -- so you end up back at energy. The Kilowatt hour is basically a really clumsy multiple of the Joule -- or at least, it's really clumsy if you're dealing with Joules, which would imply you're dealing with physics and engineering. The fact that electrical appliances are rated in watts means that a kilowatt hour is still quite convenient if you want to know, say, how much it's going to cost to run a box with a 250-watt power supply 24/7, or a 700-watt, 70-inch HDTV for a few hours a day, or how long it'll take for a CFL to pay for itself, and so on.
Degrees Celsius vs Kelvin. Kelvin is a lot more useful if you need to do actual calculations -- again, physics/engineering -- but the difference between 273 kelvin and 313 kelvin doesn't really mean as much as the difference between 0 degrees Celsius and 40 degrees Celsius.
Or angle measures -- degrees are much easier for humans to work with than radians when just trying to figure out the angle, but radians are a much more natural angle to do any sort of calculations in, especially since they technically aren't even units. You can do crazy things like take that 7200 RPMs your hard disk spins at, convert it to radians/second, and multiply it by the radius of your hard disk in whatever units you want, and you'll get the linear velocity of the edge of that disk in those same units.
That is, 45 degrees is a lot easier for humans to learn than pi/4 radians, but if you know you've got pi/4 radians, that's a lot easier to apply to almost anything.
I could go on, and that's just off the top of my head, from what is theoretically a freshman physics course.
None of this, by the way, is a justification for imperial measurements. Those are just retarded. Unfortunately, I haven't been able to learn Celsius for temperature -- I know what 40 degrees Fahrenheit feels like, but I have no idea (until I convert it) what 5 degrees Celsius feels like. Still, I'd be the first to suggest any shift towards better units -- maybe while we're at it, we can fix the whole minute/hour/day weirdness and start dividing the day by powers of 10 instead.
Thermite is cheap. Granted, a device capable of actually holding a melting hard drive might be more expensive, but I have to imagine that taking a trip to an appropriate location several times a year would be relatively cheap. It'd certainly be a lot more fun.
Based on 'ar'? Really? Written in C/C++, not easily hackable. Then there's apt -- things like apt-cache work great in theory, but I never got it working in practice.
I came to Ubuntu from Gentoo. I don't miss having to compile everything, but I do miss packages being dirt-simple bash scripts that any idiot could write. Rubygems is similarly easy to work with.
And while I'm not sure I see any other distros doing better (aside from maybe Gentoo, occasionally), there's the fact that with dpkg, you need to re-download an entire copy of whatever it is -- patches, binary or otherwise, aren't actually delivered as patches.
Note, I'm not really saying that these are things "wrong" with dpkg, just that calling Ubuntu "the edge" isn't true here.
Was Pulse Audio to soon? Yes but on the other hand, other distro's still think OSS is cutting edge...
The distro I was using before Ubuntu had no problem with ALSA. Come to think of it, I've been wondering if it isn't time to go back to ALSA for awhile. After all...
And now that Pulse Audio does work,
For some values of "work". Loading up a web game in Chrome shouldn't be able to bring my audio system to 100% CPU.
All the whiners complain about Pulse Audio and STILL use Ubuntu because...
Because it's worked reasonably well, and because I've been too lazy to try the others.
As for long term support. Just fucking upgrade already.
This is the bit that made me reply. Really? The whole point of LTS is for you to not have to "just fucking upgrade", so that you don't need to keep up with the 6-month release cycle. I don't use it on my laptop, but I do tend to give each release a few months, because they always break something huge for no apparent reason.
Ubuntu pushes the edge.
No, Ubuntu ends up somewhere in the middle. They certainly aren't "the edge" -- they're still using dpkg, FFS -- but they do frequently push random crap that's nowhere near ready into my distro. That's not "desktop", by the way -- most desktop users would like their GUIs to be reasonably stable and consistent, and not randomly crash or lag.
If I wanted "the edge", I'd be using a beta Ubuntu, or I'd be on Gentoo or Arch, or trying something more exotic like Gobo. I certainly don't recall any of these thinking OSS was cutting edge, and that was years ago. Then again, I don't recall any of these cutting the OSS compatibility layer from the kernel so that I now have to find OSS apps and launch them with a "padsp" wrapper. Seriously, WTF are you doing forcing a desktop user to understand the difference between OSS, ALSA, and PulseAudio, and explaining why most things work, and Lugaru doesn't, and how to fix it by only launching Lugaru with "padsp lugaru" from the commandline?
I miss reading a Slashdot article about a 0-day (within hours of the actual vulnerability), then going to patch it and discover I'd already patched via my distro's repository.
3. belief in god or in the doctrines or teachings of religion
4. belief in anything, as a code of ethics, standards of merit, etc.
...and so on.
There are multiple distinct definitions here. I was wrong to say that it isn't "faith" by any definition, but the definitions here are talking about entirely different things. Take the above list -- do you really think definition 1 is talking about the same thing as definition 2?
Religion is a subset of all things in which one can have faith.
But is that faith based on definition 2 or definition 1? If it's definition 2 faith, we need some evidence that religion actually works, that it actually is the truth, and that we can rely on it for, well, anything.
If it's definition 1 faith, then that's not the same thing at all, as I hope I've already demonstrated: My trust (or "faith" if you insist) in science is based on evidence, including the repeatable test I outlined for exploring and testing the veracity of and support for scientific claims.
Saying you don't understand, nor is it possible for you to understand, all aspects of science, thus you must take some of it on faith (in the scientific method and the people that are applying it) has absolutely nothing to do with religion.
I think we agree on this much, but the problem is when people make an equivocation fallacy and want to say that my "faith" in science is of exactly the same sort as their own religious faith. This article, and many commenters here, present exactly this fallacy, in a form which ultimately boils down to "We all have faith, you have faith in science and I have faith in God, therefore religion is at least as valid as science."
It's the same sort of problem as saying that evolution is "just a theory."
I suppose I'll just ask that you trust me when I say these are issues which are still better handled by a plugin.
...no, not really, not without at least this much:
(mixing, looping to specific points in a track, synchronization, etc.).
Mixing works fine. Looping and synchronization, I can see, though they really shouldn't be that far off -- and I do see people doing both mixing and looping without plugins.
the web isn't ready to completely shed itself of plugins.
Here's where I disagree: Yes, there are things you can do with a plugin that you cannot (yet) do with straight HTML. However, as soon as you introduce that plugin, you lose a lot of the advantage of making it a web app in the first place. If that plugin is proprietary, you've also guaranteed that you are going to make things worse as long as this plugin's alive, and that whatever it's replaced with will be different enough so as to not be compatible without some sort of wrappers.
From my perspective, if it can't be done without plugins, it's probably not worth doing on the Web in the first place. More often, it can be done without plugins, but it's harder.
So what I'm saying here is that I don't see this as the web shedding itself of anything, I see plugins as not part of the web in the first place, and their continued attempts to encroach on that territory are damaging to what the web actually is.
After all, replace Flash with a custom ActiveX control. Do you still feel like the Web isn't ready to shed itself of ActiveX? Or is it that ActiveX is not and never was something that belonged on the Web?
Ultimately, this has forced me back to using a Flash plugin for audio support, simply because I don't feel like storing multiple copies of every sound file
...or you could store a format that Flash can read also, and provide that in audio tag format for browsers which support it. And personally, between Flash and multiple copies, I'll take the multiple copies.
The whole web-as-a-platform thing has indeed come a long way, but I don't think we're ready for a plugin-free web experience yet.
The fact that others have been able to make this work tells me that maybe it's still painful to develop on, but it can still work. And the problems you've described sound fairly mild, compared to real cross-platform issues if I tried to, say, keep a common codebase between.NET/Silverlight, Mono/Moonlight, a desktop.NET client, an xbox...
I do, however, have evidence that this is the case.
Whenever I've wanted to verify a scientific idea, I can. Whenever I read into it, I find all sorts of documentation of what sorts of experiments were performed and how the results were interpreted, and when I've performed the experiments myself, I've gotten very nearly the same results. (Earlier on, I wasn't quite as skilled in the mechanics of lab work, so my results were less accurate and less consistent, though they still tended to suggest the same thing.)
More than that, I find that science tends to converge. While it is always improving itself, and always changing, the ideas which are generally accepted today (particularly those which have stood the test of time) tend to be the result of convergent lines of evidence, and this is also something I can verify. When a thousand scientists study the same phenomenon, with a thousand different opinions on what they think will happen, you're still not going to get a thousand different results -- you're going to get one, maybe two if you're particularly unlucky.
So, I don't know a lot about quantum physics, for instance. I do, however, know things about motion, and friction, and gravity, and electromagnetism.
Even as a layperson, there is a sense in which I perform dozens (hundreds? thousands?) of experiments every day, just by using technology. Even if I didn't understand how this computer worked -- and I seem to be getting a clearer picture all the time -- I do understand that it was created by scientists who were in the process of doing science. I may not necessarily know that they are telling me the truth about science, but I do know that science works. If there wasn't anything to science, I wouldn't be able to post here -- that, or Dell is composed of magicians rather than technicians, which really, really doesn't seem likely.
Now, compare this to what people tend to mean when they say "faith."
There is a fantastic diversity of religions -- Christianity, Hinduism, Shintoism, Taoism, Islam, Judaism, Scientology... But even ignoring that, there are at least as many varieties of Christianity as there are Christians. I hardly have to study religion to know that it isn't science; I can perform the simple experiment of asking two believers some fundamental questions about their religion.
Try this experiment: Ask two scientists about the difference between charge and voltage. Or ask them whether the speed of light is constant. Or ask them which direction the friction force on a ball rolling down a slope is.
Now, ask two believers whether those who have never heard of Jesus will go to hell. Or ask them which passages of the Bible were meant to be taken literally -- did God really speak the world into existence six thousand years ago, or was it a day-age thing, or was it purely allegorical? Is it "once saved, always saved," or can a former believer go to Hell? Is Homosexuality a sin? Is Jesus God, was he a prophet, or was he just a guy with some good ideas?
Given this background on the reliability of science, it's not faith, it's trust. I generally trust the scientific consensus, because every time I try to verify it, I find the evidence to be absolutely solid, while every time I try to even get consensus on a religious claim, I get no agreement on whether it's true or not, or if it is true, how to verify it, and the procedures suggested don't work.
I understand how the Web works, and that a well-designed website can also function as a GUI. I think you underestimate just how broken it can be, unless the method you're suggesting actually drives a browser. Suppose the web GUI is trying to do everything in JavaScript, with fancy AJAX calls and the like? Reverse-engineering that can't be easier than running a well-documented, well-understood CLI.
And at the point where you're talking about discovering the name of a GUI widget, understand that you are, in fact, trying to script something which was never intended to be scripted. Yes, it can be done, but as arcane as the shell can be, I can't imagine it approaches this.
I didn't mean to imply that this is impossible without an API, but I do think you need an API to bring it up to the convenience and reliability you get almost for free from a well-designed commandline environment -- or from the better Web frameworks, I will admit.
If you're going to claim Android isn't real Java, you may as well mention that Mono isn't real.NET.
And "maybe"? Keep in mind that Mono includes C dependencies. What a platform like.NET/Java buys you is easier portability and a kinder runtime, but the only place you can run a.NET or Java app that you can't run a C app is where you're deliberately sandboxed.
As for web stuff, the web is actually a decent platform now. I wish people would stop fucking that up with plugins. If you really want to make a browser game, make an actual browser game -- all the major mobile platforms have real browsers now, so it shouldn't be a problem.
That depends how big and ugly the diagram is, and whether it's something I'm doing manually (for planning purposes) or automatically (just how is this network configured right now?)
It's not always the best tool, but Graphviz is pretty cool.
Slashdot Mirror
Stop already ok?
No. Stop reading, if you have a problem.
It is *free* and *available* for 98% of the world. The last 2% includes linux, bsd, plan 9, etc.
Note how you listed more desktop OSes in your "last 2%" than support Silverlight for those "98%".
Oh, and does it support iPhone? Android? There are large chunks of the world, significantly more than 2%, for which their phone is their computer. That's not an exaggeration, either -- people who actually do have a phone, but no desktop, laptop, etc.
You know what? You made a concious choice to use an OS that is not only in the minority, but is miniscule in use compared to win and osx. You knew that
What does this have to do with the ideals of free information and education?
so stop whining when a company makes a product
And herein lies the problem. I wasn't whining about Netflix, I just don't use it.
But to buy the rights to these lectures and to put them up in a form which requires a fucking corporate product in order to watch them, when it would have been exactly as much effort to slap a Creative Commons license on them and put them up in a form that everyone can use, is just insulting.
To tell that last 2% to shut up and stop whining because you deliberately made a conscious choice to support 98% instead of 100% is asinine.
And before you go off your horse again, I use FreeBSD most of the time
That's exactly as relevant as saying you have a gay friend and expecting it to excuse a homophobic rant. I don't care if your OS is made of Richard Stallman's tears, your argument is invalid.
it is their perogative, not mine.
They have every right to be assholes, yes. And I have every right to call them on it.
And you have every right to read my entire fucking post, and take some extra time to tell me to "stop whining". It just makes you kind of a dousche.
If you don't live decently close to the equator solar panels on your roof are nothing but an expensive status symbol.
Actually, small amounts of wind and solar can and do pay for themselves. It doesn't make you independent -- you still need the grid to compensate for a cloudy, calm day -- but when you generate an excess of power, your meter runs backwards and the power company pays you. Even if this doesn't happen, it's still reducing your power bill significantly.
Also, while I agree with your overall point,
a coal miner here a gas worker there and every now and then someone dies installing panels on their roof.
If we weren't dependent on coal and gas, maybe these could be avoided, but if people weren't installing those panels on your roof, surely they'd be doing something very similar. Maybe installing a satellite dish on someone else's roof.
It also makes your argument sound very, very weak when you complain that people react irrationally to one big, scary anecdote, but you haven't actually provided any statistics for how many coal, gas, or solar workers die every year, nor do you seem to be considering wind -- or even sources like hydro, which would make your point much stronger.
What does FOSS have for a web framework that is a viable alternative to Silverlight or Flash?
What about HTML5 isn't viable?
More than that, how about not making it streaming-only? While I'd prefer a free codec, I can play pretty much anything in mplayer or VLC, if you give me a download URL (or a torrent). And these are things I'd want to keep around.
...that only works for Flash, and you all hate Flash too.
Well, it's tricky. In theory, I like Silverlight better than Flash, because Moonlight seems to be much more stable and complete than Gnash. But in practice, there actually is a native Flash player for Linux, and the nspluginwrapper crap isn't really worse than Flash in a 32-bit browser, which is all you get on Windows anyway -- whereas both Moonlight and Gnash only work on a ridiculously small subset of the Silverlight and Flash content out there.
Add to this the fact that the DRM in Silverlight does not work on Moonlight, so while this particular site might work, Netflix, for example, will not. So even if Moonlight was flawless, you'd still have content that requires the official Silverlight.
And if that wasn't enough, with the few videos I've watched, Moonlight didn't do anti-aliasing. I think Silverlight did, but I'm not sure. Flash does, and you better believe mplayer does.
having competing (albeit commercial) frameworks to choose from is a Good Thing[TM] IMHO.
Nope. Having multiple competing implementations is a Good Thing. Having multiple competing standards is a problem, especially when several of them are proprietary. I have no problem that IE exists, so long as we can develop to web standards and, with minimal hackery, have our websites work on all major browsers, including IE. I did have a problem when IE was the defacto standard.
Where's the FOSS alternative, and which major site's require me to use it for the best experience?
Erm, since when did we judge standards based on which ones we're forced to use? WTF makes you think that's a good criterion?
By that logic, the fact that so many apps force you to use Windows means Windows should be the standard, and people should stop bashing it, and nobody should complain if these Feynman lectures -- or, for that matter, our tax forms -- are Windows-only. (Right now, they're Flash-only, which is an improvement, but still retarded.)
What barrier, a free download barrier?
For some value of "free".
Free download, but only available for Windows and OS X. If you're on Linux, there's Mono, but that tends to lag behind -- I usually have to get some bleeding-edge version whenever I actually need some Silverlight content. And contrary to popular belief, neither Windows nor OS X is "free".
What's insulting about this, especially to Feynman's legacy, is that there's a very simple right way to do this: HTML5. And that actually is behind a free download -- Chrome, Firefox, etc, assuming you don't already have a browser capable of playing it. Or, for that matter, multiple technologies at once, if you're afraid of the codec issue -- put it in, say, H.264, then you should be able to develop Flash and Silverlight shims for browsers which don't support H.264 in HTML5.
Well, cool, but how useful is it, really? Of the factors you mentioned, I don't know of anyone using 5, 10, 15, or 30 -- that is, no one talks about needing exactly twelve minutes, or four minutes, etc. That leaves us with intervals of 5 minutes, 10, 15, 20, 30, and 60.
Of those, the problematic ones are 5, 10, and 20, and I contend that the 5 and 10 minutes are less about dividing the hour into one sixth or one twelfth, and more about multiplying the minute by the very decimal values of five or ten. I could make the same argument about twenty, but we actually _do_ see people wanting to divide the hour into thirds -- though here, 3.3 and 6.6 (or 33 and 66) would be close enough. After all, if we want to go by number of factors alone, 60 only buys us two more factors than 100.
60 seconds per minute seems even less useful, aside from the symmetry with sixty minutes per hour. Problem with that argument is that there aren't 60 hours in a day, there are 24. So, again, it seems entirely archaic and arbitrary, and if we're going with archaic and arbitrary, why not go with a convention that's intuitively easy, well understood, and cuts across all standard units? We have the language already -- instead of hours, we could use decidays. Instead of minutes, millidays, and so on. Definitely easier to go that way and come up with anything meaningful than to start with the current SI unit of a second and go the other way -- a day is then just over 86 kiloseconds? That's not terribly useful.
I agree that stubborn pig-headedness is why we don't change, but that doesn't change the fact that I set every temperature source I check to Celsius, and I kept having to check Fahrenheit. I'd basically have to force myself to change by not looking at Fahrenheit, but just going outside.
I admit, I've been living under a rock, but what is this Digital Economy Act?
Power - I prefer HP to KWH's
...what?
The kilowatt hour isn't a unit of power. It's a unit of energy. Watts are power. The Kilowatt Hour is a bastardization of the Joule (energy), and a Watt is a Joule per second.
And I can't think of a single reason to prefer horsepower to watts, unless you just like the idea of comparing your car to a horse. For everything else, watts make a lot more sense. Even for energy, much as I prefer Joules for science, everything electrical is already in watts, so it's a lot easier to figure out how many Kilowatt Hours a given appliance will use per month than how many Joules, or how many... horsepower-seconds?
Temperature - Keep Fahrenheit. Celsius is good for Science, but I much prefer a 75 F degree afternoon to a 24 C degree one...
I don't know, a lot of those intermediate ranges are useless. Can you really tell the difference between 40 F and 42 F? But the difference between 22 C and 24 C is a bit more significant.
That said, I never was able to properly train myself to use Celsius. I gave up and just use Fahrenheit, for now.
When they came for my 10 inch...
...it became over 25 centimeters, and you're complaining about that?
Yes, there are some.
Sometimes it's a matter of scale -- for instance, units based on electron-volts are useful when you want to talk about the energy in a single photon or electron, or things on that scale, while the Joule is a much more accessible unit when you're at the scale where metric units make sense -- can't get much simpler than a kilogram meter per second. It's not just a matter of a nano-Joule vs a Joule -- it's a matter of the electron-Volt being based on how elementary particles actually behave, while the Joule is based on fairly arbitrary (but convenient) metric units.
Sometimes it's a matter of who is using the unit, and what they're using it for. I often talk shit about the Kilowatt Hour -- Watts are Joules per second (energy per time) and the Kilowatt Hour is Kilowatts per Hour (power per time, where power is energy per time) -- so you end up back at energy. The Kilowatt hour is basically a really clumsy multiple of the Joule -- or at least, it's really clumsy if you're dealing with Joules, which would imply you're dealing with physics and engineering. The fact that electrical appliances are rated in watts means that a kilowatt hour is still quite convenient if you want to know, say, how much it's going to cost to run a box with a 250-watt power supply 24/7, or a 700-watt, 70-inch HDTV for a few hours a day, or how long it'll take for a CFL to pay for itself, and so on.
Degrees Celsius vs Kelvin. Kelvin is a lot more useful if you need to do actual calculations -- again, physics/engineering -- but the difference between 273 kelvin and 313 kelvin doesn't really mean as much as the difference between 0 degrees Celsius and 40 degrees Celsius.
Or angle measures -- degrees are much easier for humans to work with than radians when just trying to figure out the angle, but radians are a much more natural angle to do any sort of calculations in, especially since they technically aren't even units. You can do crazy things like take that 7200 RPMs your hard disk spins at, convert it to radians/second, and multiply it by the radius of your hard disk in whatever units you want, and you'll get the linear velocity of the edge of that disk in those same units.
That is, 45 degrees is a lot easier for humans to learn than pi/4 radians, but if you know you've got pi/4 radians, that's a lot easier to apply to almost anything.
I could go on, and that's just off the top of my head, from what is theoretically a freshman physics course.
None of this, by the way, is a justification for imperial measurements. Those are just retarded. Unfortunately, I haven't been able to learn Celsius for temperature -- I know what 40 degrees Fahrenheit feels like, but I have no idea (until I convert it) what 5 degrees Celsius feels like. Still, I'd be the first to suggest any shift towards better units -- maybe while we're at it, we can fix the whole minute/hour/day weirdness and start dividing the day by powers of 10 instead.
Thermite is cheap. Granted, a device capable of actually holding a melting hard drive might be more expensive, but I have to imagine that taking a trip to an appropriate location several times a year would be relatively cheap. It'd certainly be a lot more fun.
Based on 'ar'? Really? Written in C/C++, not easily hackable. Then there's apt -- things like apt-cache work great in theory, but I never got it working in practice.
I came to Ubuntu from Gentoo. I don't miss having to compile everything, but I do miss packages being dirt-simple bash scripts that any idiot could write. Rubygems is similarly easy to work with.
And while I'm not sure I see any other distros doing better (aside from maybe Gentoo, occasionally), there's the fact that with dpkg, you need to re-download an entire copy of whatever it is -- patches, binary or otherwise, aren't actually delivered as patches.
Note, I'm not really saying that these are things "wrong" with dpkg, just that calling Ubuntu "the edge" isn't true here.
This is an odd rant to make it to insightful...
Was Pulse Audio to soon? Yes but on the other hand, other distro's still think OSS is cutting edge...
The distro I was using before Ubuntu had no problem with ALSA. Come to think of it, I've been wondering if it isn't time to go back to ALSA for awhile. After all...
And now that Pulse Audio does work,
For some values of "work". Loading up a web game in Chrome shouldn't be able to bring my audio system to 100% CPU.
All the whiners complain about Pulse Audio and STILL use Ubuntu because...
Because it's worked reasonably well, and because I've been too lazy to try the others.
As for long term support. Just fucking upgrade already.
This is the bit that made me reply. Really? The whole point of LTS is for you to not have to "just fucking upgrade", so that you don't need to keep up with the 6-month release cycle. I don't use it on my laptop, but I do tend to give each release a few months, because they always break something huge for no apparent reason.
Ubuntu pushes the edge.
No, Ubuntu ends up somewhere in the middle. They certainly aren't "the edge" -- they're still using dpkg, FFS -- but they do frequently push random crap that's nowhere near ready into my distro. That's not "desktop", by the way -- most desktop users would like their GUIs to be reasonably stable and consistent, and not randomly crash or lag.
If I wanted "the edge", I'd be using a beta Ubuntu, or I'd be on Gentoo or Arch, or trying something more exotic like Gobo. I certainly don't recall any of these thinking OSS was cutting edge, and that was years ago. Then again, I don't recall any of these cutting the OSS compatibility layer from the kernel so that I now have to find OSS apps and launch them with a "padsp" wrapper. Seriously, WTF are you doing forcing a desktop user to understand the difference between OSS, ALSA, and PulseAudio, and explaining why most things work, and Lugaru doesn't, and how to fix it by only launching Lugaru with "padsp lugaru" from the commandline?
Sure there is: Netflix. YouTube. Online backup.
The fact that you can't come up with a reasonable way doesn't mean that there is no reasonable way.
Could be that it's Slashdotted, not that it's inherently slow.
I miss reading a Slashdot article about a 0-day (within hours of the actual vulnerability), then going to patch it and discover I'd already patched via my distro's repository.
It is built to run on the JVM, uses static typing, and supports high while maintaining a strong focus on being easy learn and easy to read.
Supports high what? Does it just support being high?
Faith is defined as trust in something (or someone).
Before you make up a definition, you could look it up:
Or how about this one:
...and so on.
There are multiple distinct definitions here. I was wrong to say that it isn't "faith" by any definition, but the definitions here are talking about entirely different things. Take the above list -- do you really think definition 1 is talking about the same thing as definition 2?
Religion is a subset of all things in which one can have faith.
But is that faith based on definition 2 or definition 1? If it's definition 2 faith, we need some evidence that religion actually works, that it actually is the truth, and that we can rely on it for, well, anything.
If it's definition 1 faith, then that's not the same thing at all, as I hope I've already demonstrated: My trust (or "faith" if you insist) in science is based on evidence, including the repeatable test I outlined for exploring and testing the veracity of and support for scientific claims.
Saying you don't understand, nor is it possible for you to understand, all aspects of science, thus you must take some of it on faith (in the scientific method and the people that are applying it) has absolutely nothing to do with religion.
I think we agree on this much, but the problem is when people make an equivocation fallacy and want to say that my "faith" in science is of exactly the same sort as their own religious faith. This article, and many commenters here, present exactly this fallacy, in a form which ultimately boils down to "We all have faith, you have faith in science and I have faith in God, therefore religion is at least as valid as science."
It's the same sort of problem as saying that evolution is "just a theory."
I suppose I'll just ask that you trust me when I say these are issues which are still better handled by a plugin.
...no, not really, not without at least this much:
(mixing, looping to specific points in a track, synchronization, etc.).
Mixing works fine. Looping and synchronization, I can see, though they really shouldn't be that far off -- and I do see people doing both mixing and looping without plugins.
the web isn't ready to completely shed itself of plugins.
Here's where I disagree: Yes, there are things you can do with a plugin that you cannot (yet) do with straight HTML. However, as soon as you introduce that plugin, you lose a lot of the advantage of making it a web app in the first place. If that plugin is proprietary, you've also guaranteed that you are going to make things worse as long as this plugin's alive, and that whatever it's replaced with will be different enough so as to not be compatible without some sort of wrappers.
From my perspective, if it can't be done without plugins, it's probably not worth doing on the Web in the first place. More often, it can be done without plugins, but it's harder.
So what I'm saying here is that I don't see this as the web shedding itself of anything, I see plugins as not part of the web in the first place, and their continued attempts to encroach on that territory are damaging to what the web actually is.
After all, replace Flash with a custom ActiveX control. Do you still feel like the Web isn't ready to shed itself of ActiveX? Or is it that ActiveX is not and never was something that belonged on the Web?
Ultimately, this has forced me back to using a Flash plugin for audio support, simply because I don't feel like storing multiple copies of every sound file
...or you could store a format that Flash can read also, and provide that in audio tag format for browsers which support it. And personally, between Flash and multiple copies, I'll take the multiple copies.
The whole web-as-a-platform thing has indeed come a long way, but I don't think we're ready for a plugin-free web experience yet.
The fact that others have been able to make this work tells me that maybe it's still painful to develop on, but it can still work. And the problems you've described sound fairly mild, compared to real cross-platform issues if I tried to, say, keep a common codebase between .NET/Silverlight, Mono/Moonlight, a desktop .NET client, an xbox...
I do, however, have evidence that this is the case.
Whenever I've wanted to verify a scientific idea, I can. Whenever I read into it, I find all sorts of documentation of what sorts of experiments were performed and how the results were interpreted, and when I've performed the experiments myself, I've gotten very nearly the same results. (Earlier on, I wasn't quite as skilled in the mechanics of lab work, so my results were less accurate and less consistent, though they still tended to suggest the same thing.)
More than that, I find that science tends to converge. While it is always improving itself, and always changing, the ideas which are generally accepted today (particularly those which have stood the test of time) tend to be the result of convergent lines of evidence, and this is also something I can verify. When a thousand scientists study the same phenomenon, with a thousand different opinions on what they think will happen, you're still not going to get a thousand different results -- you're going to get one, maybe two if you're particularly unlucky.
So, I don't know a lot about quantum physics, for instance. I do, however, know things about motion, and friction, and gravity, and electromagnetism.
Even as a layperson, there is a sense in which I perform dozens (hundreds? thousands?) of experiments every day, just by using technology. Even if I didn't understand how this computer worked -- and I seem to be getting a clearer picture all the time -- I do understand that it was created by scientists who were in the process of doing science. I may not necessarily know that they are telling me the truth about science, but I do know that science works. If there wasn't anything to science, I wouldn't be able to post here -- that, or Dell is composed of magicians rather than technicians, which really, really doesn't seem likely.
Now, compare this to what people tend to mean when they say "faith."
There is a fantastic diversity of religions -- Christianity, Hinduism, Shintoism, Taoism, Islam, Judaism, Scientology... But even ignoring that, there are at least as many varieties of Christianity as there are Christians. I hardly have to study religion to know that it isn't science; I can perform the simple experiment of asking two believers some fundamental questions about their religion.
Try this experiment: Ask two scientists about the difference between charge and voltage. Or ask them whether the speed of light is constant. Or ask them which direction the friction force on a ball rolling down a slope is.
Now, ask two believers whether those who have never heard of Jesus will go to hell. Or ask them which passages of the Bible were meant to be taken literally -- did God really speak the world into existence six thousand years ago, or was it a day-age thing, or was it purely allegorical? Is it "once saved, always saved," or can a former believer go to Hell? Is Homosexuality a sin? Is Jesus God, was he a prophet, or was he just a guy with some good ideas?
Given this background on the reliability of science, it's not faith, it's trust. I generally trust the scientific consensus, because every time I try to verify it, I find the evidence to be absolutely solid, while every time I try to even get consensus on a religious claim, I get no agreement on whether it's true or not, or if it is true, how to verify it, and the procedures suggested don't work.
I understand how the Web works, and that a well-designed website can also function as a GUI. I think you underestimate just how broken it can be, unless the method you're suggesting actually drives a browser. Suppose the web GUI is trying to do everything in JavaScript, with fancy AJAX calls and the like? Reverse-engineering that can't be easier than running a well-documented, well-understood CLI.
And at the point where you're talking about discovering the name of a GUI widget, understand that you are, in fact, trying to script something which was never intended to be scripted. Yes, it can be done, but as arcane as the shell can be, I can't imagine it approaches this.
I didn't mean to imply that this is impossible without an API, but I do think you need an API to bring it up to the convenience and reliability you get almost for free from a well-designed commandline environment -- or from the better Web frameworks, I will admit.
If you're going to claim Android isn't real Java, you may as well mention that Mono isn't real .NET.
And "maybe"? Keep in mind that Mono includes C dependencies. What a platform like .NET/Java buys you is easier portability and a kinder runtime, but the only place you can run a .NET or Java app that you can't run a C app is where you're deliberately sandboxed.
As for web stuff, the web is actually a decent platform now. I wish people would stop fucking that up with plugins. If you really want to make a browser game, make an actual browser game -- all the major mobile platforms have real browsers now, so it shouldn't be a problem.
parsed the data into URLs that call the web interface for the "GUI" only router,
At that point, the router also has an API. A GUI alone won't cut it.
You wouldn't use a CLI to draw a network diagram.
That depends how big and ugly the diagram is, and whether it's something I'm doing manually (for planning purposes) or automatically (just how is this network configured right now?)
It's not always the best tool, but Graphviz is pretty cool.