i can count the number of rebuilds i've needed to do with windows for my own systems on half of one hand in the past 3 years
I can count the number of rebuilds of Debian I needed to do with my laptop over 10 years using one finger. And that includes upgrades from one version to another. Apparently, I am not the only one. I just read few weeks ago about an issue being fixed for people that upgraded their system from Potato to Wheezy all the way...
Not specifically joe and mtr, but being able to rebuild basic ncurses tools that needs the autotools layer, libc6, and other non-fancy very common libs (zlib, libstdc++, and other stuff like that) is for sure one of the requirements. I'm on purpose avoiding any graphical (X, GTK, Qt, etc.) depends to my definition, but some might not agree. There's about 30k packages in Debian, about 10k in CentOS 5.x, being able to rebuild (or even at least, cross-compile) a vast majority of the non-graphical ones seems a reasonable requirement to be able to be called a distribution. That is for being a *Unix* distro, not even Linux (eg: you aren't bound to a particular kernel here, which I don't think is the important bit). dpkg and rpm aren't even involved, you could well use other types of packages if you like it (see the variety of choice we have in this world: tar.gz with Slackware, deb, rpm, gentoo and FreeBSD ports, and there's all sorts of other less common types of packages (.ipk of openwrt anyone?)). Being bound to a particular language in a device (java, in the case of Android) isn't really helping.
No, windows itself wouldn't, but cygwin for sure IS a GNU distro. Windows is simply capable of running it. Here, you are simply pretending that Linux == kernel and that's it, like many others in this/. entry. If you want to think it this way, then go on. But never ever, the Android platform can be called a "distribution" (Linux or not).
See how much a waste of time it would have been to state arguments that you know already! Yourself, you know already that running Linux is all about the GNU tools and all the libraries we have, and the ease of "porting" one app to another arch/platform. In Maemo, I just typed "dpkg-buildpackage" and I had both joe and mtr available for my phone, and I use them both very often. Until I can do that in Android, you can't call it a Unix distribution (and yet, even less a Linux distro.).
Do you know how selling software works? If you just throw together a.deb or.rpm, and "be done with it", you'll find an inbox full of angry customers demanding support. Oh, you packaged your.deb based on Debian Squeeze? Well, package foo upon which you software was tested doesn't work the same in Ubuntu Natty Narwhal. And never mind all the requests from Gentoo users who want to know which flags they need to set to recompile their software to fix some obscure bug that pops up when trying to run the software as installed by your.rpm.
Has someone every said you should ship ONE type of package? It isn't that hard to build binaries for both Debian and Ubuntu, it's just a dpkg-buildpackage away, and virtualization (like virtualbox) makes it *very* easy to have many build platforms. Now, let's say you want to cover 90% of the cases, then you'd be doing packages for Debian, Ubuntu and CentOS, with both 32 and 64 bits. In fact, that's down to doing one deb and one rpm, and then build. Yet, you might find some funny distro like Gentoo, you are right, and "they" wont like it, but you can still live with it... Also, most of the time, there's no ABI breakage, so it will work even without recompiling (see the horrible work that Skype did: even if it's horrible job, it does work).
Oh, and where does your package install to? Is it/usr/local/? Or maybe/opt/? Wherever you pick, you will piss off someone who wants it somewhere else.
There's a standard filesystem hierarchy here, it's not about pissing somebody off, it's about respecting standards, like having your configuration files in/etc, your read only files in/usr, and so on. There's no reason to put anything in/opt unless you are on the non-brainer Maemo platform, or things in/usr/local unless you are writing a FreeBSD port.
And for systems that aren't based on.rpm or.deb packages, so they don't have proper dependancies? *You* might say they are ignorable, they *they* sure won't see it that way.
But *you* might don't care about 5% of the users if you've covered most of the others with deb/rpm. So let them say...
Most Linux Distros -> GNU/Linux.
Linux is the kernel. Shared kernel means it's Linux.
Period. end of story.
Where would you classify Debian k-FreeBSD, which is Debian running over the FreeBSD kernel? And Debian GNU/Hurd? Are they "less" a linux distro than other flavor of Debian?
Android is a Linux distro.
Hell no. Android is not a distribution. From wikipedia: "The operating system will consist of the Linux kernel and, usually, a set of libraries and utilities from the GNU project, with graphics support from the X Window System. Distributions optimized for size may not contain X and tend to use more compact alternatives to the GNU utilities, such as Busybox, uClibc, or dietlibc. There are currently over six hundred Linux distributions. Over three hundred of those are in active development, constantly being revised and improved."
Where are my GNU tools in Android? It's definitely NOT a Unix distribution. It barely has a fork of the Linux kernel, which in many cases, we don't have the patches for and all the drivers.
Oh really, so do you think you could have taken RHL or Ubuntu and popularize Linux as a mobile OS?
That's what a bunch of Debian developers are trying to do, yes. They are doing so by backporting the open source Meego APPs into Debian so that your phone can eventually run them. It seems to work pretty well in fact (faster than Maemo which is damned slow).
I could write arguments about why it's not, but the simple fact that you wrote it means you don't understand what a Linux distribution is.
Next, your speculations about Maemo vs Android are quite silly and misses the hole point. Nokia didn't care promoting its own product, or sharing some vital drivers that were shipped as binary blobs, didn't have a clear path on the development, and even made very bad announcements right after the n900 was out. How do you want other companies to follow such a dangerous path, or even use the OS of a competitor which aggressively uses patents when pretending to do open source stuff?
And then there are the taxes.... American business has to go overseas for the cheap labor and the lower taxes in order to compete with the rest of the World.
I don't know where you read that taxes are low in China, but it's wrong. In fact, taxes on labor in China are really as heavy as in France for example. The difference is that for lower salaries (bellow 2 or 3000 yuan per month IIRC), people are exempted from taxes. But read it carefully: I wrote "people" not companies. On average, you'd pay as much taxes as there is salaries. Of course, that doesn't represent so much, because of the fact that salaries are low. But that doesn't make it a low tax country anyway.
In the meantime, the entire World spirals down economically...
Last time I checked, Germany, China and Brazil (for example) are doing pretty well. Not the entire world, your world.
My understanding of the situation (although, full disclaimer I do not speak Chinese nor have I ever been to China) is that the companies simply don't follow regulation.
You are talking as if there was on one side private companies, and on another the Chinese gov. That's not the way it works in China.
If China introduces "regulation" that would stunt their free market, the free market simply circumvents it one way or another [slashdot.org].
Ditto. There's no such thing as a "free market" the way you imagine it. You can't just decide to build a power company in China without strong government support.
And when a country trades with China, they're just exporting their pollution.
That, and also using very cheap labor with very harsh work conditions. Just to make sure: do you own a Foxconn-made iPhone?
Let's put it this way too: how long do you think a global warming scientist will hold his grant if he discovers that there wont be, in fact, global warming? Do you really think so called climatologists would be pleased to cut the three branch they are sitting on?
Now, forget about what I just wrote above, and let's just agree that greediness, politics and money is an issue here, which ever side you want to be with.
No, the fact the language doesn't deal with the real programmer's life issues with (eventually) bad text editor, or simply because there will always people with bad or diverging habits is python downside. The fact that one may or may not use a text editor that you consider "good" or that you think tabs or spaces are the bad way is irrelevant. You may "like it", and I can see why (like it's forcing people into doing good indentation), but do not dismiss the issues associated with it.
Yet, in Afghanistan, the authorities REFUSED that OLPC are equipped with WiFi connection, refusing kids a free access to the Internet, base on the false pretext of the dangerousness of its content for kids.
Even the idea of having an electronic voting machine is scary. Even a mechanical one. Of course, everything has to be done manually, and watching how the voting and counting process is being done should be granted by constitutional rights (like in France, for example). Not only this: the method to do the manual counting should be written in the stone, because even manually, there are ways to cheat.
I didn't read TFA (because I don't care about phones), but I don't need to do it to tell that the news title in/. is stupid. Having the stats of 14% of problems being due to hardware doesn't tell the global failure rate. So, let's say we have 0.00001% of failure rate with android phones, and 1% with others, with still 14% of problems being due to hardware on the Android platform, that doesn't make Android more susceptible for hardware issue, does it? So, either the linked article is stupid, either the summary's title is.
And there's as well Courtillot who asks for the data, never had them, and did exactly what you said above: tried to gather the data by himself. Unfortunately, it was a very long and uneasy task, and he couldn't get all the data he wanted, so himself said that his results aren't as good as one could expect, because of the lack of data.
Continue to say/think what you want, this retention of data IS a problem, however you put it.
Do you think that killing half of a population is a scientific possibility that we should talk in a book as well? Is eugenicist theories also acceptable in that kind of book, just because it is a possibility? If not, why does making forced sterilization acceptable, even if you seem to consider it morally bad as well?
I don't care what kind of statements he made AFTER the book was published, because people have been pointing at him. It's too late, he wrote it. Now, even the fact that this is a "possibility" in a "brainstorm" is horrible.
As for the quote, it's right in the book, read it. I don't like at all the fact that one may say "I'm only the co-author". If he signed it, it's his writings, that's it.
Sure, the head of the IPCC is made of politics rather than scientists. So I'll rephrase: he was the head of the scientists, together with NASA and MIT, that built the AR4, which is the main document made by the IPCC in 2007. He clearly lead the report building though. Sorry for the shortcut, but I thought it was so obvious... Does that make all what I said as invalid? Are you contesting his role as a scientist in the AR4? And remember, my point was the hate between VC and PJ... You're just trying to play with words and make it looks bad, it's silly.
First, let me say his name correctly. It's "holdren". Second, I'm not pushing for it, I'm discussed by it, and want to expose Holdren, because people don't know enough his horrible book. 3rd, I'm a self made man, founder and CEO of www.gplhost.com, a Debian Developer, and leading all sorts of open source projects related to hosting. Do you call that a pariah?
I'm saying that the model from the IPCC is wrong, and pointed out to the studies of Courtillot.
You wrote that there was "no consensus on the quantification". When I show you consensus, you admit that there is consensus but that it is only among climatologists who are all accomplices in a global conspiracy to suppress dissent by not funding it.
NO! This isn't what I wrote, you are the one talking about the James Bond like Climatologist consensus doing a conspiracy. If you don't like Courtillot, and think he is involved in a conspiracy, or even that he isn't respectable because his main field isn't Climatology, we have nothing to discuss anymore. I am not interested in such topics, and I'll leave you with your illusions of consensus...
There's all sorts of electric bi-wheelers for sales in China. Take your pick. From the most simple bike with a battery, to the most heavy scooter with double battery and long range. This really should become a best-seller everywhere, because IT IS very efficient, cheap, and clean. You can search counter-arguments if you want, but I wont change my mind on that, I saw how much it is so popular in China, and how convenient it is, as I own one myself. There's absolutely no reason why policy makers wouldn't push for it in the west, yet it's not happening.
If the goal is to stop using petrol, so we can move to alternatives, then let's tax the petrol itself (even more than it is today), not CO2. If we tax it to the point that it becomes cheaper to use electricity, it will happen. A tax on CO2 doesn't set such goals.
That is very silly. So, let's tax pollution to create not-pollution energy. How about deciding to not pollute at all in the first place, so we don't need a tax? The "CO2 neutral" thing is really a big bullshit though. Planting a tree on the other side of earth wont remove the pollution you are generating in your own city.
Slashdot Mirror
i can count the number of rebuilds i've needed to do with windows for my own systems on half of one hand in the past 3 years
I can count the number of rebuilds of Debian I needed to do with my laptop over 10 years using one finger. And that includes upgrades from one version to another. Apparently, I am not the only one. I just read few weeks ago about an issue being fixed for people that upgraded their system from Potato to Wheezy all the way...
Not specifically joe and mtr, but being able to rebuild basic ncurses tools that needs the autotools layer, libc6, and other non-fancy very common libs (zlib, libstdc++, and other stuff like that) is for sure one of the requirements. I'm on purpose avoiding any graphical (X, GTK, Qt, etc.) depends to my definition, but some might not agree. There's about 30k packages in Debian, about 10k in CentOS 5.x, being able to rebuild (or even at least, cross-compile) a vast majority of the non-graphical ones seems a reasonable requirement to be able to be called a distribution. That is for being a *Unix* distro, not even Linux (eg: you aren't bound to a particular kernel here, which I don't think is the important bit). dpkg and rpm aren't even involved, you could well use other types of packages if you like it (see the variety of choice we have in this world: tar.gz with Slackware, deb, rpm, gentoo and FreeBSD ports, and there's all sorts of other less common types of packages (.ipk of openwrt anyone?)). Being bound to a particular language in a device (java, in the case of Android) isn't really helping.
No, windows itself wouldn't, but cygwin for sure IS a GNU distro. Windows is simply capable of running it. Here, you are simply pretending that Linux == kernel and that's it, like many others in this /. entry. If you want to think it this way, then go on. But never ever, the Android platform can be called a "distribution" (Linux or not).
See how much a waste of time it would have been to state arguments that you know already! Yourself, you know already that running Linux is all about the GNU tools and all the libraries we have, and the ease of "porting" one app to another arch/platform. In Maemo, I just typed "dpkg-buildpackage" and I had both joe and mtr available for my phone, and I use them both very often. Until I can do that in Android, you can't call it a Unix distribution (and yet, even less a Linux distro.).
Do you know how selling software works? If you just throw together a .deb or .rpm, and "be done with it", you'll find an inbox full of angry customers demanding support. Oh, you packaged your .deb based on Debian Squeeze? Well, package foo upon which you software was tested doesn't work the same in Ubuntu Natty Narwhal. And never mind all the requests from Gentoo users who want to know which flags they need to set to recompile their software to fix some obscure bug that pops up when trying to run the software as installed by your .rpm.
Has someone every said you should ship ONE type of package? It isn't that hard to build binaries for both Debian and Ubuntu, it's just a dpkg-buildpackage away, and virtualization (like virtualbox) makes it *very* easy to have many build platforms. Now, let's say you want to cover 90% of the cases, then you'd be doing packages for Debian, Ubuntu and CentOS, with both 32 and 64 bits. In fact, that's down to doing one deb and one rpm, and then build. Yet, you might find some funny distro like Gentoo, you are right, and "they" wont like it, but you can still live with it... Also, most of the time, there's no ABI breakage, so it will work even without recompiling (see the horrible work that Skype did: even if it's horrible job, it does work).
Oh, and where does your package install to? Is it /usr/local/? Or maybe /opt/? Wherever you pick, you will piss off someone who wants it somewhere else.
There's a standard filesystem hierarchy here, it's not about pissing somebody off, it's about respecting standards, like having your configuration files in /etc, your read only files in /usr, and so on. There's no reason to put anything in /opt unless you are on the non-brainer Maemo platform, or things in /usr/local unless you are writing a FreeBSD port.
And for systems that aren't based on .rpm or .deb packages, so they don't have proper dependancies? *You* might say they are ignorable, they *they* sure won't see it that way.
But *you* might don't care about 5% of the users if you've covered most of the others with deb/rpm. So let them say...
Most Linux Distros -> GNU/Linux. Linux is the kernel. Shared kernel means it's Linux. Period. end of story.
Where would you classify Debian k-FreeBSD, which is Debian running over the FreeBSD kernel? And Debian GNU/Hurd? Are they "less" a linux distro than other flavor of Debian?
Android is a Linux distro.
Hell no. Android is not a distribution. From wikipedia: "The operating system will consist of the Linux kernel and, usually, a set of libraries and utilities from the GNU project, with graphics support from the X Window System. Distributions optimized for size may not contain X and tend to use more compact alternatives to the GNU utilities, such as Busybox, uClibc, or dietlibc. There are currently over six hundred Linux distributions. Over three hundred of those are in active development, constantly being revised and improved."
Where are my GNU tools in Android? It's definitely NOT a Unix distribution. It barely has a fork of the Linux kernel, which in many cases, we don't have the patches for and all the drivers.
Oh really, so do you think you could have taken RHL or Ubuntu and popularize Linux as a mobile OS?
That's what a bunch of Debian developers are trying to do, yes. They are doing so by backporting the open source Meego APPs into Debian so that your phone can eventually run them. It seems to work pretty well in fact (faster than Maemo which is damned slow).
As it stands, Android is a Linux distribution
No, it is NOT.
I could write arguments about why it's not, but the simple fact that you wrote it means you don't understand what a Linux distribution is.
Next, your speculations about Maemo vs Android are quite silly and misses the hole point. Nokia didn't care promoting its own product, or sharing some vital drivers that were shipped as binary blobs, didn't have a clear path on the development, and even made very bad announcements right after the n900 was out. How do you want other companies to follow such a dangerous path, or even use the OS of a competitor which aggressively uses patents when pretending to do open source stuff?
And then there are the taxes .... American business has to go overseas for the cheap labor and the lower taxes in order to compete with the rest of the World.
I don't know where you read that taxes are low in China, but it's wrong. In fact, taxes on labor in China are really as heavy as in France for example. The difference is that for lower salaries (bellow 2 or 3000 yuan per month IIRC), people are exempted from taxes. But read it carefully: I wrote "people" not companies. On average, you'd pay as much taxes as there is salaries. Of course, that doesn't represent so much, because of the fact that salaries are low. But that doesn't make it a low tax country anyway.
In the meantime, the entire World spirals down economically...
Last time I checked, Germany, China and Brazil (for example) are doing pretty well. Not the entire world, your world.
My understanding of the situation (although, full disclaimer I do not speak Chinese nor have I ever been to China) is that the companies simply don't follow regulation.
You are talking as if there was on one side private companies, and on another the Chinese gov. That's not the way it works in China.
If China introduces "regulation" that would stunt their free market, the free market simply circumvents it one way or another [slashdot.org].
Ditto. There's no such thing as a "free market" the way you imagine it. You can't just decide to build a power company in China without strong government support.
And when a country trades with China, they're just exporting their pollution.
That, and also using very cheap labor with very harsh work conditions. Just to make sure: do you own a Foxconn-made iPhone?
Here http://solidot.org/ ?
Let's put it this way too: how long do you think a global warming scientist will hold his grant if he discovers that there wont be, in fact, global warming? Do you really think so called climatologists would be pleased to cut the three branch they are sitting on?
Now, forget about what I just wrote above, and let's just agree that greediness, politics and money is an issue here, which ever side you want to be with.
You need a get text editor. :p
No, the fact the language doesn't deal with the real programmer's life issues with (eventually) bad text editor, or simply because there will always people with bad or diverging habits is python downside. The fact that one may or may not use a text editor that you consider "good" or that you think tabs or spaces are the bad way is irrelevant. You may "like it", and I can see why (like it's forcing people into doing good indentation), but do not dismiss the issues associated with it.
Yet, in Afghanistan, the authorities REFUSED that OLPC are equipped with WiFi connection, refusing kids a free access to the Internet, base on the false pretext of the dangerousness of its content for kids.
Even the idea of having an electronic voting machine is scary. Even a mechanical one. Of course, everything has to be done manually, and watching how the voting and counting process is being done should be granted by constitutional rights (like in France, for example). Not only this: the method to do the manual counting should be written in the stone, because even manually, there are ways to cheat.
I didn't read TFA (because I don't care about phones), but I don't need to do it to tell that the news title in /. is stupid. Having the stats of 14% of problems being due to hardware doesn't tell the global failure rate. So, let's say we have 0.00001% of failure rate with android phones, and 1% with others, with still 14% of problems being due to hardware on the Android platform, that doesn't make Android more susceptible for hardware issue, does it? So, either the linked article is stupid, either the summary's title is.
And there's as well Courtillot who asks for the data, never had them, and did exactly what you said above: tried to gather the data by himself. Unfortunately, it was a very long and uneasy task, and he couldn't get all the data he wanted, so himself said that his results aren't as good as one could expect, because of the lack of data.
Continue to say/think what you want, this retention of data IS a problem, however you put it.
Do you think that killing half of a population is a scientific possibility that we should talk in a book as well? Is eugenicist theories also acceptable in that kind of book, just because it is a possibility? If not, why does making forced sterilization acceptable, even if you seem to consider it morally bad as well?
I don't care what kind of statements he made AFTER the book was published, because people have been pointing at him. It's too late, he wrote it. Now, even the fact that this is a "possibility" in a "brainstorm" is horrible.
As for the quote, it's right in the book, read it. I don't like at all the fact that one may say "I'm only the co-author". If he signed it, it's his writings, that's it.
Sure, the head of the IPCC is made of politics rather than scientists. So I'll rephrase: he was the head of the scientists, together with NASA and MIT, that built the AR4, which is the main document made by the IPCC in 2007. He clearly lead the report building though. Sorry for the shortcut, but I thought it was so obvious... Does that make all what I said as invalid? Are you contesting his role as a scientist in the AR4? And remember, my point was the hate between VC and PJ... You're just trying to play with words and make it looks bad, it's silly.
First, let me say his name correctly. It's "holdren". Second, I'm not pushing for it, I'm discussed by it, and want to expose Holdren, because people don't know enough his horrible book. 3rd, I'm a self made man, founder and CEO of www.gplhost.com, a Debian Developer, and leading all sorts of open source projects related to hosting. Do you call that a pariah?
You wrote that there was "no consensus on the quantification". When I show you consensus, you admit that there is consensus but that it is only among climatologists who are all accomplices in a global conspiracy to suppress dissent by not funding it.
NO! This isn't what I wrote, you are the one talking about the James Bond like Climatologist consensus doing a conspiracy. If you don't like Courtillot, and think he is involved in a conspiracy, or even that he isn't respectable because his main field isn't Climatology, we have nothing to discuss anymore. I am not interested in such topics, and I'll leave you with your illusions of consensus...
There's all sorts of electric bi-wheelers for sales in China. Take your pick. From the most simple bike with a battery, to the most heavy scooter with double battery and long range. This really should become a best-seller everywhere, because IT IS very efficient, cheap, and clean. You can search counter-arguments if you want, but I wont change my mind on that, I saw how much it is so popular in China, and how convenient it is, as I own one myself. There's absolutely no reason why policy makers wouldn't push for it in the west, yet it's not happening.
If the goal is to stop using petrol, so we can move to alternatives, then let's tax the petrol itself (even more than it is today), not CO2. If we tax it to the point that it becomes cheaper to use electricity, it will happen. A tax on CO2 doesn't set such goals.
That is very silly. So, let's tax pollution to create not-pollution energy. How about deciding to not pollute at all in the first place, so we don't need a tax? The "CO2 neutral" thing is really a big bullshit though. Planting a tree on the other side of earth wont remove the pollution you are generating in your own city.