For the last couple of years, I've been mostly ignoring Slashdot, occasionally reading a headline, wishing back for the good old days when the incompetence was charming. This is not charming. This is hateful.
I have serious doubts that Canonical is able to deliver on this: they do not have a history of delivering top-notch software, unless you count their press-releases and boundless enthusiasm as software.
Aside from a few interesting things (upstart being among the few projects adopted outside of Ubuntu), they've basically decided to ignore whatever the rest of the community is doing and implement their own (buggy) stuff which is "better". Canonical's stuff makes GNOME3 look usable. That takes some doing.
Aside from my doubts about their ability, I also find the concept deeply flawed. Cheap support infrastructure does not currently exist for a dockable phone. Sure, you can use it as a desktop, you just need to buy a dock that you carry around, or a dock for every desk you usually use. Sure, you can use it as a phone, you just need a bluetooth headset that you have to keep charged when you're using it as a desktop. Sure, it's dual-boot, it just means that you can't phone or use the desktop when you switch modes. Sure it can do all of the above, but you have no battery life.
People who need to navigate and use their phone a lot tend to have TWO devices: a GPS or built-in satnav an a phone. Convergence is a great idea, but you're going to pay a lot in battery life for all those features. Running out of juice is NOT FUN these days.
It appears Shuttleworth is trying to emulate companies like Apple, Microsoft and Google by doing the opposite of what used to be done in the spirit of Linux. The copyright clause in all Canonical software, Mir, forking GNOME into Unity and the doublespeak pouring out of the community spokesdrones have been in stark contrast to the early days of Debian, Slackware and open culture. Maybe he really believes he's Steve Jobs and Bill Gates reincarnated and rolled into one: I really think he's got the remorselessness of the one and the ruthlessness of the other.
I believe Ubuntu has single-handedly done more to bring down the quality of Linux on the desktop than any other distro.
I believe the reason Ubuntu is so successful is because of marketing. NOT because of technical quality. This is why I believe that the human race is getting stupider every year. Ah well.
While I don't very much like the *code* that Poettering contributes (pulseAudio took a LONG while to become stable), I see sense in a lot of his arguments:
1) SystemV initscripts are fine for systems that were designed 20 years ago. Things have changed quite a bit now. 2) After seeing some of the Apple launchctl things in action, I want some on Linux! 3) If we stick to POSIX, we might as well decide to throw in the towel, break out the old Slackware 1.0 distros and grow beards. If we can design a better interface/system that's more future proof, then DO IT. 4) Letting Upstart/SysV/OpenRC and whatever compete is *not* a good thing. It's the equivalent of having 3 incomplete kernels that allow you to run your audio, graphics or disk, one at a time. 5) Turns out pulseAudio got better AFTER PEOPLE FIXED IT UP. The architecture and the idea wasn't busted, but the execution was, for a long time.
The only reason Red Hat is upstream is because they contribute so damn much to the code. But, as Mark Shuttleworth said, Canonical contributes users and bugreports (sometimes directly to Red Hat, hilariously). Turns out, you have less control over code than the authors, go figure.
Finally, from my point of view, Unity and GNOME 3 are both abominations that should be killed with fire. I stand 100% behind Linus's statements about compatibility and ABI breakage. The fact that your app can only run on a specific distribution with a specific set of libraries is very rarely a good way to keep guys interested in developing for your desktop.
Strange, last I heard was that Debian added it as an *alternative*. Gentoo's had their own initsystem (they switched to openrc right about when I left 'em), but, to be honest, the average Gentoo user could probably boot his PC just by flicking the power randomly to clock the bits into RAM.
Your viewpoint on how Ubuntu and Canonical contributes back to the community notwithstanding, there seems to be a stark difference between the management style of Red Hat and that of Canonical.
The perception raised with Unity is that Canonical has decided to diverge from upstream more and more: this is evident from the problems that the Debian project (which contributes the majority of code to Ubuntu) is facing, as well as GNOME and the dissent with the upcoming signed boot EFI implementations.
Red Hat (and the Fedora project) is trying to prevent the balkanization of Linux userspace with projects like systemd, which only Ubuntu rejects.
Red Hat's business model seems to be very successful, and Canonical, despite it's massive desktop market share, doesn't seem to be able to match it in reputation or revenue. Would you attribute this to Red Hat's deeper involvement in the kernel community and higher technical skills?
Then it only took 8 years to make a 64-bit port from the date of the first available amd64 machine. If you take into account the date of the first full spec released to the public, it's almost 11 years.
Now if only complex software like the Linux Kernel could be ported in shorter time....
First off, a short comment on legality: Linus views binary drivers as legal ONLY IF they were "ported" to Linux. His view is, if they were written for the kernel from scratch, they're derivative work. Otherwise they don't derive from GPL code and (though undesired) are considered legal. Note that this "Linus" I am referring to may have no actual relation to the "Linus Torvalds" person who wrote Linux, it's just what I think I remember him saying.
Now that's out of the way, the first curveball that hit me after Win32 world was the idea of no stable ABI. First let's get some misconceptions out of the way:
1) Windows doesn't have a stable driver ABI. Ask NVidia. Ask Intel. Every service pack that can potentially break a driver means that the ABI has changed! How many times did you have to update a driver after a service pack? How many times did a driver "require" a new service pack to install? Yeah, I thought so. 2) Distributions have "stablish" driver ABIs. SLES and RHEL have got reference kernels with stable ABI calls that's well documented AND YOU HAVE THE SOURCE. It's much simpler to update drivers for Linux than for a new Windows service pack since the Enterprise versions normally stick to the ultra-stable stuff.
But, noooooo! I hear. Screw this stable stuff, stable stuff is old! And that's the crux of it. On a desktop, you don't necessarily want 99.9999% availability, you'd rather have 20 more features, but a non-serious hang once a month. The difference between the Windows model and the Linux model is: YOU NEVER GET WINDOWS KERNEL FEATURE UPDATES! You have to pay for them! And wait years and years... Yah, the service packs add features, but to most people they are considered bugfixes;-)
So that's why the kernel shouldn't have a stable API (let alone ABI) for drivers. It prevents big architectural decisions from being taken at the opportune moment. With the kernel under git management, it's even easier to maintain out-of-tree drivers (case in point: alsa). What developers who clamour for a stable ABI don't understand is that there _is_ one! Any fork of Linux they make can be as stable as they want it to be. If they want to benefit from continual improvement, they _have_ to make some sacrifices.
Second-to-lastly, the guys biting of the short stick here are the distro maintainers. Again, there are many options: follow the source route like Gentoo, allow only free hardware like Debian, rely on the user base to spin driver packages like Fedora, Ubuntu and OpenSuSE, or go the stable route like SLES and RHEL. If they really want a stable video ABI so NVIDIA's drivers are even easier to install, get THEM to work together on it. If it's good, it'll get in to the kernel. If not, it won't.
Lastly, try the following experiment. Buy a shiny new Windows XP Home (or Pro, for that matter) CD, and install it on your brand-spanking-new hardware. Now, try to write a USB driver using only software that you don't need to pay for, and for which you'd own the binaries after you'd written it. Finding it difficult? If you got it right, PLEASE TELL ME FREAKING HOW, SINCE I DON'T KNOW! Next step, PCI driver... but hey, Rome wasn't built in a day!
Half-life was released in November of 1998, and it did not run well on the typical gamer system then. I had a P2 400 and 128MB RAM with a TNT2 graphics card and got about 25 FPS in 640x480 after much tweaking.
I'm sorry, I had a P2 266 with a Permedia 2 chipset graphics card (Creative Labs Graphics Blaster Exxtreme if I recall correctly). 640x480 netted me 25 fps and when I upgraded to a TNT (NOT TNT2) I had 1024x768 at a very decent framerate (probably more than 30, it never jerked). BTW, both were at full detail.
Maybe you were thinking of a P2 200 and a Riva128?
My personal view is that there is only two stumbling blocks for Linux, neither of them technical or really solvable without (it seems) large amounts of money for bribes^H^H^H^H^H^Hinvestment. The first is hardware drivers (which might be solvable by throwing money at developers) and the second is patents and associated legal hassles. Would you suggest that we take the battle to the politicians/policymakers eventually? Or do you think that it wouldn't be a problem in the long run?
This might be a bit of a sensitive issue, but do you plan to make money off this project, and other open source projects eventually, or is the funding a (VERY GENEROUS!) gift to the community? I assume that profitability would be a long long long term (10+ years in the computer industry!) goal of any project, but I get a sense "profitability" is not monetary only in this case.
There's really only one inconsistency that bothered me in Cryptonomicon, and that is the character of Enoch Root. Do we have enough evidence by now (I haven't yet read books 2 and 3 of the Baroque Cycle) to solve this conundrum? Would it be solvable?
Finally, was there any particular reasoning behind using "Finux" and not "Linux" as a mythical OS in Cryptonomicon? Or is this an alternate reality where Linus Torvalds went to work for Microsoft?
Wrong, wrong, wrong. When last have you installed a Linux distro? USB/Firewire hotplugging works better under Linux than under Windows for me. Try any modern distribution with a kernel later than 2.4.22 and you'll have support for ACPI. In fact, kernel 2.6.0-test8-mm1 on Gentoo supports all devices (including the Zoltrix Genie-Wonder-Pro that Windows XP doesn't support) on my system without a single glitch. USB 2.0 works fine. Firewire runs perfectly and my motherboard's sensors get reported via gkrellm2. If you really want some shenanigans, try to load Windows 95 on some newer motherboards. Guaranteed to make you wince. Can you believe it that some motherboards aren't backwards compatible? At least Linux can now run on old stuff and new!
I got VERY nice results with Gentoo on my NForce2 board. But, then, I looove tweaking.
With the newest 2.4.22-pre4, you have to use acpi=off pci=noacpi on the boot command line. What's nice, is that specific kernel supports my IDE, AGP and using ALSA 0.9.5, my sound.
It's not nice that the NVidia people have the Linux support a couple of months behind the Windows, but it seems that that's the best you're going to get!
Well... I'd rather run an OS where I KNOW what's going on than one that tells me that I don't need to know what's going on.
The biggest problem I have with the newer kernels is probably some ACPI/IRQ routing bug in my board. It's a common problem with the NForce2 chipset (APIC doesn't work, so you have to boot with pci=noacpi or acpi=off). It's not the biggest inconvenience, but it causes half of my unused USB slots not to work...
I must say the snappiness of 2.6 is great! I'm looking forward to beta-testing. AFTER I backed up my drive, of course!
Every single person I know who has played video on Linux (except DVDs) with MPlayer fell in love with it. Really, there are only six shortcut keys you need:
right, up : ffd ffffd left, down : rev rrev f : fullscreen space : pause
That's 99.9% of what I do when I when I play movies and MPlayer does it REALLY well. No smegging around with codecs, plays broken.avi's as easily as non-broken ones AND now features Sorenson SVQ4 playback! Hint: keyboard is faster than mouse!
I'd really like to see this guy giving constructive criticisms. No, don't ask him to criticize my post, NO NO NO!
Powered, no doubt, by a slice of buttered toast strapped to the back of a cat!
I think somebody once funded a study into this and numerous conference papers later (along with the help of a couple of great Slashdot posts) it was put off as impractical.
Apparently, the amount of energy required to strap the buttered toast to the back of the cat negates any net gain from the system over time.
Also, physical experiments are inconclusive, since the lacerations take too long to heal.
As a general rule only dot releases keep C++ ABI compatibility. C compatibility is now generally regarded as stable. One of the main goals for GCC 3.0 was a stable C++ ABI, but it seems we'll have to wait for at least GCC 3.3 for that;-)
So, to sum up, 2.95.*, "2.96", 3.0.*, 3.1.* and 3.2.* will have issues if you want to link one's code to the other using C++. Also, 2.96 can't really peacefully co-exist with 2.95 because of library name conflicts...
But hey, source distributions don't have that problem! Well-written source code generally compiles on any one of the three... besides you can always go back to assembly...
Short answer: the gcc crew is lazy, inconsiderate, or both.
Look, in all fairness you have to extend that to the ANSI/ISO team too. We have waited YEARS to get the C++ standard out and EVERY SINGLE DRAFT changed the language immensely. (Can you still remember the for [int i](i=0;i10;i++) debacle?)
Not only that, but most of the standard breaks existing code. Most of the existing code. In fact, I would say, 90% of existing code. (The idiotic.h header extension change forces code to emit warnings or even errors).
In one sense, this is good. It promotes non-platform dependend coding. In another sense, this is bad. It promotes compilers that are widely disparate and temperamental.
So, to sum up: the GCC team is doing a fine job. Implementing ISO with a set of moving goalposts is done very well. After everyone got uptight because GCC 3.0 was delayed (because of this Red Hat probably forked off the "bastard son of gcc", 2.96) the GCC team decided to make the release schedule faster.
You ARE GOING TO GET BUGS IF YOU RELEASE FASTER! So there, you have two reasons to whine, one if GCC is bug-free and one if it is not! Enjoy!
The date is July 11th 1992. The GNU people are once again safe from patent law... gzip has saved the LZW-hampered compress from kneeing the GNU project in the other groin.
RMS is still not relaxed however. He knows the Achilles heel of his plan to free the hearts and minds of all thinking beings in the world is patents. Restlessly, he searches for an answer, but as yet, it has eluded him. Later that day, one of his minions shows up and senses something is on his master's mind.
Minion0: What is wrong, master? Has someone threatened freedom again? RMS: Rest easy young minion. Today, at least the battle has been won, but I fear the war may still be lost. Minion0: You can't mean... RMS: No, no, my facade as a badly-dressed over-zealous hippie will not be breeched. Be assured that it is inconcievable for anyone to see past my BO. Minion0: Phew. But yet, I sense something's amiss. RMS: Patents may yet be our downfall. The dark forces has threatened us before with them, and soon they will again. Minion0: But, cannot we impress upon the minds of our leaders the importance of this? RMS: Alas, in that respect our actions will be in vain. Our only chance lies with being more wise than them. Minion0: I was worried there for a moment... here, let me get the champagne! RMS: Perhaps it is good that we rejoice in our small victory. But the lull will not last and yet... Minion0: And yet? RMS: The time is not yet ripe. Rest assured, this patent war, we shall win!
And so, after a brief victory celebration, the GNU people return to their secret identities of coders who are brilliant, yet in need of basic hygiene.
Yet RMS continues to plan, in the dark and twisting corners of his mind....
We rejoin RMS on April 1 2001, in the boardroom of a relatively unknown company named VTEL. The board members of this struggling company wonder what the darkly dressed stranger has to say.
RMS: My plan is nearly complete. One this is done, you will not only have saved freedom from certain doom, but have struck a terrible blow at the very heart of the dark forces opposing us. Boardmember0: But, would this not cause us to become unpopular and even shunned by our families? Boardmember1: And what if there are complications? RMS: All of the above has been taken care off... My... minion... here, will pose as a new board member. After the deed is done, he will fake his own death, and in the resulting chaos, all blame will be passed on to him. Minion0: Naturally, you will never see me or RMS again after this meeting. RMS: And so, we have prepared new identities in any country of your choice for you and your loved ones. We chose your company because it would make the least impact on your social lives. Boardmember0: But this name change? Would it not raise questions to the observer? RMS: Only to those who are free from the dark influences. Fortunately for us, our counterparts are oft blinded by the smell of money...
And so, we reach our present day. The dominoes of fate are all in place, and we can only hope that freedom will prevail!
* Java support is broken. Reason: The currently available Java is not compiled with GCC 3.1 and therefore does not work with our packages.
What is so imperative about going to gcc 3.1 that you have to break java?
Not to harp too much, but I've got a shiny new Gentoo system compiled from scratch with GCC 3.1
1) GCC 3.1 makes bigger code 2) GCC 3.1 makes faster code (most of the times) 3) GCC 3.1 actually tries to conform to a standard other than "just GCC" 4) Binary compatability is ONLY broken for C++ (maybe some other languages, but definitely not C) 5) Java (Sun's JDK) works if you compile it from scratch. 6) Plugins with Mozilla is a bit tricky. 7) In order to maintain Red Hat compatability they HAVE to use GCC 3.1 8) The GCC team will shortly break binary compatability yet again (by renaming the GCC 3.1 branch to GCC 3.2). 9) My is fine. Maybe not stable, but much nicer for a desktop. 10) Once more people jump on the GCC 3-series bandwagon again (They jumped off when Red Hat did the gcc 2.96 doodoo) GCC will have a nice, stable ABI that won't be broken anytime soon.
...but where's the lowlights? It couldn't have been roses all the time! What about the unstable bits in the 2.4 kernels and the VM debacle?
Frankly, if they decide to show all the Microsoft lowlights, they might as well show all the Linux lowlights too!
Make no mistake. If we don't learn from our mistakes in the past, we will not continue. A timeline that lists all the obstacles Linux had to overcome (legal, architectural and social) would be much nicer.
Most machine intelligence techniques I have come across (like neural nets, genetic algorithms and expert systems) require some for of training. A "reward algorithm", if you will, that reinforces certain behaviour mechanisms so that the system "trains" to do something you want.
I would assume that humans derive these training inputs much the same way, since pain receptors and pleasure sensations influence our behaviour much more than we would think at first.
The question is: For a "true" AI that mimics real intelligence as close as possible, what do you think would be used as training influences? Perhaps a neural net (or statistical analysis) could decide on which input should be used to train the system?
Are people worrying about moral ramifications, training an artificial Hitler, for example, or one with a God complex? (This last question is totally philosophical and I would be sincerely surprised if I ever see it affect me during my lifetime.)
Slashdot Mirror
Goodbye Slashdot.
For the last couple of years, I've been mostly ignoring Slashdot, occasionally reading a headline, wishing back for the good old days when the incompetence was charming. This is not charming. This is hateful.
RANT COMMENCING!
I have serious doubts that Canonical is able to deliver on this: they do not have a history of delivering top-notch software, unless you count their press-releases and boundless enthusiasm as software.
Aside from a few interesting things (upstart being among the few projects adopted outside of Ubuntu), they've basically decided to ignore whatever the rest of the community is doing and implement their own (buggy) stuff which is "better". Canonical's stuff makes GNOME3 look usable. That takes some doing.
Aside from my doubts about their ability, I also find the concept deeply flawed. Cheap support infrastructure does not currently exist for a dockable phone. Sure, you can use it as a desktop, you just need to buy a dock that you carry around, or a dock for every desk you usually use. Sure, you can use it as a phone, you just need a bluetooth headset that you have to keep charged when you're using it as a desktop. Sure, it's dual-boot, it just means that you can't phone or use the desktop when you switch modes. Sure it can do all of the above, but you have no battery life.
People who need to navigate and use their phone a lot tend to have TWO devices: a GPS or built-in satnav an a phone. Convergence is a great idea, but you're going to pay a lot in battery life for all those features. Running out of juice is NOT FUN these days.
It appears Shuttleworth is trying to emulate companies like Apple, Microsoft and Google by doing the opposite of what used to be done in the spirit of Linux. The copyright clause in all Canonical software, Mir, forking GNOME into Unity and the doublespeak pouring out of the community spokesdrones have been in stark contrast to the early days of Debian, Slackware and open culture. Maybe he really believes he's Steve Jobs and Bill Gates reincarnated and rolled into one: I really think he's got the remorselessness of the one and the ruthlessness of the other.
I believe Ubuntu has single-handedly done more to bring down the quality of Linux on the desktop than any other distro.
I believe the reason Ubuntu is so successful is because of marketing. NOT because of technical quality. This is why I believe that the human race is getting stupider every year. Ah well.
RANT CONCLUDED!
While I don't very much like the *code* that Poettering contributes (pulseAudio took a LONG while to become stable), I see sense in a lot of his arguments:
1) SystemV initscripts are fine for systems that were designed 20 years ago. Things have changed quite a bit now.
2) After seeing some of the Apple launchctl things in action, I want some on Linux!
3) If we stick to POSIX, we might as well decide to throw in the towel, break out the old Slackware 1.0 distros and grow beards. If we can design a better interface/system that's more future proof, then DO IT.
4) Letting Upstart/SysV/OpenRC and whatever compete is *not* a good thing. It's the equivalent of having 3 incomplete kernels that allow you to run your audio, graphics or disk, one at a time.
5) Turns out pulseAudio got better AFTER PEOPLE FIXED IT UP. The architecture and the idea wasn't busted, but the execution was, for a long time.
The only reason Red Hat is upstream is because they contribute so damn much to the code. But, as Mark Shuttleworth said, Canonical contributes users and bugreports (sometimes directly to Red Hat, hilariously). Turns out, you have less control over code than the authors, go figure.
Finally, from my point of view, Unity and GNOME 3 are both abominations that should be killed with fire. I stand 100% behind Linus's statements about compatibility and ABI breakage. The fact that your app can only run on a specific distribution with a specific set of libraries is very rarely a good way to keep guys interested in developing for your desktop.
http://wiki.debian.org/systemd
http://en.gentoo-wiki.com/wiki/Systemd
Strange, last I heard was that Debian added it as an *alternative*. Gentoo's had their own initsystem (they switched to openrc right about when I left 'em), but, to be honest, the average Gentoo user could probably boot his PC just by flicking the power randomly to clock the bits into RAM.
Your viewpoint on how Ubuntu and Canonical contributes back to the community notwithstanding, there seems to be a stark difference between the management style of Red Hat and that of Canonical.
The perception raised with Unity is that Canonical has decided to diverge from upstream more and more: this is evident from the problems that the Debian project (which contributes the majority of code to Ubuntu) is facing, as well as GNOME and the dissent with the upcoming signed boot EFI implementations.
Red Hat (and the Fedora project) is trying to prevent the balkanization of Linux userspace with projects like systemd, which only Ubuntu rejects.
Red Hat's business model seems to be very successful, and Canonical, despite it's massive desktop market share, doesn't seem to be able to match it in reputation or revenue. Would you attribute this to Red Hat's deeper involvement in the kernel community and higher technical skills?
If you look at the timeline of the amd64 architecture: http://en.wikipedia.org/wiki/X86-64#History_of_AMD64
Then it only took 8 years to make a 64-bit port from the date of the first available amd64 machine. If you take into account the date of the first full spec released to the public, it's almost 11 years.
Now if only complex software like the Linux Kernel could be ported in shorter time....
First off, a short comment on legality: Linus views binary drivers as legal ONLY IF they were "ported" to Linux. His view is, if they were written for the kernel from scratch, they're derivative work. Otherwise they don't derive from GPL code and (though undesired) are considered legal. Note that this "Linus" I am referring to may have no actual relation to the "Linus Torvalds" person who wrote Linux, it's just what I think I remember him saying.
;-)
Now that's out of the way, the first curveball that hit me after Win32 world was the idea of no stable ABI. First let's get some misconceptions out of the way:
1) Windows doesn't have a stable driver ABI.
Ask NVidia. Ask Intel. Every service pack that can potentially break a driver means that the ABI has changed! How many times did you have to update a driver after a service pack? How many times did a driver "require" a new service pack to install? Yeah, I thought so.
2) Distributions have "stablish" driver ABIs.
SLES and RHEL have got reference kernels with stable ABI calls that's well documented AND YOU HAVE THE SOURCE. It's much simpler to update drivers for Linux than for a new Windows service pack since the Enterprise versions normally stick to the ultra-stable stuff.
But, noooooo! I hear. Screw this stable stuff, stable stuff is old! And that's the crux of it. On a desktop, you don't necessarily want 99.9999% availability, you'd rather have 20 more features, but a non-serious hang once a month. The difference between the Windows model and the Linux model is: YOU NEVER GET WINDOWS KERNEL FEATURE UPDATES! You have to pay for them! And wait years and years... Yah, the service packs add features, but to most people they are considered bugfixes
So that's why the kernel shouldn't have a stable API (let alone ABI) for drivers. It prevents big architectural decisions from being taken at the opportune moment. With the kernel under git management, it's even easier to maintain out-of-tree drivers (case in point: alsa). What developers who clamour for a stable ABI don't understand is that there _is_ one! Any fork of Linux they make can be as stable as they want it to be. If they want to benefit from continual improvement, they _have_ to make some sacrifices.
Second-to-lastly, the guys biting of the short stick here are the distro maintainers. Again, there are many options: follow the source route like Gentoo, allow only free hardware like Debian, rely on the user base to spin driver packages like Fedora, Ubuntu and OpenSuSE, or go the stable route like SLES and RHEL. If they really want a stable video ABI so NVIDIA's drivers are even easier to install, get THEM to work together on it. If it's good, it'll get in to the kernel. If not, it won't.
Lastly, try the following experiment. Buy a shiny new Windows XP Home (or Pro, for that matter) CD, and install it on your brand-spanking-new hardware. Now, try to write a USB driver using only software that you don't need to pay for, and for which you'd own the binaries after you'd written it. Finding it difficult? If you got it right, PLEASE TELL ME FREAKING HOW, SINCE I DON'T KNOW! Next step, PCI driver... but hey, Rome wasn't built in a day!
I'm sorry, I had a P2 266 with a Permedia 2 chipset graphics card (Creative Labs Graphics Blaster Exxtreme if I recall correctly). 640x480 netted me 25 fps and when I upgraded to a TNT (NOT TNT2) I had 1024x768 at a very decent framerate (probably more than 30, it never jerked). BTW, both were at full detail.
Maybe you were thinking of a P2 200 and a Riva128?
My personal view is that there is only two stumbling blocks for Linux, neither of them technical or really solvable without (it seems) large amounts of money for bribes^H^H^H^H^H^Hinvestment. The first is hardware drivers (which might be solvable by throwing money at developers) and the second is patents and associated legal hassles. Would you suggest that we take the battle to the politicians/policymakers eventually? Or do you think that it wouldn't be a problem in the long run?
This might be a bit of a sensitive issue, but do you plan to make money off this project, and other open source projects eventually, or is the funding a (VERY GENEROUS!) gift to the community? I assume that profitability would be a long long long term (10+ years in the computer industry!) goal of any project, but I get a sense "profitability" is not monetary only in this case.
There's really only one inconsistency that bothered me in Cryptonomicon, and that is the character of Enoch Root. Do we have enough evidence by now (I haven't yet read books 2 and 3 of the Baroque Cycle) to solve this conundrum? Would it be solvable?
Finally, was there any particular reasoning behind using "Finux" and not "Linux" as a mythical OS in Cryptonomicon? Or is this an alternate reality where Linus Torvalds went to work for Microsoft?
Thanks for a great and entertaining read!
Wrong, wrong, wrong. When last have you installed a Linux distro? USB/Firewire hotplugging works better under Linux than under Windows for me. Try any modern distribution with a kernel later than 2.4.22 and you'll have support for ACPI. In fact, kernel 2.6.0-test8-mm1 on Gentoo supports all devices (including the Zoltrix Genie-Wonder-Pro that Windows XP doesn't support) on my system without a single glitch. USB 2.0 works fine. Firewire runs perfectly and my motherboard's sensors get reported via gkrellm2. If you really want some shenanigans, try to load Windows 95 on some newer motherboards. Guaranteed to make you wince. Can you believe it that some motherboards aren't backwards compatible? At least Linux can now run on old stuff and new!
I got VERY nice results with Gentoo on my NForce2 board. But, then, I looove tweaking.
With the newest 2.4.22-pre4, you have to use acpi=off pci=noacpi on the boot command line. What's nice, is that specific kernel supports my IDE, AGP and using ALSA 0.9.5, my sound.
It's not nice that the NVidia people have the Linux support a couple of months behind the Windows, but it seems that that's the best you're going to get!
Well... I'd rather run an OS where I KNOW what's going on than one that tells me that I don't need to know what's going on.
Last time I looked at ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/ must-fix/ there were still some showstoppers. It seems like they were updated about a month ago, so I guess progress must have been made on them...
The biggest problem I have with the newer kernels is probably some ACPI/IRQ routing bug in my board. It's a common problem with the NForce2 chipset (APIC doesn't work, so you have to boot with pci=noacpi or acpi=off). It's not the biggest inconvenience, but it causes half of my unused USB slots not to work...
I must say the snappiness of 2.6 is great! I'm looking forward to beta-testing. AFTER I backed up my drive, of course!
Every single person I know who has played video on Linux (except DVDs) with MPlayer fell in love with it. Really, there are only six shortcut keys you need:
.avi's as easily as non-broken ones AND now features Sorenson SVQ4 playback! Hint: keyboard is faster than mouse!
right, up : ffd ffffd
left, down : rev rrev
f : fullscreen
space : pause
That's 99.9% of what I do when I when I play movies and MPlayer does it REALLY well. No smegging around with codecs, plays broken
I'd really like to see this guy giving constructive criticisms. No, don't ask him to criticize my post, NO NO NO!
from the insert-group-sex-joke-here dept.
Y?
Powered, no doubt, by a slice of buttered toast strapped to the back of a cat!
I think somebody once funded a study into this and numerous conference papers later (along with the help of a couple of great Slashdot posts) it was put off as impractical.
Apparently, the amount of energy required to strap the buttered toast to the back of the cat negates any net gain from the system over time.
Also, physical experiments are inconclusive, since the lacerations take too long to heal.
As a general rule only dot releases keep C++ ABI compatibility. C compatibility is now generally regarded as stable. One of the main goals for GCC 3.0 was a stable C++ ABI, but it seems we'll have to wait for at least GCC 3.3 for that ;-)
So, to sum up, 2.95.*, "2.96", 3.0.*, 3.1.* and 3.2.* will have issues if you want to link one's code to the other using C++. Also, 2.96 can't really peacefully co-exist with 2.95 because of library name conflicts...
But hey, source distributions don't have that problem! Well-written source code
generally compiles on any one of the three... besides you can always go back to assembly...
Short answer: the gcc crew is lazy, inconsiderate, or both.
.h header extension change forces code to emit warnings or even errors).
Look, in all fairness you have to extend that to the ANSI/ISO team too. We have waited YEARS to get the C++ standard out and EVERY SINGLE DRAFT changed the language immensely. (Can you still remember the for [int i](i=0;i10;i++) debacle?)
Not only that, but most of the standard breaks existing code. Most of the existing code. In fact, I would say, 90% of existing code. (The idiotic
In one sense, this is good. It promotes non-platform dependend coding. In another sense, this is bad. It promotes compilers that are widely disparate and temperamental.
So, to sum up: the GCC team is doing a fine job. Implementing ISO with a set of moving goalposts is done very well. After everyone got uptight because GCC 3.0 was delayed (because of this Red Hat probably forked off the "bastard son of gcc", 2.96) the GCC team decided to make the release schedule faster.
You ARE GOING TO GET BUGS IF YOU RELEASE FASTER! So there, you have two reasons to whine, one if GCC is bug-free and one if it is not! Enjoy!
The date is July 11th 1992. The GNU people are once again safe from patent law... gzip has saved the LZW-hampered compress from kneeing the GNU project in the other groin.
...
... minion ... here, will pose as a new board member. After the deed is done, he will fake his own death, and in the resulting chaos, all blame will be passed on to him.
RMS is still not relaxed however. He knows the Achilles heel of his plan to free the hearts and minds of all thinking beings in the world is patents. Restlessly, he searches for an answer, but as yet, it has eluded him. Later that day, one of his minions shows up and senses something is on his master's mind.
Minion0: What is wrong, master? Has someone threatened freedom again?
RMS: Rest easy young minion. Today, at least the battle has been won, but I fear the war may still be lost.
Minion0: You can't mean...
RMS: No, no, my facade as a badly-dressed over-zealous hippie will not be breeched. Be assured that it is inconcievable for anyone to see past my BO.
Minion0: Phew. But yet, I sense something's amiss.
RMS: Patents may yet be our downfall. The dark forces has threatened us before with them, and soon they will again.
Minion0: But, cannot we impress upon the minds of our leaders the importance of this?
RMS: Alas, in that respect our actions will be in vain. Our only chance lies with being more wise than them.
Minion0: I was worried there for a moment... here, let me get the champagne!
RMS: Perhaps it is good that we rejoice in our small victory. But the lull will not last and yet...
Minion0: And yet?
RMS: The time is not yet ripe. Rest assured, this patent war, we shall win!
And so, after a brief victory celebration, the GNU people return to their secret identities of coders who are brilliant, yet in need of basic hygiene.
Yet RMS continues to plan, in the dark and twisting corners of his mind.
We rejoin RMS on April 1 2001, in the boardroom of a relatively unknown company named VTEL. The board members of this struggling company wonder what the darkly dressed stranger has to say.
RMS: My plan is nearly complete. One this is done, you will not only have saved freedom from certain doom, but have struck a terrible blow at the very heart of the dark forces opposing us.
Boardmember0: But, would this not cause us to become unpopular and even shunned by our families?
Boardmember1: And what if there are complications?
RMS: All of the above has been taken care off... My
Minion0: Naturally, you will never see me or RMS again after this meeting.
RMS: And so, we have prepared new identities in any country of your choice for you and your loved ones. We chose your company because it would make the least impact on your social lives.
Boardmember0: But this name change? Would it not raise questions to the observer?
RMS: Only to those who are free from the dark influences. Fortunately for us, our counterparts are oft blinded by the smell of money...
And so, we reach our present day. The dominoes of fate are all in place, and we can only hope that freedom will prevail!
* Java support is broken. Reason: The currently available Java is not compiled with GCC 3.1 and therefore does not work with our packages.
What is so imperative about going to gcc 3.1 that you have to break java?
Not to harp too much, but I've got a shiny new Gentoo system compiled from scratch with GCC 3.1
1) GCC 3.1 makes bigger code
2) GCC 3.1 makes faster code (most of the times)
3) GCC 3.1 actually tries to conform to a standard other than "just GCC"
4) Binary compatability is ONLY broken for C++ (maybe some other languages, but definitely not C)
5) Java (Sun's JDK) works if you compile it from scratch.
6) Plugins with Mozilla is a bit tricky.
7) In order to maintain Red Hat compatability they HAVE to use GCC 3.1
8) The GCC team will shortly break binary compatability yet again (by renaming the GCC 3.1 branch to GCC 3.2).
9) My is fine. Maybe not stable, but much nicer for a desktop.
10) Once more people jump on the GCC 3-series bandwagon again (They jumped off when Red Hat did the gcc 2.96 doodoo) GCC will have a nice, stable ABI that won't be broken anytime soon.
Damn. Harped too much... oh well...
...but where's the lowlights? It couldn't have been roses all the time! What about the unstable bits in the 2.4 kernels and the VM debacle?
Frankly, if they decide to show all the Microsoft lowlights, they might as well show all the Linux lowlights too!
Make no mistake. If we don't learn from our mistakes in the past, we will not continue. A timeline that lists all the obstacles Linux had to overcome (legal, architectural and social) would be much nicer.
Ah well... I'm biased. I dig Linux.
Most machine intelligence techniques I have come across (like neural nets, genetic algorithms and expert systems) require some for of training. A "reward algorithm", if you will, that reinforces certain behaviour mechanisms so that the system "trains" to do something you want.
I would assume that humans derive these training inputs much the same way, since pain receptors and pleasure sensations influence our behaviour much more than we would think at first.
The question is: For a "true" AI that mimics real intelligence as close as possible, what do you think would be used as training influences? Perhaps a neural net (or statistical analysis) could decide on which input should be used to train the system?
Are people worrying about moral ramifications, training an artificial Hitler, for example, or one with a God complex? (This last question is totally philosophical and I would be sincerely surprised if I ever see it affect me during my lifetime.)
v4l2
This would allow the latest bttv driver to be packaged with the kernel once more...
After the preemptive patch, that's the first patch I run on a plain vanilla kernel.
For the sane people out there... here's some unit conversions:
25 Miles = 40.2336 Kilometers
1.68 Mach [dry air, 273 Kelvin] = 2005.6498391999999 kilometers/hour
For the insane people out there... here's some unit conversions:
1 Kilometer = 0.621371 Miles
12 Inches = 1 Foot
3 Feet = 1 Yard
1 Mile = 1,760 Yards
1 Mile = 5,280 Feet
1 Miles = 63,360 Inches
Just to give you a taste of some saner things:
1000 millimeter = 1 meter
100 centimeter = 1 meter
1000 meter = 1 kilometer
But wait! There's more!
1000 milliliter = 1 liter
100 centiliter = 1 liter
1000 liter = 1 kiloliter
Just go to onlineconversions and have fun!