MS can provide a single buggy driver update and create chaos in your customerbase. All they need to say "oops" after it. Customers will blame YOU, not them.
Why would someone blame ASUS when a driver update pushed out by microsoft breaks a previously working system?
You mean like stackoverflow.com? Dear slashdot, we know your traffic figures are declining and you're dying for cheap and effective (ad revenue) content. But really, this is becoming a tragedy...
Which is, from today's point of view, a total waste of ressources on all sides (read: a paper tiger).
At least 50% of all computer users are completely clueless and will use whatever came with their PeeCee. The blue e is "the internet" for them. For once and for all. Joe Sixpack doesn't know and doesn't care that he can remove his blue e or "upgrade" it to a red fox. Why should he? All the sites he goes to work more or or less on the blue e and even if they work less - he's unlikely to notice notice that the boxes don't align or that the rounded corners are off.
Microsoft will keep the lions share of the browser market unless one the following happens:
a) Windows loses significant market share (Linux on the desktop anyone?) b) Windows ships without a webbrowser pre-installed. Instead there is a small widget after installation that asks the user to choose a browser to install ("Ooh, look the cute fox"). c) *Major* sites (CNN, google, porn,...) start blocking IE and tell Joe Sixpack to "upgrade to the red fox".
Google "Human genome project" or pretty much any source about basic genetics. What do you think causes one of your siblings to develop a different skin color? Right, her genes. And we get better at decoding these every day.
There's no "technically" anything when it comes to race. Race is a fiction with no scientific basis. You can't take a cell sample from somebody and pop it in an analyzer and come back and say "this person is 93% black and 7% white".
To paraphrase obama: Yes, you can. The different skin-colors stem from genetic differences and those can be detected. I'm not saying that distinguishing between skin-colors or "races" makes any sense - but technically it's possible to determine someone's skin-color from a DNA-sample, with increasing accuracy.
Network transparency is IMO the killer feature of X, it would be a shame to throw the baby out with the bathwater.
You got it right, almost. Network transparency is the feature that kills X. Network transparency was nice and revolutionary (cough) when X was invented many years ago. Nowadays it's just a huge pain in the ass because it kills performance and makes the whole thing insanely complex.
Reality has taught us that the whole network transparency thing is better tackled independently from the underlying platform. Free software like NoMachine and VNC exists today and handles the Remote Desktop Use-Case much better (faster!) than the native X protocol. Moreover they are cross-platform, they don't even care whether your client/server are a unix, Windows or Mac.
A new X server should most definately not even attempt to provide network transparency. It should focus on solid driver support, fast compositing and a sane programming API so people can actually build stuff on top of it. Anyone who has ever worked with Xlib knows that the ridiculous API is the #1 reason why "linux on the desktop" is just not happening.
The primary reason for QT and GTK being the monstrosities they are is that they both have to deal with an effectively broken foundation.
Pidgin is a very poor jabber client. It doesn't support transports at all and lacks pretty much all advanced jabber features. A better jabber client for linux would be psi (there may be others worth looking at, i don't know) and I'm sure there's something better than pidgin for windows, too.
Maybe google for "jive jabber". They make an opensource jabber server and I think they have a cross platform jabber client, too.
I've never (ten years or so) had a local hardware issue extend into the host network. It seems to be fairly hard to do that if you're not an idiot (and if your own equipment is truly solid, which mine is).
But it's not hard to do at all *when* you're an idiot. I can't count the number of times someone plugged both ends of a cat5-cable into the switch...
Replying to my own post, to sum up what I learned: Nothing...
Maybe I'm an anti-social fag but I *still* don't get it. And the feeling gets stronger that I never will.
Yes I do have friends in the real world but for some reason I really don't care what they are doing at every given point in time. If they do something interesting then I'm sure they'll tell me anyways, probably in more than 128 characters. Moreover I'm already using two messenger programs (Skype and pidgin for all the rest), so I'm already getting more "thinking bubbles", away messages, birthday reminders etc. than I would ever care to read.
From all your arguments only the "persistence" and "zero barrier of entry (no clue required)" stick with me. I guess these two make the big deal out of twitter. People just love to leave their piss-marks everywhere - and stupid people even more so.
How this got *that* big is still beyond me, though. I guess we, as a species, still have a long way to go...
Maybe someone can finally explain that twitter "phenomena" to me? I just don't get it. They basically reinvented IRC and instant messaging poorly and put it on the web. Hm, okay, but why do the unwashed masses flock to it like that?
Back on topic: So the guy that couldn't even get the trivial use-case of a large scale pub/sub right is now complaining about the internet architecture? Too much cocaine?
I don't see why you think that ec2 should be discredited as a whole based on a use-case that they're not even remotely targeting? Ec2 is for application hosting or number crunching. Static content (like porn) is the job for a CDN. Ec2 is not a CDN.
And honestly, claiming you are bandwidth bound, using apache and serving a miserable 100 hits/sec per node all in the same paragraph does not help your credibility much. You're either doing it really, really wrong (cf. epic fail) or you're just trying to sound important without having ever really touched a large scale system.
For reference, an async-io server like nginx, lighty, zeus or similar will easily saturate a Gbit uplink on moderate hardware, either with small files (then it will push upwards of 1000 reqs/sec) or with large files (then your "100 reqs/sec"-statement makes even less sense). Sorry, but thanks for playing.
I just mentioned that because a lot of people really seem to believe that Python is for people who can't handle anything else.
Doh, are people really thinking that? In reality python is just as much java done right as it is perl done right. Remember python (and ruby FWIW) are higher level languages than java. Not the stuff that your average PHP or Perl monkey is going to grasp without a steep learning curve.
Thing is, your dying "$49/month" hosting companies tend to suddenly become *very* expensive when and if you need to scale. In reality the common route is indeed to start out on such a discounter, until it assplodes. And then rent a rack somewhere, fill it with own hardware and move as fast as you can.
Compared to *this* route (which many startups can sing a song about) the amazon prices don't look so hefty anymore.
And frankly, it is definately that part of the scalability story that you should watch out for. What does it matter whether you pay $49/month or ~$90/month (amazon) during the early days?
I don't think those $600 bucks that you may save in the first year will ease you much when you get to spent 5 digits on hardware to handle load spikes that, on amazon, would only cost you a few hundred bucks each...
No. Write it off as yet another rant about the declining quality here. I remember the days when most articles were worth reading and when (gasp!) most of the blurbs actually made sense.
What's happened here in the last 2 years is just sad.
Calm down mate. I'm a bit baffled at how hysteric you get over this. We're talking about filesystems, right?
It's difficult to find a point in your 3rd grade polemics, so I'll just state the obvious: There is no such thing as "the *nix heads" who have decided in a central board meeting to "FUCK NOVELL" and collectively boycot NSS.
The simple fact that NSS has been around for a while but hasn't gained any significant popularity is a strong indicator that it is apparently not as perfect as you paint it here.
The article has 10 pages, each carries 5 banners. Let's assume they are paid a very conservative $.50 USD per one thousand unique visitors for each of these banners. Let's further assume slashdot drove 2 million unique's to the article. Let's further assume those people, on average, clicked through 3 pages before they realized there is nothing to see.
That's a solid $15000 USD, under fairly pessimisic assumptions. They probably made closer to $30000 by the time you are reading this.
Dude, you're making a fool of yourself. If NSS really is this "killer" filesystem (wasn't that term reserved for reiser?) that is better than anything else out there... then why did no "bright dev" ever bother to port it all over the place? Why did nobody even bother to at least make it available on other linux distros?
Why do all these bright hackers rather work on porting ZFS or developing new filesystems (hammer, butter, ext4)?
Maybe because here in the real world, NSS has a reputation for being slow and buggy?
I would think that starts and landings are *the* situations where we would benefit the most from auto-pilot. Ofcourse there should always be an "override" button for when shit really hits the fan.
But in the general case (i.e. landings in heavy weather and with zero sight) a computer should be able to compensate much faster and more accurately than a human. There are no real decisions to make in those situations, it's all about reaction time. Furthermore a computer can "see" and "feel" much better than a human, given a wide enough array of sensors. For example, a ground station could report unpredictable turbulences to the plane before they even arrive there and the plane would prepare to compensate - all in a subsecond timeframe.
As for your question: I know for sure that unmanned drones can start and land safely in auto-pilot. I *think* I have even watched a TV-documentary where they claimed that civil aircraft (with passengers on board) would commonly land (partially?) on auto-pilot, but I'm not entirely sure on that last one.
NSS does have a proprietary tail: It only runs on SuSE linux with the custom SuSE kernel, as I explained in my previous post. Ofcourse you can correct me if I'm wrong, just post a link to an installation tutorial for any other distro (Debian, Gentoo, Redhat etc.).
Well, NSS may be the best filesystem on the planet (many think the same about ZFS) but its close ties to SuSE linux with all the proprietary long-tail is still a showstopper for many if not *most* potential users. Most sysadmins I know (and I know a few) wouldn't touch SuSE with a 10ft pole and not remotely consider it for a server.
Furthermore the features you named are not earth-shaking. Feature-wise it seems to rival ZFS but with a much smaller user-base and much less momentum. It's honestly the first time I hear at all about NSS and I'm sure I'm not alone...
Anyways, I still stand to my point. It'd be nice if novell opened up and made NSS available on all linux distros. New filesystem options on linux are always welcome (reiser being dead, xfs being flakey and ext2/3 showing their age), so even if Novell charged a *reasonable* fee for commercial use I'd think it could gain some ground - if it really proves as solid in production as you describe it.
In reality, though, they'll probably stick to their proprietary route and NSS will stay in its niche until it's eventually obsoleted by a new mainsteam linux filesystem like hammer, butter, ext5, zfs-for-linux or so....
Well, it looks interesting feature-wise but they seem to be explicitly targeting SuSE - which is a no-go for most people. From a glance at the docs (hey, at least they have docs, that's a plus) it also seems like it's tied to specific versions of EVMS and other parts of the kernel, thus if you don't run a "blessed, certified" SuSE kernel with all the nasty patches then you're on your own.
Just google for "debian|gentoo|redhat|... novell nss filesystem". Apparently nobody even tried to run NSS on another distro, or at least didn't write about it.
I, for one, would only touch this on a blackbox, vendor-supported appliance but never consider it for a production server of my own (none of which run SuSE). If they worked towards integrating it into the mainline kernel, now that would be nice.
Slashdot Mirror
Why would someone blame ASUS when a driver update pushed out by microsoft breaks a previously working system?
You mean like stackoverflow.com?
Dear slashdot, we know your traffic figures are declining and you're dying for cheap and effective (ad revenue) content. But really, this is becoming a tragedy...
I doubt that. Slashdot wasn't even around in 1789. Heck, the internet didn't exist in 1789, unless you count offline poker sites as partypoker "beta".
Which is, from today's point of view, a total waste of ressources on all sides (read: a paper tiger).
At least 50% of all computer users are completely clueless and will use whatever came with their PeeCee.
The blue e is "the internet" for them. For once and for all. Joe Sixpack doesn't know and doesn't care that he can remove his blue e or "upgrade" it to a red fox. Why should he? All the sites he goes to work more or or less on the blue e and even if they work less - he's unlikely to notice notice that the boxes don't align or that the rounded corners are off.
Microsoft will keep the lions share of the browser market unless one the following happens:
a) Windows loses significant market share (Linux on the desktop anyone?) ...) start blocking IE and tell Joe Sixpack to "upgrade to the red fox".
b) Windows ships without a webbrowser pre-installed. Instead there is a small widget after installation that asks the user to choose a browser to install ("Ooh, look the cute fox").
c) *Major* sites (CNN, google, porn,
Much easier:
Go to the start-position (from where to delete)
Press v
Move down to the end position
Press d
Google "Human genome project" or pretty much any source about basic genetics.
What do you think causes one of your siblings to develop a different skin color? Right, her genes. And we get better at decoding these every day.
To paraphrase obama: Yes, you can. The different skin-colors stem from genetic differences and those can be detected. I'm not saying that distinguishing between skin-colors or "races" makes any sense - but technically it's possible to determine someone's skin-color from a DNA-sample, with increasing accuracy.
I think you missed the point.
You got it right, almost.
Network transparency is the feature that kills X. Network transparency was nice and revolutionary (cough) when X was invented many years ago. Nowadays it's just a huge pain in the ass because it kills performance and makes the whole thing insanely complex.
Reality has taught us that the whole network transparency thing is better tackled independently from the underlying platform.
Free software like NoMachine and VNC exists today and handles the Remote Desktop Use-Case much better (faster!) than the native X protocol. Moreover they are cross-platform, they don't even care whether your client/server are a unix, Windows or Mac.
A new X server should most definately not even attempt to provide network transparency. It should focus on solid driver support, fast compositing and a sane programming API so people can actually build stuff on top of it. Anyone who has ever worked with Xlib knows that the ridiculous API is the #1 reason why "linux on the desktop" is just not happening.
The primary reason for QT and GTK being the monstrosities they are is that they both have to deal with an effectively broken foundation.
Pidgin is a very poor jabber client. It doesn't support transports at all and lacks pretty much all advanced jabber features.
A better jabber client for linux would be psi (there may be others worth looking at, i don't know) and I'm sure there's something better than pidgin for windows, too.
Maybe google for "jive jabber". They make an opensource jabber server and I think they have a cross platform jabber client, too.
Maybe try digg?
But it's not hard to do at all *when* you're an idiot. I can't count the number of times someone plugged both ends of a cat5-cable into the switch...
Replying to my own post, to sum up what I learned: Nothing...
Maybe I'm an anti-social fag but I *still* don't get it. And the feeling gets stronger that I never will.
Yes I do have friends in the real world but for some reason I really don't care what they are doing at every given point in time. If they do something interesting then I'm sure they'll tell me anyways, probably in more than 128 characters. Moreover I'm already using two messenger programs (Skype and pidgin for all the rest), so I'm already getting more "thinking bubbles", away messages, birthday reminders etc. than I would ever care to read.
From all your arguments only the "persistence" and "zero barrier of entry (no clue required)" stick with me.
I guess these two make the big deal out of twitter. People just love to leave their piss-marks everywhere - and stupid people even more so.
How this got *that* big is still beyond me, though. I guess we, as a species, still have a long way to go...
Maybe someone can finally explain that twitter "phenomena" to me?
I just don't get it. They basically reinvented IRC and instant messaging poorly and put it on the web. Hm, okay, but why do the unwashed masses flock to it like that?
Back on topic: So the guy that couldn't even get the trivial use-case of a large scale pub/sub right is now complaining about the internet architecture? Too much cocaine?
I don't see why you think that ec2 should be discredited as a whole based on a use-case that they're not even remotely targeting?
Ec2 is for application hosting or number crunching. Static content (like porn) is the job for a CDN. Ec2 is not a CDN.
And honestly, claiming you are bandwidth bound, using apache and serving a miserable 100 hits/sec per node all in the same paragraph does not help your credibility much.
You're either doing it really, really wrong (cf. epic fail) or you're just trying to sound important without having ever really touched a large scale system.
For reference, an async-io server like nginx, lighty, zeus or similar will easily saturate a Gbit uplink on moderate hardware, either with small files (then it will push upwards of 1000 reqs/sec) or with large files (then your "100 reqs/sec"-statement makes even less sense). Sorry, but thanks for playing.
Doh, are people really thinking that?
In reality python is just as much java done right as it is perl done right. Remember python (and ruby FWIW) are higher level languages than java. Not the stuff that your average PHP or Perl monkey is going to grasp without a steep learning curve.
Your definition of "cheap" amuses me.
Thing is, your dying "$49/month" hosting companies tend to suddenly become *very* expensive when and if you need to scale.
In reality the common route is indeed to start out on such a discounter, until it assplodes. And then rent a rack somewhere, fill it with own hardware and move as fast as you can.
Compared to *this* route (which many startups can sing a song about) the amazon prices don't look so hefty anymore.
And frankly, it is definately that part of the scalability story that you should watch out for.
What does it matter whether you pay $49/month or ~$90/month (amazon) during the early days?
I don't think those $600 bucks that you may save in the first year will ease you much when you get to spent 5 digits on hardware to handle load spikes that, on amazon, would only cost you a few hundred bucks each...
No. Write it off as yet another rant about the declining quality here.
I remember the days when most articles were worth reading and when (gasp!) most of the blurbs actually made sense.
What's happened here in the last 2 years is just sad.
Calm down mate. I'm a bit baffled at how hysteric you get over this. We're talking about filesystems, right?
It's difficult to find a point in your 3rd grade polemics, so I'll just state the obvious:
There is no such thing as "the *nix heads" who have decided in a central board meeting to "FUCK NOVELL" and collectively boycot NSS.
The simple fact that NSS has been around for a while but hasn't gained any significant popularity is a strong indicator that it is apparently not as perfect as you paint it here.
The "we have no clue but slashvertisments pay"-kind.
Slashdot is declining but still attracts roughly 8 million page views per day.
The article has 10 pages, each carries 5 banners.
Let's assume they are paid a very conservative $.50 USD per one thousand unique visitors for each of these banners.
Let's further assume slashdot drove 2 million unique's to the article.
Let's further assume those people, on average, clicked through 3 pages before they realized there is nothing to see.
That's a solid $15000 USD, under fairly pessimisic assumptions. They probably made closer to $30000 by the time you are reading this.
Dude, you're making a fool of yourself. If NSS really is this "killer" filesystem (wasn't that term reserved for reiser?) that is better than anything else out there... then why did no "bright dev" ever bother to port it all over the place? Why did nobody even bother to at least make it available on other linux distros?
Why do all these bright hackers rather work on porting ZFS or developing new filesystems (hammer, butter, ext4)?
Maybe because here in the real world, NSS has a reputation for being slow and buggy?
I would think that starts and landings are *the* situations where we would benefit the most from auto-pilot. Ofcourse there should always be an "override" button for when shit really hits the fan.
But in the general case (i.e. landings in heavy weather and with zero sight) a computer should be able to compensate much faster and more accurately than a human. There are no real decisions to make in those situations, it's all about reaction time. Furthermore a computer can "see" and "feel" much better than a human, given a wide enough array of sensors. For example, a ground station could report unpredictable turbulences to the plane before they even arrive there and the plane would prepare to compensate - all in a subsecond timeframe.
As for your question: I know for sure that unmanned drones can start and land safely in auto-pilot. I *think* I have even watched a TV-documentary where they claimed that civil aircraft (with passengers on board) would commonly land (partially?) on auto-pilot, but I'm not entirely sure on that last one.
I'm sure a real pilot can clear this up :)
NSS does have a proprietary tail: It only runs on SuSE linux with the custom SuSE kernel, as I explained in my previous post.
Ofcourse you can correct me if I'm wrong, just post a link to an installation tutorial for any other distro (Debian, Gentoo, Redhat etc.).
Well, NSS may be the best filesystem on the planet (many think the same about ZFS) but its close ties to SuSE linux with all the proprietary long-tail is still a showstopper for many if not *most* potential users. Most sysadmins I know (and I know a few) wouldn't touch SuSE with a 10ft pole and not remotely consider it for a server.
Furthermore the features you named are not earth-shaking. Feature-wise it seems to rival ZFS but with a much smaller user-base and much less momentum. It's honestly the first time I hear at all about NSS and I'm sure I'm not alone...
Anyways, I still stand to my point. It'd be nice if novell opened up and made NSS available on all linux distros. New filesystem options on linux are always welcome (reiser being dead, xfs being flakey and ext2/3 showing their age), so even if Novell charged a *reasonable* fee for commercial use I'd think it could gain some ground - if it really proves as solid in production as you describe it.
In reality, though, they'll probably stick to their proprietary route and NSS will stay in its niche until it's eventually obsoleted by a new mainsteam linux filesystem like hammer, butter, ext5, zfs-for-linux or so....
Well, it looks interesting feature-wise but they seem to be explicitly targeting SuSE - which is a no-go for most people.
From a glance at the docs (hey, at least they have docs, that's a plus) it also seems like it's tied to specific versions of EVMS and other parts of the kernel, thus if you don't run a "blessed, certified" SuSE kernel with all the nasty patches then you're on your own.
Just google for "debian|gentoo|redhat|... novell nss filesystem". Apparently nobody even tried to run NSS on another distro, or at least didn't write about it.
I, for one, would only touch this on a blackbox, vendor-supported appliance but never consider it for a production server of my own (none of which run SuSE).
If they worked towards integrating it into the mainline kernel, now that would be nice.