I go to sleep when I'm sleepy, and I wake up when I'm rested. This seems to result in a 25 to 28 hour cycle, depending on what time of day I start the "day" and what I'm doing. (For some reason, having new video games around makes the cycle longer . ..)
Every once in a while I run into problems at work thanks to this (I'm a coder), and some people just don't seem to understand it. It can also make meeting up with friends more difficult. However, I'm far more productive - partly because I often end up working at midnight when there's nobody around, and partly because a 24-hour schedule just doesn't work for me. (I tried it for something like six years - it stopped working in high school, and I dropped out of two colleges thanks to that 24-hour schedule before I said "fuck this".)
I don't consider my sleeping pattern to be a strength - it's what works for me, but it would probably be better if the 24-hour cycle worked for me. It doesn't, though, so I do this.
I'm writing this at 1:17 AM while I eat breakfast.:)
Some people don't want their code to be used in any situation. They want to guarantee that anyone trying to profit off their code will basically have to contribute in one way or another - the code can't just be copied and closed.
Personally I use both GPL and BSD for different projects, but saying "the GPL is too complex and that's why people should use BSD/MIT" really ignores the reason why many people use the GPL in the first place. I agree that we could use a simpler version of the GPL - but BSD isn't it.
Neither. The problem wasn't a bug in the implementation - it was a dubious file format. WINE provided full support for the file format, which they shouldn't have, because the file format was crappy to begin with.
Imagine if zip files contained an arbitrary binary payload that was supposed to be executed when any file was extracted - anyone who "properly" implemented the standard would be vulnerable to that binary payload containing a virus. That's basically what happened here.
How, exactly, is this substantially different from a reboot?:P
Okay. The kernel doesn't reset. So what? If all the servers go down, I don't much care about the kernel. And, y'know, not all servers shut themselves down properly in all cases . . . it's just safer, simpler, and less error-prone to reboot the damn thing. Windows *or* Linux.
I've only seen them do so on Windows if they're extremely invasive (to the point where you need administrator privs to run them, because they do stuff on the kernel level) or if you have flaky drivers.
For that matter, recently the only problem along those lines I've had was flaky-drivers.
Seriously - 80% of the time, it's the drivers. 19.9% of the time, it's crappy hardware. And 0.1% of the time, it's cosmic rays.
Obviously all these numbers go out the window if you have spyware.
Applications crash. Drivers crash. Hardware crashes. Windows itself is quite stable. I had a 150-day uptime on the box I'm typing on right now (WinXP) until I had a power outage.
I've seen drivers crash, and I've seen flaky hardware cause problems, and I've seen combinations of the two become an issue. But Windows itself? Pretty damn near rock solid. It gets a nice reputation for instability because so many manufacturers put it on bottom-basement gimpy hardware, but I seriously doubt Linux would fare any better.
Linux might have better drivers, of course - as long as you don't ask it to do anything heavy in 3d.
I do something very similar - zorba-slashdotcomment@pavlovian.net for example. (Now if I get spam harvested, I'll know where it came from.;) )
Amusingly, I found that 99% of my spam comes from people just bruteforcing everything at pavlovian.net - everything at that domain, prefixed with "zorba" or not, gets redirected to me. So I've got two addresses that I've never used, anywhere, ever, blackholed.
Still get an amazing amount of spam on those. Go figure.
It'd simply never allow unpopular entries to be edited. A while back I made some updates to the article on the Top Ten comic series - how many months would that have taken to show up with this system?
You should weight it based on the popularity of the page.
You know why they don't support the HD resolutions?
Because the consoles aren't fast enough.
Yes, they *technically* have support for those modes . . . but it takes far more of the hardware's capability than the low-res modes, and the graphics suffer as a result.
That's true, but that's why his objection of "I don't see why anyone would share unfinished files" doesn't apply in this case.:) I agree that it makes perfect sense across, say, FTP or IRC. But in the (now) more common case of P2P networks, sharing incomplete files is a Good Thing.
As the other responder mentioned, most (if not all) modern networks will cheerfully share the parts of incomplete files that you've already downloaded. It improves download rates, since by the time the uploader has sent the entire file once, you can easily have hundreds of people with half of it complete.
The holes they know of are fixed. This is an attempt to make sure new inevitable holes (which all large programs have, Firefox is no exception) won't be nearly as serious.
The thing that replaces the WWW won't do it through long rambling vague papers. It won't be accomplished by proclamations, or promises, or explanations of Why It's Better And Why You Should All Use My System, Please.
It'll be an experiment that someone creates. And someone else will take it and say "hey, this is kind of neat" and do something with it. And then it will spread, and five years later we'll wonder why we didn't do it long ago.
An interview like this, as far as I'm concerned, is rock-solid proof that what he's talking about will never happen - because if it was going to, he could just write it and it would become the standard.
Honestly, I'm finding that I need just as many reboots on Linux now as I do Windows. My main Windows box has reached 88 day uptime now, and the only recent reboots (i.e. in the past year) have been power outages. (One of them was my UPS dying. Sigh.) I admittedly don't install patches often, but then again it's behind a firewall.
Meanwhile, the last time I updated my Linux box, it told me that it was updating libc and that I should reboot. So I did. Its uptime is now 34 days. Not as impressive.
Of course, the aforementioned firewall runs BSD and is up to 177 days. But it doesn't run anything all that interesting, and never gets updated - it's practically an embedded device as far as I'm concerned.
My games Windows box gets "rebooted" far more often - namely because it gets shut down at night, and also because it has some interesting heat problems and occasionally reboots itself thanks to that. However I find it very doubtful that Linux would do much better in that situation.:P
At work, my Linux and Windows boxes need to be rebooted roughly equally - the Windows box because of the occasional security patch, and the Linux box because of bizarre bugs they're still trying to fix. (In fairness, the Linux box does far more than the Windows box does.) But in any case, Windows has come a long way from its "reboot once a day" legacy, while Linux has come a long way from its "never need to reboot for anything" legacy. They both seem to be converging at a happy medium.
Personally, I use the "unread" flag for this. I have my inbox sorted first by whether I've read or not, and second by date. When I go into my email-read cycle, I start at the bottom unread message that I haven't seen (I generally have at most a dozen outstanding "this needs to be taken care of" or "these are notes to myself" emails at a time) and go up, marking them as unread again after I've glanced at them.
So my inbox always has less than a dozen unread important messages after I've finished reading. Works pretty well.
'Course, it has thousands of read messages, but those I just ignore. They're Not Important. Every once in a while they go into a big archive directory and just live there for eternity (I never delete emails, disk space is too cheap and information is too expensive.)
Slashdot Mirror
Liar!
I go to sleep when I'm sleepy, and I wake up when I'm rested. This seems to result in a 25 to 28 hour cycle, depending on what time of day I start the "day" and what I'm doing. (For some reason, having new video games around makes the cycle longer . . .)
:)
Every once in a while I run into problems at work thanks to this (I'm a coder), and some people just don't seem to understand it. It can also make meeting up with friends more difficult. However, I'm far more productive - partly because I often end up working at midnight when there's nobody around, and partly because a 24-hour schedule just doesn't work for me. (I tried it for something like six years - it stopped working in high school, and I dropped out of two colleges thanks to that 24-hour schedule before I said "fuck this".)
I don't consider my sleeping pattern to be a strength - it's what works for me, but it would probably be better if the 24-hour cycle worked for me. It doesn't, though, so I do this.
I'm writing this at 1:17 AM while I eat breakfast.
So you wake up, find yourself useless and hating life, say "well, it was worth a try", and kill yourself. What's the problem?
Personally, I think I'm going to do this - I mean, it can't be *worse* than just dying.
Some people don't want their code to be used in any situation. They want to guarantee that anyone trying to profit off their code will basically have to contribute in one way or another - the code can't just be copied and closed.
Personally I use both GPL and BSD for different projects, but saying "the GPL is too complex and that's why people should use BSD/MIT" really ignores the reason why many people use the GPL in the first place. I agree that we could use a simpler version of the GPL - but BSD isn't it.
Now that's just stupid, and I don't believe it for a second.
Neither. The problem wasn't a bug in the implementation - it was a dubious file format. WINE provided full support for the file format, which they shouldn't have, because the file format was crappy to begin with.
Imagine if zip files contained an arbitrary binary payload that was supposed to be executed when any file was extracted - anyone who "properly" implemented the standard would be vulnerable to that binary payload containing a virus. That's basically what happened here.
True.
:)
But if the design error is a government backdoor, open-source didn't provide any immunity to it.
Yes, because it's impossible for an identical problem to exist in WINE, and therefore open source solves all problems.
How, exactly, is this substantially different from a reboot? :P
Okay. The kernel doesn't reset. So what? If all the servers go down, I don't much care about the kernel. And, y'know, not all servers shut themselves down properly in all cases . . . it's just safer, simpler, and less error-prone to reboot the damn thing. Windows *or* Linux.
I've only seen them do so on Windows if they're extremely invasive (to the point where you need administrator privs to run them, because they do stuff on the kernel level) or if you have flaky drivers.
For that matter, recently the only problem along those lines I've had was flaky-drivers.
Seriously - 80% of the time, it's the drivers. 19.9% of the time, it's crappy hardware. And 0.1% of the time, it's cosmic rays.
Obviously all these numbers go out the window if you have spyware.
I use Firefox and Thunderbird, and have a very restrictive BSD firewall between me and the Internet.
I never said Windows was great on security - just that the core OS was stable.
(Amusingly, I actually had to reboot my Linux server more recently than my Windows box - Debian updated glibc.)
Windows doesn't crash.
Applications crash. Drivers crash. Hardware crashes. Windows itself is quite stable. I had a 150-day uptime on the box I'm typing on right now (WinXP) until I had a power outage.
I've seen drivers crash, and I've seen flaky hardware cause problems, and I've seen combinations of the two become an issue. But Windows itself? Pretty damn near rock solid. It gets a nice reputation for instability because so many manufacturers put it on bottom-basement gimpy hardware, but I seriously doubt Linux would fare any better.
Linux might have better drivers, of course - as long as you don't ask it to do anything heavy in 3d.
I do something very similar - zorba-slashdotcomment@pavlovian.net for example. (Now if I get spam harvested, I'll know where it came from. ;) )
Amusingly, I found that 99% of my spam comes from people just bruteforcing everything at pavlovian.net - everything at that domain, prefixed with "zorba" or not, gets redirected to me. So I've got two addresses that I've never used, anywhere, ever, blackholed.
Still get an amazing amount of spam on those. Go figure.
It'd simply never allow unpopular entries to be edited. A while back I made some updates to the article on the Top Ten comic series - how many months would that have taken to show up with this system?
You should weight it based on the popularity of the page.
That would require the admins and editors to do work. I've seen very little sign of interest in that.
If it is, they've sacrificed polygons, fill rate, and/or video RAM for it. Or some combination of the three. TANSTAAFL.
You know why they don't support the HD resolutions?
Because the consoles aren't fast enough.
Yes, they *technically* have support for those modes . . . but it takes far more of the hardware's capability than the low-res modes, and the graphics suffer as a result.
That's true, but that's why his objection of "I don't see why anyone would share unfinished files" doesn't apply in this case. :) I agree that it makes perfect sense across, say, FTP or IRC. But in the (now) more common case of P2P networks, sharing incomplete files is a Good Thing.
As the other responder mentioned, most (if not all) modern networks will cheerfully share the parts of incomplete files that you've already downloaded. It improves download rates, since by the time the uploader has sent the entire file once, you can easily have hundreds of people with half of it complete.
The holes they know of are fixed. This is an attempt to make sure new inevitable holes (which all large programs have, Firefox is no exception) won't be nearly as serious.
Toothpaste costs a few dollars.
Laptops cost a few thousand dollars.
See why buying a laptop is a little more stressful than buying toothpaste?
This assumes the offices and the data center are in close physical proximity. That's not always - or even often, from what I know - the case.
The thing that replaces the WWW won't do it through long rambling vague papers. It won't be accomplished by proclamations, or promises, or explanations of Why It's Better And Why You Should All Use My System, Please.
It'll be an experiment that someone creates. And someone else will take it and say "hey, this is kind of neat" and do something with it. And then it will spread, and five years later we'll wonder why we didn't do it long ago.
An interview like this, as far as I'm concerned, is rock-solid proof that what he's talking about will never happen - because if it was going to, he could just write it and it would become the standard.
Honestly, I'm finding that I need just as many reboots on Linux now as I do Windows. My main Windows box has reached 88 day uptime now, and the only recent reboots (i.e. in the past year) have been power outages. (One of them was my UPS dying. Sigh.) I admittedly don't install patches often, but then again it's behind a firewall.
:P
Meanwhile, the last time I updated my Linux box, it told me that it was updating libc and that I should reboot. So I did. Its uptime is now 34 days. Not as impressive.
Of course, the aforementioned firewall runs BSD and is up to 177 days. But it doesn't run anything all that interesting, and never gets updated - it's practically an embedded device as far as I'm concerned.
My games Windows box gets "rebooted" far more often - namely because it gets shut down at night, and also because it has some interesting heat problems and occasionally reboots itself thanks to that. However I find it very doubtful that Linux would do much better in that situation.
At work, my Linux and Windows boxes need to be rebooted roughly equally - the Windows box because of the occasional security patch, and the Linux box because of bizarre bugs they're still trying to fix. (In fairness, the Linux box does far more than the Windows box does.) But in any case, Windows has come a long way from its "reboot once a day" legacy, while Linux has come a long way from its "never need to reboot for anything" legacy. They both seem to be converging at a happy medium.
Personally, I use the "unread" flag for this. I have my inbox sorted first by whether I've read or not, and second by date. When I go into my email-read cycle, I start at the bottom unread message that I haven't seen (I generally have at most a dozen outstanding "this needs to be taken care of" or "these are notes to myself" emails at a time) and go up, marking them as unread again after I've glanced at them.
So my inbox always has less than a dozen unread important messages after I've finished reading. Works pretty well.
'Course, it has thousands of read messages, but those I just ignore. They're Not Important. Every once in a while they go into a big archive directory and just live there for eternity (I never delete emails, disk space is too cheap and information is too expensive.)