SCUBA divers experience sudden pressure changes in the realm of 15 PSI all the time. They don't "explode," they just get the bends.
Yeah well, they never experience absolute pressures below 15 PSI though. Maybe your organs can withstand pressure loads better than tensile loads:-P
I thought they launch from Florida. What does it matter what some restaurant owner in Houston says about how much astronauts drink while off-duty? Or do they travel from Houston to then cape less than 12 hours before launch?
OK, here's my results (copying bitmap images, just tried):
- Gimp -> OOo works (impressive indeed)
- Opera -> OOo works
- FF -> OOo doesn't work (no "copy" in FF)
- Gimp -> Skencil doesn't work
- Gimp -> Dia doesn't
- Gimp/Opera -> Inkscape doesn't ("nothing in the clipboard" says Inkscape)
Copying text works better generally, but not universally either. Copying into GNU Emacs (e.g. from Eclipse or Acrobat reader) sometimes works, sometimes Emacs in its unfathomable, transcendent mind decides to cling to its own clipboard contents. (Yes I did (setq x-select-enable-clipboard t), and I even did grok http://www.jwz.org/doc/x-cut-and-paste.html). Sometimes it helps to just change the clipboard in Emacs first, sometimes it helps to not copy directly from app A to B, but from A to xterm (or xfce4-terminal in my case), then from xterm to B.
I'm sure all those phenomenae weren't created intentionally; they just arise by chance, because the APIs are too strange, some programmers interpret the ICCCM differently than others, or whatever.
Yeah, just convince all UI programmers to put in place a moratorium on new features until copy&paste works between all applications, all the time, with all commonly used kinds of data. Once that's done (2030 or so), world domination will become reality on short notice.
..in which Al-Qaida terrorists planned and executed the 9/11 attacks, in which 12 Americans landed on the moon and returned safely to the earth, in which passive smoking is dangerous, in which a man-made global warming takes place, and in which a single depressed loser named Lee Harvey Oswald killed John F. Kennedy.
Because this is all too boring to bear (and too unrewarding for book authors with large egos), people invent themselves a parallel universe in which all of the above is an illusion carefully crafted by the CIA, large corporations, scientists or the government.
This is a general advice to everybody: It's stupid enough to not RTFA generally, but here it is particularly idiotic because the FA specifically counters common arguments brought up by deniers, so if you just think "this is the 101st global warming discussion, so instead of R'ingTFA let's make cheap points by throwing in in our generic anti-GW prejud^H^H^H^H^H^Harguments", you're destined to make yourself look like a complete dumbass. So go right ahead, read the damn thing already.
We can point to stuff that correlates to *continuing* climate change in the past. Once climate change starts, you can point to some thing that seems to track with it well. This seems to be where CO2 sits mostly. That silly 800 year lag before CO2 starts tracking temperature.
But we don't have any idea what event or combination of events was the initial cause of any climate change event in the past.
Astonishing -- even people with very low Slashdot IDs don't bother to RTFA.
So just RTFA, it addresses these points specifically -- we have a very good idea what initially caused climate changes in the past glacial-interglacial transitions, and what the feedback factors were, and where that 800 year time lag comes from.
The greenhouse gas emissions created by the human race are about 3-5% of the total. The rest comes from the planet itself.
Get your numbers right (don't take them from obscure global warming sceptics' sites, for starters). You're confusing the oceans' (and land masses') total CO2 emissions (which are indeed much higher than ours) with the ocean's *net* CO2 emissions (which are *negative* -- the oceans currently absorb more than they emit, slowing the CO2 level increase in the atmosphere -- CO2 concentrations in the oceans are rising, all measurements show that). The CO2 concentration in the air is higher today than it was in the last 600,000 years or more, we also have direct evidence (carbon isotopes) that much of the CO2 in the atmosphere comes from fossil fuel burning, and if you want, you can take the total amount of CO2 released into the air since 1800, divide it by the total number of molecules in the atmosphere and see for yourself that the current CO2 concentration is not a "thing much bigger than we are". About one in three CO2 molecules in the atmosphere originates from human activities, there is no scientific dispute about that.
| Think of subsystems as being like shells with system-specific behavior.
Or they are complete OS environments that on top of the NT core.
Well, considering that those "subsystems" run in user space, how is such a "subsystem" different from a binary execution format and a bunch of libraries that provide some API that differs from the kernel's?
And then, the GDI and window manager are not in any subsystem, but in the kernel, which is certainly not where they belong. Various window manager attributes like initial window positions etc. were slapped into the process creation APIs for good measure, which is quite a violation of layering principles -- as opposed to X, which, for all its shortcomings, provides just an ordinary client-side library that applications may or may not use.
Kernel space drivers have the ability of taking down the whole system where userland drivers do not.
If you access the graphics hardware from userspace via hacks like/dev/kmem, like XFree86 traditionally did (and still does in many installations), you have a userspace process that has the ability of taking down the whole system. How is that any better?
In fact, if you want to do graphics, some piece of software has to access the graphics hardware directly, and that piece always has the ability to take down the whole system. If that's the case, it's a much better idea to put that piece into the kernel where it belongs anyway.
Oh, nonsense. Look inside just about any router, and you'll see lots of code that knows about connection-oriented protocols like TCP and RTP. Routers can and do implement various QoS schemes. There are a number of big companies that would love to sell you boxes that do such things. And if you can't find a router that implements exactly the stuff you want, buy yourself a linux or BSD box, install all the source code, and implement it yourself.
I'm not talking about software routers. I also have a neat shiny blinking self-built linux router here that does all kinds of stuff, from stateful packet filtering to NAT and traffic shaping. However, while this is nice for home or company LAN routers, it's too slow for backbone routers. I'm not an expert, but I know that backbone routers route in hardware for speed reasons. No routing software running on any available processor would be fast enough to handle the traffic that backbone routers must handle. And those routers *do not* know about TCP (or RTP). Unfortunately, implementing hard realtime constraints for connection-oriented protocols would only work if all routers along the end-to-end path implemented it.
Well, there are things that can't be implemented on top of the existing internet infrastructure, but would have to be buried into the core protocols and infrastructure itself. E.g. configurable, guaranteed upper bounds for the latency and/or lower bounds for the throughput, and not just on the packet level, but throughout the lifetime of a connection. You can't do that with the current IP protocol and router infrastructure because routers only know about IP and have no concept of connections, let alone required QoS properties of connections.
Well, dissent against random stupid government policy XYZ may be "the highest form of patriotism", dissent against scientific consensus for no good reason is just stupid. Who would call dissenting from gravity (or evolution) "patriotic"?
The reason I suggested staying with Etch for a little while is that there is likely to be some breakage in Testing as the backlog of Unstable updates move into Testing.
I may be mistaken here, but isn't the old unstable (sid) the new unstable, not the new testing? I was under the impression that the previous testing (Etch) is just duplicated into stable and renamed to "Lenny" in the testing branch, and nothing happens to unstable (i.e. it's not like the old unstable is becoming the new testing or anything).
To support POP3, they would also have to support SMTP to send out. And my guess is that by doing this, they can't lock down your email as tight as they apparently want. By giving up POP3 and SMTP, they ensure students won't be Spamming or accidentally letting a trojan on a student's computer attempt to send infected email.
Sorry but this sounds like plain FUD on your part. The exchange protocol certainly has some groupware features that POP/IMAP/SMTP lack, but how is it inherently better than, say, SMTP-Auth over SSL, in preventing users from unintentionally sending spam? If the user can send mail, he can send spam, and if some malicious program is running, it can send mass mails automatically. And of course you can still "lock the control to the server" with SMTP as well -- all mails have to go through it, after all. The server can implement some advanced sender authentication schemes like SPF or challenge-response schemes like hash cash or whatever. So, how can you say that the exchange protocol, which you admit you don't even know, is better here?
With that in mind I'd surely choose a protocol that leaves me free choice to use whatever advanced mail client I want, as opposed to the choice between a single specific MS-only client and a web client, thank you very much.
I learnt JavaScript from the standard. Took about two weeks. Yes, the standard is difficult to read, but at least it's complete, which cannot be said about virtually all the books and tutorials about JavaScript -- indeed, if you've read those, you may think you know everything when in fact half the features of the language have completely eluded you because they weren't even mentioned.
Unfortunately, I found out about Douglas Crockford's lectures only *after* reading the spec:-\
And it has been statistically proven, many times, that the more trained citizens carry guns, the lower the crime rate in the area is.
Huh? The murder rate in the US is higher than in virtually all other western societies, despite the latter having much, much stricter gun control laws and much less draconic punishments for murder or other crimes. Or doesn't the US count as an "area"?:-P
And if you really think your police is "overworked" I suggest hiring more policemen...
Technical definition, please. And how does the system know? Isn't it broken as designed that the system has to know about "setup programs"? I always thought they're just programs like any other one (as they should be).
Well, I can't read the names on those graphs, but a system call is only ever entered and left, it never calls out to another system call, so there can't be a spaghettiesque graph of system calls calling each other.
It has fewer reported vulnerabilities, fewer unpatched vulnerabilities (zero), etc.
You're not counting all the reports published by Linux distributors against any of the thousands of programs that make up an average Linux distribution against all the reports Microsoft publishes against the core Windows OS, are you? MS Word alone has 3 or 4(?) unpatched vulnerabilities atm., iirc.
Slashdot Mirror
SCUBA divers experience sudden pressure changes in the realm of 15 PSI all the time. They don't "explode," they just get the bends. Yeah well, they never experience absolute pressures below 15 PSI though. Maybe your organs can withstand pressure loads better than tensile loads :-P
I do remember that one. IIRC, Crusher's advice to Geordi was not to exhale under any circumstances. If the FA is right, that seems like a bad idea.
I thought they launch from Florida. What does it matter what some restaurant owner in Houston says about how much astronauts drink while off-duty? Or do they travel from Houston to then cape less than 12 hours before launch?
- Gimp -> OOo works (impressive indeed)
- Opera -> OOo works
- FF -> OOo doesn't work (no "copy" in FF)
- Gimp -> Skencil doesn't work
- Gimp -> Dia doesn't
- Gimp/Opera -> Inkscape doesn't ("nothing in the clipboard" says Inkscape)
Copying text works better generally, but not universally either. Copying into GNU Emacs (e.g. from Eclipse or Acrobat reader) sometimes works, sometimes Emacs in its unfathomable, transcendent mind decides to cling to its own clipboard contents. (Yes I did (setq x-select-enable-clipboard t), and I even did grok http://www.jwz.org/doc/x-cut-and-paste.html). Sometimes it helps to just change the clipboard in Emacs first, sometimes it helps to not copy directly from app A to B, but from A to xterm (or xfce4-terminal in my case), then from xterm to B.
I'm sure all those phenomenae weren't created intentionally; they just arise by chance, because the APIs are too strange, some programmers interpret the ICCCM differently than others, or whatever.
Yeah, just convince all UI programmers to put in place a moratorium on new features until copy&paste works between all applications, all the time, with all commonly used kinds of data. Once that's done (2030 or so), world domination will become reality on short notice.
Because this is all too boring to bear (and too unrewarding for book authors with large egos), people invent themselves a parallel universe in which all of the above is an illusion carefully crafted by the CIA, large corporations, scientists or the government.
This is a general advice to everybody: It's stupid enough to not RTFA generally, but here it is particularly idiotic because the FA specifically counters common arguments brought up by deniers, so if you just think "this is the 101st global warming discussion, so instead of R'ingTFA let's make cheap points by throwing in in our generic anti-GW prejud^H^H^H^H^H^Harguments", you're destined to make yourself look like a complete dumbass. So go right ahead, read the damn thing already.
So just RTFA, it addresses these points specifically -- we have a very good idea what initially caused climate changes in the past glacial-interglacial transitions, and what the feedback factors were, and where that 800 year time lag comes from.
Get your numbers right (don't take them from obscure global warming sceptics' sites, for starters). You're confusing the oceans' (and land masses') total CO2 emissions (which are indeed much higher than ours) with the ocean's *net* CO2 emissions (which are *negative* -- the oceans currently absorb more than they emit, slowing the CO2 level increase in the atmosphere -- CO2 concentrations in the oceans are rising, all measurements show that). The CO2 concentration in the air is higher today than it was in the last 600,000 years or more, we also have direct evidence (carbon isotopes) that much of the CO2 in the atmosphere comes from fossil fuel burning, and if you want, you can take the total amount of CO2 released into the air since 1800, divide it by the total number of molecules in the atmosphere and see for yourself that the current CO2 concentration is not a "thing much bigger than we are". About one in three CO2 molecules in the atmosphere originates from human activities, there is no scientific dispute about that.
Well, considering that those "subsystems" run in user space, how is such a "subsystem" different from a binary execution format and a bunch of libraries that provide some API that differs from the kernel's?
And then, the GDI and window manager are not in any subsystem, but in the kernel, which is certainly not where they belong. Various window manager attributes like initial window positions etc. were slapped into the process creation APIs for good measure, which is quite a violation of layering principles -- as opposed to X, which, for all its shortcomings, provides just an ordinary client-side library that applications may or may not use.
In fact, if you want to do graphics, some piece of software has to access the graphics hardware directly, and that piece always has the ability to take down the whole system. If that's the case, it's a much better idea to put that piece into the kernel where it belongs anyway.
I'm not talking about software routers. I also have a neat shiny blinking self-built linux router here that does all kinds of stuff, from stateful packet filtering to NAT and traffic shaping. However, while this is nice for home or company LAN routers, it's too slow for backbone routers. I'm not an expert, but I know that backbone routers route in hardware for speed reasons. No routing software running on any available processor would be fast enough to handle the traffic that backbone routers must handle. And those routers *do not* know about TCP (or RTP). Unfortunately, implementing hard realtime constraints for connection-oriented protocols would only work if all routers along the end-to-end path implemented it.
Well, there are things that can't be implemented on top of the existing internet infrastructure, but would have to be buried into the core protocols and infrastructure itself. E.g. configurable, guaranteed upper bounds for the latency and/or lower bounds for the throughput, and not just on the packet level, but throughout the lifetime of a connection. You can't do that with the current IP protocol and router infrastructure because routers only know about IP and have no concept of connections, let alone required QoS properties of connections.
Well, dissent against random stupid government policy XYZ may be "the highest form of patriotism", dissent against scientific consensus for no good reason is just stupid. Who would call dissenting from gravity (or evolution) "patriotic"?
I may be mistaken here, but isn't the old unstable (sid) the new unstable, not the new testing? I was under the impression that the previous testing (Etch) is just duplicated into stable and renamed to "Lenny" in the testing branch, and nothing happens to unstable (i.e. it's not like the old unstable is becoming the new testing or anything).
Sorry but this sounds like plain FUD on your part. The exchange protocol certainly has some groupware features that POP/IMAP/SMTP lack, but how is it inherently better than, say, SMTP-Auth over SSL, in preventing users from unintentionally sending spam? If the user can send mail, he can send spam, and if some malicious program is running, it can send mass mails automatically. And of course you can still "lock the control to the server" with SMTP as well -- all mails have to go through it, after all. The server can implement some advanced sender authentication schemes like SPF or challenge-response schemes like hash cash or whatever. So, how can you say that the exchange protocol, which you admit you don't even know, is better here?
With that in mind I'd surely choose a protocol that leaves me free choice to use whatever advanced mail client I want, as opposed to the choice between a single specific MS-only client and a web client, thank you very much.
Unfortunately, I found out about Douglas Crockford's lectures only *after* reading the spec :-\
Huh? The murder rate in the US is higher than in virtually all other western societies, despite the latter having much, much stricter gun control laws and much less draconic punishments for murder or other crimes. Or doesn't the US count as an "area"? :-P
And if you really think your police is "overworked" I suggest hiring more policemen...
Technical definition, please. And how does the system know? Isn't it broken as designed that the system has to know about "setup programs"? I always thought they're just programs like any other one (as they should be).
Well, I can't read the names on those graphs, but a system call is only ever entered and left, it never calls out to another system call, so there can't be a spaghettiesque graph of system calls calling each other.
You're not counting all the reports published by Linux distributors against any of the thousands of programs that make up an average Linux distribution against all the reports Microsoft publishes against the core Windows OS, are you? MS Word alone has 3 or 4(?) unpatched vulnerabilities atm., iirc.
Flash 9 for Linux uses Alsa (and only Alsa, no OSS anymore), and it "plays nice" with other Alsa-using sound apps just fine.