The funny thing is... that his joke was actually hacker-funny, because he got the syntax right: Forth is *postfix*, not *backward infix*--the verbs go at the end (top) of the sentence (stack).
I could say that like this:
postfix Forth is infix backward Forth is not and sentence end verbs goto therefor
Actually those "Ext4" data corruption issues that set the Internet all ablaze (including Slashdot) were mainly due to KDE 4 not handling metadata correctly. In the end, it wasn't an Ext4 issue. However, feel free to spread FUD.
Bullshit. The error was very generic and not limited to KDE
You're right: it's a common application programming error.
At what frequency does a bug cease to be a bug?
I wish I'd known that I could promote myself up out of the rank of `amateur' just by pointing out that there were plenty of other people who weren't any more skilled than I was--that would have obviated all of these years of study and hard work.
The real problem with satellites is that the predictable orbit allows the enemy to hide his shit when they're overhead.
Only if he can get it out, use it, and then put it away again all within the time that it takes a satellite to complete an orbit, which takes a lot less time than you think it does (hint: look for the word "minutes").
I read an anecdote somewhere that went something like this:
I asked a programmer friend, whose native language was something other than English, whether he was bothered by all of the hold of English over programming.
He responded by asking me, "Are you at all musically inclined?"
When I said that I was, he asked, "Does it bother you that all of the musical vocabulary is Italian?"
When I said, "No, of course not.", he said, "Well, it's the same thing--it's just an artifact, that the thing has a vocabulary from wherever it developed."
This is not appropriate for desktop systems -- desktop systems have to be robust against all kinds of stupid situations, like sudden power losses, users hitting the reset button because an application is hanging, and so forth.
First off, PC hardware is not robust against sudden power-losses: it's literally possible for a write to be `half-done' inside the HDD, and no amount of higher-level `protection' can do anything about that.
Secondly, the atomicity of rename() (or any other operation) isn't contagious: rename() does an unlink() and a link() in `one operation' (and that's the only amount of atomicity rename() claims--it's even specifically not atomic in that it's possible to see the original link and the new link at the same time), but that doesn't have any impact--or even any relation to other `nearby' operations.
write() and rename() aren't even operating on the same object--write() is operating on the file, and rename() is operating on the directory (or directories that do/will contain links to the file). I don't quite get the `metadata' arguments, because `filenames' aren't `file metadata'--they're directory data, which is what allows you to have any number of links to the same file from any number of directories. File-metadata are things like timestamps, ownership, permissions....
Lastly, having said all that: the reason that we go through the write-close-rename sequence is to prevent a race-condition while the system is running, and (to a lesser extent) to guard against failure of the acting process itself, not failure of the system as a whole.
Well, this might work for young unmarried developers with little work experience, but try to convince your wife that you are doing everything you can to get a job while programming open source projects for free.
What's so hard about telling her that you're working on adding some additional experience to your CV? As long as you are actually doing that (and not posting to slashdot), I would expect most wives to understand, at least mine would.
Indeed--my wife did `understand':
After a month of being unemployed, I had to start hacking on Open Source projects--part of it was that being productive like that helped to pick my morale back up off the floor, but the other (and bigger, more pragmatic) reason was that hacking on Open Source projects kept my skillset from rotting.
She did question me a couple of times, but I just told her that.
In the end, I got top-notch job--which I wouldn't have been able to get if I'd shown up to the interview and been unable to answer engineering-questions due to having not done any engineering for 3+ months. Being able to show my prospective employer which projects were using my code (and show the code) was a definite plus, too--I'm certain that it helped me win-out over the other candidates.
Unfortunately, there are only 2 canned keybinding-sets shipped with GTK: "Default" (which is Windowsy, these days) and "Emacs" (which makes things behave like the GNU tools (Emacs, Bash, etc.) traditionally behave.
These canned sets are defined in/usr/share/themes/*/gtk-2.0-key/gtkrc; I think that, if you make a new directory (where the "*" is), and create your own `gtkrc' file under it, the that name becomes valid in GConf. You may be able to find a Mac-alike gtkrc file that someone else has already produced, now that you know what to look for.
Though, actually: are the keybindings similar enough that you could just remap the "command" key so that it acts like "control"? Or does Mac OS use the "control" strokes for something else?
When I buy a brick, it's my brick. When I buy a car, it's my car. When I buy an iPhone, it's... my iphone. So I can do whatever I want with it.
Suddenly, a person who purchased an iphone can't do whatever s/he wants with it, and that is apparently NORMAL, while the brick owner can do whatever s/he wants with it, and that's also normal (duh).
Presumably, you are free to install OpenMoko or some other OS onto/your/ iPhone hardware.
If you execute chroot() and then a seteuid(uid) where uid>0, then you prevent a hole/bug in your program from being exploited in a way that will allow file access/execution outside the chroot. That *is* a security advantage.
According to the page about breaking out of chroot linked from the discussion, the seteuid() is not effective because the process still has a real UID of 0 and can call seteuid(0) at any time.
So don't use seteuid--use setuid (without the "e"); this is what the article that you just cited indicates:
If this [setuid to non-root] is done there should be no way to gain root privilages unless an attacker uses something within the chroot() jail to gain those privilages.
I can't really see a reason against this except cost and regulation. Honestly I wouldn't care.
Really it's not for you (unless you want it) it's for the people who are morally opposed to certain things. You can do it for Kosher food, why not have it for cloned food (and possibly genetically enhanced food). The fact is there's going to be people against cloned food....
Are there really people who are morally opposed to kosher food?
Isn't the idea of requiring labels for cloned food more akin to requiring labels for all non-kosher food?
Isn't kosher-labelling more of an allowance than a requirement?
Aren't all of the non-clone producers (who pushing this legislation) already free to label their products as "Clone-Free", just like `Cage-Free' eggs or `Organic' foods?
They could even form a `Clone-Free Consortium' to buy a "Clone-Free" trademark, just like the kosher-authorities have done.
Requiring clone-labelling seems much more like requiring that...:
* all non-kosher foods must be labelled
* all non-organic foods must be labelled as such
* cage-raised eggs must be labelled as such
I store all of my e-mail in my inbox, including my sent mail--for one thing, this makes it/vastly/ easier to go back through threads.
Unread/new messages are... not marked as Read.
By default, all I see are the unread messages, and whatever I've flagged.
Alternate views (e.g.: everything that I sent, everything that another particular person/entity sent, all messages about some topic, messages within some date-range...) are defined primarily by searches.
"Resist these urges. On tests, including the SATs and APs, fancy laptops and super-calculators are not allowed, and proctors will confiscate what they aren't sure about."
No kidding--anything they're not sure about. Anything. Not just fancy electronics, but/anything/:
When I took the SAT-2 in 1998, I went in with a sliderule.
They wouldn't let me use it.
"No, I don't see `sliderule' on the list or approved calculators....", they said.
I should have asked if `fingers' were on the list--some people can say that their hands are deadly weapons, and I'm not one of them..., but my hands are effective calculating-devices: they contain two five-bit general-purpose registers that can be treated as a unified 10-bit register, with bit-shift operators and a full complement of boolean operators.
"Ever tried to do basic drawing in The Gimp? Like, say, drawing a circle? Ask any Gimp developer why this is such a bitch and they'll tell you something like: The Gimp is an image manipulation program, not a drawing program, go use Inkscape or something if you want to draw circles."
I've never found drawing circles in Gimp to be `a bitch'--I do it the same way I do it in real life: I make a mask, and stroke along the edge of the mask with my pencil/pen/airbrush. It's possible to make more complex masks easily as intersections and/or unions of base shapes and splines (in the real world, some of those spline-masks would be French curves). So, quite the opposite from being `a bitch', it's `my bitch'--it does what I want without my having to deal with `computerisms' that don't scale.
Sometimes I do the double-mask thing (recommended at the URL that you gave), but that's something different--not what I thought of when you said "drawing circles". But, if you'd said "make a filled annulus", I'd do something similar to that, then I'd do something similar to that in real life; I'd do something similar in Gimp, but not for solid-colour fills.
When I've worked in Photoshop, I did things the same way.
Maybe you're not speaking from a graphic artist's point of view?
When I've make minemeat pie, the recipes that I've used included meat.
When I bought prepackaged mincemeat at the store, the package has clearly listed "beef" in the ingredients. `Nonesuch' brand certainly does contain meat.
- 1600 DPI, twice that of conventional high performance sensors
How many pixels (of pointer travel on screen) per point of movement? I mean, if I have a screen of 1600x1200, moving the mouse an inch would send the pointer across the whole screen width with 1:1 mapping. Pretty much unplayable, aim at a 16x16px icon with that, you need 0.01 inch hand movement precision. Of course the points get downsampled, and as result the extra resolution - wasted. Useless. [...] - Frame rate over 6400 frames per second (5.8 megapixels per second)
That means about 100 frames per screen display frame. Your character can turn 3 times around before the screen updates to show it, thanks to this mouse. More frames per second than the screen can show is useless.
But the mouse needs to be able to tell how far, and in which direction, it's moved between screen-updates. It tracks motion (delta-position), not position (except for the Wacom-style mice, I guess--they're different).
Imagine watching video of pans across the mid-section of a large carpet, or parking lot, or anything without distinguishable `landmarks'. If you have to report how the camera has moved at 1-minute intervals, what kind of frame-rate do you need to see in order to perform your task? At one frame per minute, you might be able to tell that the camera has moved, but you won't be able to tell where to or how. What if the camera moves to the left and back to the right within a single sample-period? Would you even be able to tell that it had moved, let alone how?
The `resolution' issue is presumably similar. And the sensor in the mouse is presumably not an inch square and not able to capture all of those pixels simultaneously. At a bare minimum, I imagine that you need at least a 2x2 array of pixels to see motion, and that resolution-requirement goes up as the sample-rate goes down. The more of the image you can see of your panning carpet (or whatever), and the more detail you can see, the easier it is to see where you're going.
Try navigating through a dense, unfamiliar forest, with a set of blur-inducing eye-glasses. And open your eyes only briefly, once per minute. And then try to escape the Blair Witch;)
"Oh doesn't read well", "full of dupes", "lame comments", "I'm tired of listening to your babble".
Basically man, "beep beeeeep" and get a life. You grab any opportunity to tell the world that you are "cool" because you read Slashdot but all you do is complain while reading porn in a dark corner when nibidy is looking and if you don't like the Slashdot commentary....
I am a software-developer and there is a demand for my software-writing services, so I do charge for that, just as RMS is charging for his autograph-writing services.
The question is whether RMS will charge for the ability to make photocopies or other derivative works of his autographs.
And (continuing the equivocatuion) yes, charging $500 for writing and revising software, or $200 for just writing software, is better than charging millions of dollars to watch an airplane--I usually end up charging thousands of dollars just to write the initial version or a piece of software, while watching planes usually costs nothing (unless I want to record then, in which case I usually have to buy my own media).
$30 per hour of software-development is more than $5k per week! Quite a living, eh?
"Take, for instance, any form of acquired behavior.
Wrong. Thank you for playing. Features aquired while alive (not counting genetic mutations in the sperm/egg or genome of an asexually reproducing organism) do NOT get passed on."
Well, not genetically, but they might go out the same way that they came in. Vocation, for example--I'm an X, as my father before me who taught me the way as did his before him. Or religious traditions, for another: I know plenty of people who were taught to celebrate christmast by their parents, and those parents by their parents, and so on. Religious tradition is another one: are christmast-lights transfered genetically? No. Do they transfer? Yes. Consider also, as, as he wrote, just about any other form of acquired behaviour.
"One more option. Windows boxes with Open Source software running. If there are 8 things the box must do, and one of them is available only for Windows..., use free software wherever it fits.... Proprietary here:... AVG (no Free antiviruses)"
Well his coworkers that are physically in new york use the roads to get to work. If the roads weren't there he would have no job.
Right..., and I wouldn't have my job if I hadn't worn my Brooks Brothers shirt to the interview or signed the contract with my Sheaffer pen, so... I should be paying for the utilities used by the people who made them, in both New York and Iowa?
Slashdot Mirror
The funny thing is... that his joke was actually hacker-funny, because he got the syntax right: Forth is *postfix*, not *backward infix*--the verbs go at the end (top) of the sentence (stack).
I could say that like this:
postfix Forth is infix backward Forth is not and sentence end verbs goto therefor
... but:
it (would; be; completely) {
lost(on, you);
}
Your joke, on the other hand... matches your UID.
Actually those "Ext4" data corruption issues that set the Internet all ablaze (including Slashdot) were mainly due to KDE 4 not handling metadata correctly. In the end, it wasn't an Ext4 issue. However, feel free to spread FUD.
Bullshit. The error was very generic and not limited to KDE
You're right: it's a common application programming error.
At what frequency does a bug cease to be a bug?
I wish I'd known that I could promote myself up out of the rank of `amateur' just by pointing out that there were plenty of other people who weren't any more skilled than I was--that would have obviated all of these years of study and hard work.
Only if he can get it out, use it, and then put it away again all within the time that it takes a satellite to complete an orbit, which takes a lot less time than you think it does (hint: look for the word "minutes").
I read an anecdote somewhere that went something like this:
This is not appropriate for desktop systems -- desktop systems have to be robust against all kinds of stupid situations, like sudden power losses, users hitting the reset button because an application is hanging, and so forth.
First off, PC hardware is not robust against sudden power-losses: it's literally possible for a write to be `half-done' inside the HDD, and no amount of higher-level `protection' can do anything about that.
Secondly, the atomicity of rename() (or any other operation) isn't contagious: rename() does an unlink() and a link() in `one operation' (and that's the only amount of atomicity rename() claims--it's even specifically not atomic in that it's possible to see the original link and the new link at the same time), but that doesn't have any impact--or even any relation to other `nearby' operations.
write() and rename() aren't even operating on the same object--write() is operating on the file, and rename() is operating on the directory (or directories that do/will contain links to the file). I don't quite get the `metadata' arguments, because `filenames' aren't `file metadata'--they're directory data, which is what allows you to have any number of links to the same file from any number of directories. File-metadata are things like timestamps, ownership, permissions....
Lastly, having said all that: the reason that we go through the write-close-rename sequence is to prevent a race-condition while the system is running, and (to a lesser extent) to guard against failure of the acting process itself, not failure of the system as a whole.
Will the flames still be retarded?
On Slashdot? Yes, surely there are plenty of retarded flamers....
Indeed--my wife did `understand':
After a month of being unemployed, I had to start hacking on Open Source projects--part of it was that being productive like that helped to pick my morale back up off the floor, but the other (and bigger, more pragmatic) reason was that hacking on Open Source projects kept my skillset from rotting.
She did question me a couple of times, but I just told her that.
In the end, I got top-notch job--which I wouldn't have been able to get if I'd shown up to the interview and been unable to answer engineering-questions due to having not done any engineering for 3+ months. Being able to show my prospective employer which projects were using my code (and show the code) was a definite plus, too--I'm certain that it helped me win-out over the other candidates.
Just like the insurance-companies do now? What's the difference?
Or you can edit your .gtkrc file.
Unfortunately, there are only 2 canned keybinding-sets shipped with GTK: "Default" (which is Windowsy, these days) and "Emacs" (which makes things behave like the GNU tools (Emacs, Bash, etc.) traditionally behave.
These canned sets are defined in /usr/share/themes/*/gtk-2.0-key/gtkrc; I think that, if you make a new directory (where the "*" is), and create your own `gtkrc' file under it, the that name becomes valid in GConf. You may be able to find a Mac-alike gtkrc file that someone else has already produced, now that you know what to look for.
Though, actually: are the keybindings similar enough that you could just remap the "command" key so that it acts like "control"? Or does Mac OS use the "control" strokes for something else?
Presumably, you are free to install OpenMoko or some other OS onto /your/ iPhone hardware.
So don't use seteuid--use setuid (without the "e"); this is what the article that you just cited indicates:
Are there really people who are morally opposed to kosher food?
Isn't the idea of requiring labels for cloned food more akin to requiring labels for all non-kosher food?
Isn't kosher-labelling more of an allowance than a requirement?
Aren't all of the non-clone producers (who pushing this legislation) already free to label their products as "Clone-Free", just like `Cage-Free' eggs or `Organic' foods?
They could even form a `Clone-Free Consortium' to buy a "Clone-Free" trademark, just like the kosher-authorities have done.
Requiring clone-labelling seems much more like requiring that...:
* all non-kosher foods must be labelled
* all non-organic foods must be labelled as such
* cage-raised eggs must be labelled as such
I store all of my e-mail in my inbox, including my sent mail--for one thing, this makes it /vastly/ easier to go back through threads.
Unread/new messages are... not marked as Read.
By default, all I see are the unread messages, and whatever I've flagged.
Alternate views (e.g.: everything that I sent, everything that another particular person/entity sent, all messages about some topic, messages within some date-range...) are defined primarily by searches.
"Resist these urges. On tests, including the SATs and APs, fancy laptops and super-calculators are not allowed, and proctors will confiscate what they aren't sure about."
/anything/:
No kidding--anything they're not sure about. Anything. Not just fancy electronics, but
When I took the SAT-2 in 1998, I went in with a sliderule.
They wouldn't let me use it.
"No, I don't see `sliderule' on the list or approved calculators....", they said.
I should have asked if `fingers' were on the list--some people can say that their hands are deadly weapons, and I'm not one of them..., but my hands are effective calculating-devices: they contain two five-bit general-purpose registers that can be treated as a unified 10-bit register, with bit-shift operators and a full complement of boolean operators.
"Ever tried to do basic drawing in The Gimp? Like, say, drawing a circle? Ask any Gimp developer why this is such a bitch and they'll tell you something like: The Gimp is an image manipulation program, not a drawing program, go use Inkscape or something if you want to draw circles."
I've never found drawing circles in Gimp to be `a bitch'--I do it the same way I do it in real life: I make a mask, and stroke along the edge of the mask with my pencil/pen/airbrush. It's possible to make more complex masks easily as intersections and/or unions of base shapes and splines (in the real world, some of those spline-masks would be French curves). So, quite the opposite from being `a bitch', it's `my bitch'--it does what I want without my having to deal with `computerisms' that don't scale.
Sometimes I do the double-mask thing (recommended at the URL that you gave), but that's something different--not what I thought of when you said "drawing circles". But, if you'd said "make a filled annulus", I'd do something similar to that, then I'd do something similar to that in real life; I'd do something similar in Gimp, but not for solid-colour fills.
When I've worked in Photoshop, I did things the same way.
Maybe you're not speaking from a graphic artist's point of view?
When I've make minemeat pie, the recipes that I've used included meat.
When I bought prepackaged mincemeat at the store, the package has clearly listed "beef" in the ingredients. `Nonesuch' brand certainly does contain meat.
What you talkin' 'bout, man?
But the mouse needs to be able to tell how far, and in which direction, it's moved between screen-updates. It tracks motion (delta-position), not position (except for the Wacom-style mice, I guess--they're different).
Imagine watching video of pans across the mid-section of a large carpet, or parking lot, or anything without distinguishable `landmarks'. If you have to report how the camera has moved at 1-minute intervals, what kind of frame-rate do you need to see in order to perform your task? At one frame per minute, you might be able to tell that the camera has moved, but you won't be able to tell where to or how.
What if the camera moves to the left and back to the right within a single sample-period? Would you even be able to tell that it had moved, let alone how?
The `resolution' issue is presumably similar. And the sensor in the mouse is presumably not an inch square and not able to capture all of those pixels simultaneously. At a bare minimum, I imagine that you need at least a 2x2 array of pixels to see motion, and that resolution-requirement goes up as the sample-rate goes down. The more of the image you can see of your panning carpet (or whatever), and the more detail you can see, the easier it is to see where you're going.
Try navigating through a dense, unfamiliar forest, with a set of blur-inducing eye-glasses. And open your eyes only briefly, once per minute. And then try to escape the Blair Witch
"Oh doesn't read well", "full of dupes", "lame comments", "I'm tired of listening to your babble".
Basically man, "beep beeeeep" and get a life. You grab any opportunity to tell the world that you are "cool" because you read Slashdot but all you do is complain while reading porn in a dark corner when nibidy is looking and if you don't like the Slashdot commentary....
DON'T BLOODY READ IT!
"IE7 allows for search engine options just like firefox, and IE7 doesn't use its biggest competitor by default, again just like Firefox."
I don't follow--how is a search-engine a `competitor' to a web-browser?
"It doesn't support copying yet--I think that they're working on that."
"Well, for all of the features that it does have, I can live with that little problem for a while."
And then `a while' becomes `forever'.
Eventually, acceptance sets in.
I am a software-developer and there is a demand for my software-writing services, so I do charge for that, just as RMS is charging for his autograph-writing services.
The question is whether RMS will charge for the ability to make photocopies or other derivative works of his autographs.
And (continuing the equivocatuion) yes, charging $500 for writing and
revising software, or $200 for just writing software, is better than charging millions of dollars to watch an airplane--I usually end up charging thousands of dollars just to write the initial version or a piece of software, while watching planes usually costs nothing (unless I want to record then, in which case I usually have to buy my own media).
$30 per hour of software-development is more than $5k per week! Quite a living, eh?
"Take, for instance, any form of acquired behavior.
Wrong. Thank you for playing. Features aquired while alive (not counting genetic mutations in the sperm/egg or genome of an asexually reproducing organism) do NOT get passed on."
Well, not genetically, but they might go out the same way that they came in. Vocation, for example--I'm an X, as my father before me who taught me the way as did his before him. Or religious traditions, for another: I know plenty of people who were taught to celebrate christmast by their parents, and those parents by their parents, and so on. Religious tradition is another one: are christmast-lights transfered genetically? No. Do they transfer? Yes. Consider also, as, as he wrote, just about any other form of acquired behaviour.
The Common Lisp "image" idea doesn't bother you?
"One more option. Windows boxes with Open Source software running. If there are 8 things the box must do, and one of them is available only for Windows..., use free software wherever it fits.... Proprietary here: ... AVG (no Free antiviruses)"
What about ClamAV?
Right..., and I wouldn't have my job if I hadn't worn my Brooks Brothers shirt to the interview or signed the contract with my Sheaffer pen, so... I should be paying for the utilities used by the people who made them, in both New York and Iowa?