When you are dealing with expensive hardware, or a cluster
If you're dealing with a cluster, chances are you can make improvements significantly larger than 5%.
Re:The performance of compiled code
on
A Review of GCC 4.0
·
· Score: 4, Interesting
You're obviously a small box user. Have you ever worked in the real world where huge batch runs can take weeks?
Yes.
You think companies should splash out another million or too on new hardware, just because you use a pissy little machine?
I think that companies should re-evaluate their "need" for an extra 5% performance. Here's an idea -- if you need something 10 minutes faster, why not start the process 10 minutes sooner?
5% just gets lost in the noise. You beef up your system, making it 5% faster... And then some retard in production makes a mistake and sets you back six weeks.
Re:The performance of compiled code
on
A Review of GCC 4.0
·
· Score: 3, Insightful
At some point you've got the best algorithm, you've profiled, you've hand-optimised, you've got the fastest hardware you can afford....and you *still* need that last 5%.
That's when you spend 10 hours tweaking compilers settings...
If you really, positively need an extra 5% performance, you might as well just buy a computer that's 5% faster.
Okay, I must be a total fucking geek, because at first I thought you had abbreviated "Dijkstra" as "Dijk" and left the 'i' out for some reason. Then I wondered why the hell Dijkstra would waste his time compiling code by hand...
The performance of compiled code
on
A Review of GCC 4.0
·
· Score: 5, Informative
This has always bugged me.
Some people spend 10 hours tweaking compiler settings and optimizations to get an extra 5% performance from their code.
Other people spend 2 hours selecting the proper algorithm in the first place and get an extra 500% performance from their code.
To semi-quote The Matrix: One of these endeavors... is intelligent. And one of them is not.
A stack overflow ought to overwrite unallocated space, not earlier stack frames and return addresses. It's totally insane.
Not really. You assume that all buffer overflows overflow in the "upward" direction. It's just as easy, in C, to code a loop that progresses backward through memory. I've had many reasons and occassions to do it. Simply making the stack grow upward instead of downward won't solve the underlying basic issue, which is that without proper bounds checking, the program can overwrite memory it's not supposed to.
Besides, it's incredibly convenient for the stack to grow downward. Program code and data starts at the bottom of virtual memory, and the stack starts at the top. You just map in new page frames as necessary. If the stack grew the other direction, it would either have to be limited in size, or you'd have to shift it in memory if it grew too large. Shifting it is practically impossible, since you'd have to find all program pointers into the stack and update them all to reflect the new stack. Gad, I don't even want to think about it.
But if you use bitmaps, then it will look shit when zoomed in or printed at high resolution
Not if the image is at a high enough resolution. The ones we generate are either 300 or 600 DPI depending on the user's preference. A normal person can barely tell the difference between 300 and 600 DPI for a printed page of text. Hardly anyone can tell between 600 and 1200. If your laser printer can only go up to 600 (which is typical), then a 600 DPI image is all you need.
And who cares what the image looks like when someone zooms in? Having it re-rendered in high quality is unrealistic -- when it prints, it prints to a certain resolution and that's that. If a user wants to zoom way in on an uppercase "A" and he sees pixels, who cares?
You don't seem to understand how people actually use this stuff. "Not seeing pixelation when I zoom way the fuck in" just isn't a real requirement. Having pixel-perfect rendition and printing on all PDF viewers *is* a real requirement.
Can most home users who bought a computer with Microsoft Works Suite (including Microsoft Word) pre-installed justify the price of an Acrobat license?
Probably not, but that wasn't what I was getting at. I was just trying to differentiate between PDF as a format, and Adobe's particular toolchain for dealing with it.
Each inode stores a pair of four-character codes: one "file type" code, which could be mapped to Internet Media Types, and one "creator" code, which determines what application to start when the user wants to Open the file.
I grok the creator code, because it gives you flexibility. But again, seems like the file type could get out of sync with what's actually in the file.
Out of a bit of curiosity I saved it under OpenOffice as.scx (or whatever OO's normal extension is, dont remember),.doc, and.pdf The.doc was about 4 times the size of the.scx and the.pdf was about 40 times that.
Yes, it's possible to generate truly massive PDF. This is usually due to the stupidity of the application generating the PDF, however. PDF gives you a million ways to shoot yourself in the foot. But I'd blame the tools, not the format.
BTW last version of Word I used didnt have a print to PDF unless you had some sort of addon.
I'm talking about Acrobat. If you install Acrobat, every application becomes capable of printing to PDF.
OpenOffice does though, never use it because I boycott Acrobat when I can.
Boycotting the format isn't really the same as boycotting Acrobat...
besides, file associations are another piece of braindead in-band signalling
Magic numbers are also in-band signalling. There's nothing inherently wrong with in-band signalling. Extensions suck because somebody can change it and break the file. A magic number, OTOH, stays with the file no matter what you name it or where you store it.
And are you suggesting some kind of out-of-band signalling instead? Like, say, the file inode stores its MIME type? Sounds cool at first, but again, it can get out of sync with what's actually in the file.
Extensions suck, I agree, but don't knock in-band signalling.
I wish it wasn't so slow. It even taunts you with how much crap it is doing.
Go to the plug_ins\ directory in your Acrobat or Reader installation. Move everything in that directory to another location (or just delete it, but I wouldn't). Now, none of that shit loads when Acrobat starts.
Some pdf writers basically use pdf as a bitmap format (wastefull) while others encode the text as text and let the pdf viewer render the fonts..
The reason many PDF generators use images is because they want the end product to look exactly how they want. Using fonts and text leaves the end representation up to the viewer on the other end. If the viewer is crap, the result will be crap. On the other hand, even a terrible viewer can usually render an image properly, so the document is nearly guaranteed to look correct on the end user's system.
I work in this field, and we only just recently started putting vector text in our PDF output for just these reasons. And that was only out of complete necessity. In fact, the text itself is still represented as an image, but the file also contains "invisible" text in vector format which allows the end user to search for, and copy and paste, text.
As for wastefulness... A typical black and white, letter sized page will compress to about 50 kilobytes using CCITTFaxEncode stream compression. 50 kilobytes is a small price to pay to guarantee that the page is represented correctly on the other side.
Unforunately Adobe and gang have convinced too many people that PDF is a good distribution format for reading documentation online supposedly because someone "might" want to print it.
That's not why people like it. They like it because they can whip something up in Word (or whatever else) and just hit "Print to PDF." They like it because it's brain dead simple.
And anyway, you are one of the very few enlightened ones. Most people do want to print.
I want to scream whenever I have to use a PDF that I can't fill in on the computer. Because of the retarded limitations they place on the viewer, the form *must* be set up for document input when it's created.
You do realize that the document creator has to pay big bucks to Adobe to make a PDF form that can be filled out by the free Reader, don't you? Creating forms which can be filled out by the Reader (not Acrobat, that's different) has an approximately $65,000 startup cost.
It's meant to encourage corporations to buy and use Acrobat internally instead of just the free Reader.
The exception is the US Government, which has a license from Adobe to make and distribute PDF forms for free.
PDF has DRM, you know. You can restrict user saving, printing, copying, editing, and even high-level rendering.
I'd hestitate to call it "Digital Rights Management," because it utterly fails to manage anything. Non-Adobe viewers just ignore the advisory limitations, and the PDF file itself can easily be cracked to remove them altogether.
Encrypted PDF is a bit harder, but brute force is still feasible because most users pick stupid passwords.
I fucking hate reading a page at a time. I'm on a fucking computer, not a book damit!
Hint: PDF files are meant to be printed. Printers typically use pages.
If they'd just fucking use HTML I'd get so much more done.
You are complaining that the document is being distributed in the wrong format. How is that the fault of PDF?
We already have the replacement, it's called HTML and it does just fine.
No. Because unfortunately, people still need to print things on paper for various reasons. HTML gives you practically zero control over page formatting (CSS gets you about 1% of the way there). PDF, on the other hand, is designed for page formatting.
You are complaining that when you use a hammer as a screwdriver, things don't work too well. No shit.
I agree that too many people distribute information in PDF format when they could have just used HTML. But many publishers are very print-centric. Remember that.
why would incoming application level traffic be interpreted by the winmodem firmware?
Not incoming traffic. Outgoing traffic (the ping is echoed, along with the payload). +++ is the Hayes modem escape sequence which places the modem in a command mode. ATH0 then hangs up the modem.
This wasn't supposed to be possible because there's supposed to be a three second wait before and after the escape sequence. But shoddy modems (as someone else explained, those that did not license the patent) didn't require the delay.
Re:"How to improve your phishing attack"
on
Phishing for Credit
·
· Score: 1
If no one has built it, there is no need to disarm it. In Soviet Russia, bomb disarms you.
Just because some hypothetical "we" doesn't want to build one doesn't mean the enemy won't. Similarly in this case, do you think that if academics don't do the research, the Bad Guys will never figure it out?
Next time you break into a bank and get caught while inside the vault just tell the cops you were testing the security system without the banks knowledge, but intended to give a full report later on.
Just remember to get authorization from said police before you do it. That's what these folks did.
Re:"How to improve your phishing attack"
on
Phishing for Credit
·
· Score: 1
Er... this is sorta like doing research on how to make a better bomb, buddy. This is not socially responsible computer science research, is it?
So if we shut our eyes and stick our fingers in our ears, Everything Will Be Okay?
I'd be more interested in determining out how to create a social networking site ( like whatever this "facebook" thing is ) that _can't_ be exploited in such a manner.
How would anybody have known about this exploit if nobody has studied it? To use your bomb analogy, how could we ever figure out how to disarm a thermonuclear weapon if nobody has ever built one?
Re:The More Attention This Gets, The Better
on
Phishing for Credit
·
· Score: 4, Informative
I think it's pretty clear to everyone that these students didn't follow proper procedure for research studies. When I did human experimental research, I had to have my research proposal approved by the Institutional Review Board at my college.
That's precisely what they did. The whole thing was authorized from top to bottom. They even got the okay from campus IT to "abuse" the computer systems for their purposes. Try RTFA sometime.
Slashdot Mirror
If you're dealing with a cluster, chances are you can make improvements significantly larger than 5%.
Yes.
You think companies should splash out another million or too on new hardware, just because you use a pissy little machine?
I think that companies should re-evaluate their "need" for an extra 5% performance. Here's an idea -- if you need something 10 minutes faster, why not start the process 10 minutes sooner?
5% just gets lost in the noise. You beef up your system, making it 5% faster... And then some retard in production makes a mistake and sets you back six weeks.
If you really, positively need an extra 5% performance, you might as well just buy a computer that's 5% faster.
Okay, I must be a total fucking geek, because at first I thought you had abbreviated "Dijkstra" as "Dijk" and left the 'i' out for some reason. Then I wondered why the hell Dijkstra would waste his time compiling code by hand...
Some people spend 10 hours tweaking compiler settings and optimizations to get an extra 5% performance from their code.
Other people spend 2 hours selecting the proper algorithm in the first place and get an extra 500% performance from their code.
To semi-quote The Matrix: One of these endeavors... is intelligent. And one of them is not.
Not really. You assume that all buffer overflows overflow in the "upward" direction. It's just as easy, in C, to code a loop that progresses backward through memory. I've had many reasons and occassions to do it. Simply making the stack grow upward instead of downward won't solve the underlying basic issue, which is that without proper bounds checking, the program can overwrite memory it's not supposed to.
Besides, it's incredibly convenient for the stack to grow downward. Program code and data starts at the bottom of virtual memory, and the stack starts at the top. You just map in new page frames as necessary. If the stack grew the other direction, it would either have to be limited in size, or you'd have to shift it in memory if it grew too large. Shifting it is practically impossible, since you'd have to find all program pointers into the stack and update them all to reflect the new stack. Gad, I don't even want to think about it.
Not if the image is at a high enough resolution. The ones we generate are either 300 or 600 DPI depending on the user's preference. A normal person can barely tell the difference between 300 and 600 DPI for a printed page of text. Hardly anyone can tell between 600 and 1200. If your laser printer can only go up to 600 (which is typical), then a 600 DPI image is all you need.
And who cares what the image looks like when someone zooms in? Having it re-rendered in high quality is unrealistic -- when it prints, it prints to a certain resolution and that's that. If a user wants to zoom way in on an uppercase "A" and he sees pixels, who cares?
You don't seem to understand how people actually use this stuff. "Not seeing pixelation when I zoom way the fuck in" just isn't a real requirement. Having pixel-perfect rendition and printing on all PDF viewers *is* a real requirement.
Probably not, but that wasn't what I was getting at. I was just trying to differentiate between PDF as a format, and Adobe's particular toolchain for dealing with it.
I grok the creator code, because it gives you flexibility. But again, seems like the file type could get out of sync with what's actually in the file.
Yes, it's possible to generate truly massive PDF. This is usually due to the stupidity of the application generating the PDF, however. PDF gives you a million ways to shoot yourself in the foot. But I'd blame the tools, not the format.
BTW last version of Word I used didnt have a print to PDF unless you had some sort of addon.
I'm talking about Acrobat. If you install Acrobat, every application becomes capable of printing to PDF.
OpenOffice does though, never use it because I boycott Acrobat when I can.
Boycotting the format isn't really the same as boycotting Acrobat...
Magic numbers are also in-band signalling. There's nothing inherently wrong with in-band signalling. Extensions suck because somebody can change it and break the file. A magic number, OTOH, stays with the file no matter what you name it or where you store it.
And are you suggesting some kind of out-of-band signalling instead? Like, say, the file inode stores its MIME type? Sounds cool at first, but again, it can get out of sync with what's actually in the file.
Extensions suck, I agree, but don't knock in-band signalling.
Go to the plug_ins\ directory in your Acrobat or Reader installation. Move everything in that directory to another location (or just delete it, but I wouldn't). Now, none of that shit loads when Acrobat starts.
The reason many PDF generators use images is because they want the end product to look exactly how they want. Using fonts and text leaves the end representation up to the viewer on the other end. If the viewer is crap, the result will be crap. On the other hand, even a terrible viewer can usually render an image properly, so the document is nearly guaranteed to look correct on the end user's system.
I work in this field, and we only just recently started putting vector text in our PDF output for just these reasons. And that was only out of complete necessity. In fact, the text itself is still represented as an image, but the file also contains "invisible" text in vector format which allows the end user to search for, and copy and paste, text.
As for wastefulness... A typical black and white, letter sized page will compress to about 50 kilobytes using CCITTFaxEncode stream compression. 50 kilobytes is a small price to pay to guarantee that the page is represented correctly on the other side.
That's not why people like it. They like it because they can whip something up in Word (or whatever else) and just hit "Print to PDF." They like it because it's brain dead simple.
And anyway, you are one of the very few enlightened ones. Most people do want to print.
You do realize that the document creator has to pay big bucks to Adobe to make a PDF form that can be filled out by the free Reader, don't you? Creating forms which can be filled out by the Reader (not Acrobat, that's different) has an approximately $65,000 startup cost.
It's meant to encourage corporations to buy and use Acrobat internally instead of just the free Reader.
The exception is the US Government, which has a license from Adobe to make and distribute PDF forms for free.
I'd hestitate to call it "Digital Rights Management," because it utterly fails to manage anything. Non-Adobe viewers just ignore the advisory limitations, and the PDF file itself can easily be cracked to remove them altogether.
Encrypted PDF is a bit harder, but brute force is still feasible because most users pick stupid passwords.
Hint: PDF files are meant to be printed. Printers typically use pages.
If they'd just fucking use HTML I'd get so much more done.
You are complaining that the document is being distributed in the wrong format. How is that the fault of PDF?
We already have the replacement, it's called HTML and it does just fine.
No. Because unfortunately, people still need to print things on paper for various reasons. HTML gives you practically zero control over page formatting (CSS gets you about 1% of the way there). PDF, on the other hand, is designed for page formatting.
You are complaining that when you use a hammer as a screwdriver, things don't work too well. No shit.
I agree that too many people distribute information in PDF format when they could have just used HTML. But many publishers are very print-centric. Remember that.
Not incoming traffic. Outgoing traffic (the ping is echoed, along with the payload). +++ is the Hayes modem escape sequence which places the modem in a command mode. ATH0 then hangs up the modem.
This wasn't supposed to be possible because there's supposed to be a three second wait before and after the escape sequence. But shoddy modems (as someone else explained, those that did not license the patent) didn't require the delay.
It's not ass-talking, it's old school :-)
I see you've fallen for this before, huh?
Just because some hypothetical "we" doesn't want to build one doesn't mean the enemy won't. Similarly in this case, do you think that if academics don't do the research, the Bad Guys will never figure it out?
Publicly? Can you please give the URL to the page where they posted the names of the hapless victims? I'd like to see that.
Just remember to get authorization from said police before you do it. That's what these folks did.
So if we shut our eyes and stick our fingers in our ears, Everything Will Be Okay?
I'd be more interested in determining out how to create a social networking site ( like whatever this "facebook" thing is ) that _can't_ be exploited in such a manner.
How would anybody have known about this exploit if nobody has studied it? To use your bomb analogy, how could we ever figure out how to disarm a thermonuclear weapon if nobody has ever built one?
That's precisely what they did. The whole thing was authorized from top to bottom. They even got the okay from campus IT to "abuse" the computer systems for their purposes. Try RTFA sometime.
It was cleared by the research ethics body.