Comments for Not A Number http://notanumber.net Programming, Theory, and Math Mon, 11 Jan 2010 17:27:31 +0000 http://wordpress.org/?v=2.9.1 hourly 1 Comment on Underhanded C: The Leaky Redaction by Simon Strandgaard http://notanumber.net/archives/54/underhanded-c-the-leaky-redaction/comment-page-1#comment-2907 Simon Strandgaard Mon, 11 Jan 2010 17:27:31 +0000 http://notanumber.net/?p=54#comment-2907 Pure evil. Fantastic well put together piece of code that looks fairly innocent. Thank you for opening my eyes for this kind of dangerous code. Keep up the good (or should I say evil) work! Pure evil. Fantastic well put together piece of code that looks fairly innocent. Thank you for opening my eyes for this kind of dangerous code.

Keep up the good (or should I say evil) work!

]]> Comment on Underhanded C: The Leaky Redaction by Craig http://notanumber.net/archives/54/underhanded-c-the-leaky-redaction/comment-page-1#comment-2906 Craig Sat, 02 Jan 2010 03:47:01 +0000 http://notanumber.net/?p=54#comment-2906 Wow. Wow.

]]> Comment on Underhanded C: The Leaky Redaction by Lalit Mishra http://notanumber.net/archives/54/underhanded-c-the-leaky-redaction/comment-page-1#comment-2905 Lalit Mishra Fri, 01 Jan 2010 15:10:03 +0000 http://notanumber.net/?p=54#comment-2905 Clever. Clever.

]]> Comment on Underhanded C: The Leaky Redaction by Aaron http://notanumber.net/archives/54/underhanded-c-the-leaky-redaction/comment-page-1#comment-2904 Aaron Thu, 31 Dec 2009 20:29:13 +0000 http://notanumber.net/?p=54#comment-2904 My code review would start with "Where is BUFSIZE declared?". I'm assuming there are header files being #include'd which have been omitted, or perhaps you are defining it on the command line? My code review would start with “Where is BUFSIZE declared?”. I’m assuming there are header files being #include’d which have been omitted, or perhaps you are defining it on the command line?

]]> Comment on Underhanded C: The Leaky Redaction by Thomas Beaucourt http://notanumber.net/archives/54/underhanded-c-the-leaky-redaction/comment-page-1#comment-2903 Thomas Beaucourt Wed, 30 Dec 2009 23:39:58 +0000 http://notanumber.net/?p=54#comment-2903 Congratulations, beautiful idea ! Plausible deniability is very high indeed. Congratulations, beautiful idea ! Plausible deniability is very high indeed.

]]> Comment on Underhanded C: The Leaky Redaction by ThatsGobbles http://notanumber.net/archives/54/underhanded-c-the-leaky-redaction/comment-page-1#comment-2902 ThatsGobbles Wed, 30 Dec 2009 21:12:19 +0000 http://notanumber.net/?p=54#comment-2902 Very clever. I really dig the fact that, like you said, one of the most common types of images to block out are black-and-white text images, and this program works *best* on those types. Very clever. I really dig the fact that, like you said, one of the most common types of images to block out are black-and-white text images, and this program works *best* on those types.

]]> Comment on Underhanded C: The Leaky Redaction by Anonymous http://notanumber.net/archives/54/underhanded-c-the-leaky-redaction/comment-page-1#comment-2901 Anonymous Wed, 30 Dec 2009 20:09:22 +0000 http://notanumber.net/?p=54#comment-2901 Hmm. If I were Spook, I would still be suspicious and/or worried as your answer lacks any _real_ justification as to why the obvious option of using scanf() (or rather fscanf()) would be wrong here. Your explanation seems to rely on the (correct) behavior of fgets(), which is different from fscanf(). It seems you assume fscanf() might work incorrectly (by not assigning all of the matching data within a line - you seem to imply fscanf() uses fgets() with a limited buffer internally, which may or may not be true) and because of this you implement the logic which uses fgets() directly. Such behavior would be a bug. I am not aware of such a (known) bug in any fscanf() implementation. Are you? Hmm. If I were Spook, I would still be suspicious and/or worried as your answer lacks any _real_ justification as to why the obvious option of using scanf() (or rather fscanf()) would be wrong here. Your explanation seems to rely on the (correct) behavior of fgets(), which is different from fscanf().

It seems you assume fscanf() might work incorrectly (by not assigning all of the matching data within a line – you seem to imply fscanf() uses fgets() with a limited buffer internally, which may or may not be true) and because of this you implement the logic which uses fgets() directly.

Such behavior would be a bug. I am not aware of such a (known) bug in any fscanf() implementation. Are you?

]]> Comment on Underhanded C: The Leaky Redaction by MadKat http://notanumber.net/archives/54/underhanded-c-the-leaky-redaction/comment-page-1#comment-2900 MadKat Wed, 30 Dec 2009 19:07:16 +0000 http://notanumber.net/?p=54#comment-2900 That is truly a beautiful solution. Thank you for the extended description. That is truly a beautiful solution. Thank you for the extended description.

]]>