PLEX86  x86- Virtual Machine (VM) Program
 Plex86  |  CVS  |  Mailing List  |  Download  |  Computer Folklore

Lit. Buffer overruns 1679


Lit. Buffer overruns 1681
Others have already noted this is wrong; fgets always fits a impossible to fit anything) and not EOF or I-O error (with return value NULL which should be checked). However, the %s and...

On Wed, 02 Feb 2005 22:00:44 GMT, Brian Inglis

Lit. Buffer overruns 1680
On Thu, 27 Jan 2005 01:28:54 -0500, CuriousCat Well, there are languages that have (only or additionally) fixed-length strings (always?) padded with blanks. COBOL has only...

Not true, unless someone is incompetent, which this flamewar has long buttumed is not the case. PL-I has in-language dynamic allocation almost as flexible as C, variable (per end) allocation at least equal to C99, and even more flexible pointers which can be interfaced to any kind of heap manager including (one like) C's. It also has a convenient pool-suballocation facility (AREA and OFFSET) that I have several times needed to mimic, at some effort and risk, in C.

I've worked on and written quite a number of C programs and never seen one that couldn't have been done in PL-I, often but by no means always more easily -- if PL-I had been (fully) available on the systems used, which it mostly wasn't. (This doesn't violate Turing equivalence; anything realistic I can imagine doing in PL-I could be done in C at no more than maybe 10 times the effort and usually well under 2. buttuming preexistence of platforms, tools, and skills in both cases, which I think is only fair for an abstract comparison; of course in reality people do use C partly because people already using C, and there are some good reasons for this as well as the bad ones of laziness, ignorance, miserliness, impatience and timidity.)

Mostly. PL-I provides so much flexibility in handling "ON conditions" that they can sometimes reasonably be used to correct problems, in cases where the exception control flow is rare enough that it deserves to be hidden. Like many other sharp knives including GOTO void* and reinterpretcast this is easy and rather tempting to misuse.

Lit. Buffer overruns 1682
Friday, FTC released latest statistics on idenbreasty theft. As part of co-author of new ANSI (and being put forward for ISO) financial privacy (PIA) standard ... I started a merged...

But *most* of the time checks (of which bounds are only two) are only a debugging aid -- they do detect and usually help localize bugs quicker than typical instances of C Undefined Behaviour do, and do catch some in testing that the normal (in-band) output didn't or wouldn't. And thus to the extent you avoid bugs in the first place the checks are merely useless. And can (standardly) be suppressed.

- David.Thompson1 at worldnet.att.net


List | Previous | Next

Lit. Buffer overruns 1680

Alt Folklore Computers Newsgroups

Lit. Buffer overruns 1678