h a l f b a k e r yFree set of rusty screwdrivers if you order now.
add, search, annotate, link, view, overview, recent, by name, random
news, help, about, links, report a problem
browse anonymously,
or get an account
and write.
register,
|
|
|
| |
I'm afraid it doesn't work like that. If Windows
knew what was crashing well enough to start an
autokill it would know enough to stop it
immediately.
What happens, for example,
is Internet Explorer is running an instance of Flash
that captures memory at the same time another
instance of flash is trying to work with the same
bit of memory and Flash thinks the memory is
underallocated so it keeps asking Windows for
more and more and more and Windows thinks this
is a perfectly normal request and keeps giving it
more and more, even paging as much of itself into
memory as possible.
There's no way for
Windows to tell the difference between this and
the user running My Massive Program which asks
for a gig and a half of memory because it actually
needs it. |
|
| |
Could it not just shut the computer down permanently upon
installation? |
|
| |
The problem with the idea is that it requires windows to do *something successfully. I don't see how it would work if ctrl-alt-delete and "end task" both fail.
To be fair, ever since I upgraded from XP I haven't had to reset the computer due to a program locking up like I used to. |
|
| |
You'd think that programs would be able to tell the operating system approximately how much memory they would roughly require up front, and if they go massively beyond that, then Windows would get wise enough to figure out that something wasn't right. |
|
| |
Fewer problems with Win7 64bit Ultimate than previous versions but I suspect there's already an autodestruct sequence embedded somewhere in every iteration of Windows. |
|
| |
<insert one of many rants on tap> |
|
| |
Meh - it's about time consumer computing started running all its processes in a unix-style way where, apart from kernal and system operations, user-processes are easily kill-offable should the need arise. |
|
| |
It gets more complicated when providing an interface to a physical device (disk, memory, screen, printer etc) but providing suitably sandboxed and timeoutable APIs for these exist, no user-generated call should be able to trash the system anyway. |
|
| |
How long before Microsoft decides to restructure their OS so that it runs atop some kind of Unix foundation? Surely it can only be a matter of time? |
|
| |
They're desperate not to piss off the people who bought their earlier shit products, so they keep building legacy issues into their software, like a reliance upon BIOS, which is responsible for the sheer length of time it takes for PCs to boot. |
|
| |
Fix that problem and they'll boot in seconds, like a digital TV. |
|
| |
// How long before Microsoft decides to restructure their OS // |
|
| |
Long enough so they go bankrupt, hopefully. |
|
| |
//You'd think that programs would be able to tell the operating system approximately how much memory they would roughly require up front// |
|
| |
Some programs might, other programs don't really know how much they need until they need it. Browsers, for example, can require tens of megabytes for the program and a simple HTML page or two. But you might go on YouTube and before you know it you have two gigabytes of videos open. |
|
| |
You could have some sort of sanity limit though. Say any program that burns through more than half the available virtual memory gets halted and the user gets to decide whether to close it or not. |
|
| |
"Module 'Iterate-Trajectory' in Application 'Calculate_Critical_ Re-Entry_Angle' has requested excessive amounts of memory and CPU time, interfering with Internet Explorer's ability to log and report your browsing habits to Microsoft. Windows 13 has therefore dumped this process and deleted the executable and source files. Have a nice day. Click OK to resume your descent from orbit." |
|
| |
:O I'm reminded of the saying "if it weren't for error messages, computers wouldn't have problems". |
|
| |
Might programs then be divided into the
'knows' and the 'know-nots'? |
|
| |