|
Tiny Bug Child posted:raymond chen makes me appreciate apple that much more because they don't humor all that dumb backwards compatibility stuff apple macs can't even use apps from 1982. fuckin bullshit *has osx* *runs recent software*
|
# ? Jun 17, 2015 20:41 |
|
|
# ? May 30, 2024 13:11 |
|
here we see the fight for dangerous settings and the phrasing of warnings: http://blogs.msdn.com/b/oldnewthing/archive/2013/04/16/10411267.aspx
|
# ? Jun 18, 2015 16:37 |
|
also, here is my favorite article to send people when they insisit on using window 95 style themes because they think aero is killing their processor: http://blogs.msdn.com/b/oldnewthing/archive/2013/03/27/10405554.aspx
|
# ? Jun 18, 2015 16:39 |
|
echinopsis posted:apple macs can't even use apps from 1982. fuckin bullshit actually they can thanks to the power of bsd
|
# ? Jun 18, 2015 16:42 |
|
A Wheezy Steampunk posted:also, here is my favorite article to send people when they insisit on using window 95 style themes because they think aero is killing their processor: http://blogs.msdn.com/b/oldnewthing/archive/2013/03/27/10405554.aspx lol @ the comments doing exactly what the article is making fun of
|
# ? Jun 18, 2015 17:13 |
|
A Wheezy Steampunk posted:also, here is my favorite article to send people when they insisit on using window 95 style themes because they think aero is killing their processor: http://blogs.msdn.com/b/oldnewthing/archive/2013/03/27/10405554.aspx i had a really old thinkpad with really crappy integrated video and turning off a lot of visual effects in windows 7 and visual studio 2010 really did speed a lot of things up noticeably but yeah if you have even a half decent discrete video or even a modern integrated one then i guess he's right
|
# ? Jun 18, 2015 21:02 |
|
A Wheezy Steampunk posted:also, here is my favorite article to send people when they insisit on using window 95 style themes because they think aero is killing their processor: http://blogs.msdn.com/b/oldnewthing/archive/2013/03/27/10405554.aspx it's cause win7 glass theme is loving atrocious not because it makes it any faster
|
# ? Jun 19, 2015 01:41 |
|
~Coxy posted:it's cause win7 glass theme is loving atrocious not because it makes it any faster Actually it's great. Sorry about your failintosh fail aids.
|
# ? Jun 19, 2015 13:49 |
|
maniacdevnull posted:Actually it's great. Sorry about your failintosh fail aids.
|
# ? Jun 19, 2015 13:51 |
|
here's a technical article that i don't understand: If you look at the disassembly of functions inside Windows DLLs, you'll find that they begin with the seemingly pointless instruction MOV EDI, EDI. This instruction copies a register to itself and updates no flags; it is completely meaningless. So why is it there?
|
# ? Jun 19, 2015 14:00 |
|
Nifty
|
# ? Jun 19, 2015 14:12 |
|
A Wheezy Steampunk posted:here's a technical article that i don't understand: it's to allow on-the-fly patching of DLL functions. It uses MOV EDI,EDI because that instruction does nothing, but can later be replaced with a jump instruction pointing to the patched version of the code. It has to be MOV EDI,EDI (or any other two byte do-nothing instruction) because a jump instruction needs two bytes. if you'd used two single byte NOPs then there's a change that a task could execute the first NOP, then the patch is installed, then the task tries to execute the second NOP, but it's now been replaced with the second byte of a jump instruction, so the task would fetch a byte from the middle of an instruction and try to execute it as as a valid instruction leading to unexpected results.
|
# ? Jun 19, 2015 14:19 |
|
Sweevo posted:it's to allow on-the-fly patching of DLL functions. i mean i read the rest of the article too but i don't know what edi or nop means or why the second byte of a jump instruction would lead to unexpected results
|
# ? Jun 19, 2015 14:21 |
|
its assembly voodoo
|
# ? Jun 19, 2015 14:23 |
|
A Wheezy Steampunk posted:i mean i read the rest of the article too but i don't know what edi or nop means or why the second byte of a jump instruction would lead to unexpected results jump instructions need an additional byte to create the address you are jumping to, which is why they use dummy instructions with similar instruction structure to "reserve space" for a jump instruction to be added later
|
# ? Jun 19, 2015 14:28 |
|
A Wheezy Steampunk posted:i mean i read the rest of the article too but i don't know what edi or nop means or why the second byte of a jump instruction would lead to unexpected results edi is a general purpose register nop is a no-op or an operation that does nothing, which is actually really useful an instruction is one or more bytes, and if it's more than one byte, the second byte of an instruction could also possibly be interpreted as its own instruction because they're just unstructured bytes an ancient anti-disassembly trick is a two-byte instruction that jumps to halfway into itself, throwing off some disassemblers: you can read more about it in this definitely-legally-reproduced page from a no starch press book i found on a chinese university's website: http://staff.ustc.edu.cn/~bjhua/courses/security/2014/readings/anti-disas.pdf to make a long story short: there's a useless instruction at the beginning of functions because it's better than having two useless instructions
|
# ? Jun 19, 2015 14:39 |
|
Cocoa Crispies posted:edi is a general purpose register cool
|
# ? Jun 19, 2015 14:42 |
|
A Wheezy Steampunk posted:i mean i read the rest of the article too but i don't know what edi or nop means or why the second byte of a jump instruction would lead to unexpected results x86 instructions have the annoying property of not all of them being the same length. this also means that the same sequence of bytes can be totally different instructions depending on whether the stuff immediately before it is a full instruction, or just part of one. (it also means disassembling obfuscated x86 is obnoxiously Fun, since you need to work out exactly where jumps *in* to the bit of code you're looking at happen) the reason they need to use a single two-byte no-op rather than two individual nops is that if you used two separate instructions, you might happen to overwrite the patch point right as the CPU was halfway between them. Then it would take the second half of the jump instruction (the part which overlaps with the second nop), try and treat it as the start of an instruction, and then basically all bets are off from that point.
|
# ? Jun 19, 2015 14:43 |
|
hey did anyone explain the article yet i'd like to show off that i know assembly
|
# ? Jun 19, 2015 15:31 |
|
knowing assembly is nothing to be proud of, it's a sign of poor life choices
|
# ? Jun 19, 2015 16:00 |
|
monday morning means time for some posts by raymond about windows pinball What one Windows XP feature am I most proud of? Why was Pinball removed from Windows Vista?
|
# ? Jun 22, 2015 16:36 |
|
One weird trick to reduce service ticket volume: http://blogs.msdn.com/b/oldnewthing/archive/2015/06/16/10621828.aspx
|
# ? Jun 22, 2015 17:08 |
|
this isn't technically a raymond post but it's close enough in spirit: http://blogs.technet.com/b/exchange/archive/2004/04/08/109626.aspx "So those 15,000,000 email messages collectively consumed 195,000,000,000 bytes of bandwidth. Yes, 195 gigabytes of bandwidth bouncing around between the email servers."
|
# ? Jun 23, 2015 14:49 |
|
Some people have 40gb NICs, less than a minute of traffic.
|
# ? Jun 23, 2015 17:08 |
|
MrMoo posted:Some people have 40gb NICs, less than a minute of traffic. note the date on that article
|
# ? Jun 23, 2015 17:16 |
|
Heresiarch posted:note the date on that article and the story was old then, according to https://en.wikipedia.org/wiki/Email_storm it happened in 1997
|
# ? Jun 23, 2015 17:20 |
|
the chen posted:Look at this way: Suppose you could somehow get this algorithm to run a quintillion times faster, so it finishes in under a year. Your output file is going to be 2¹²⁸ × 16 = 2¹³² bytes in size. That's around 10²⁷ terabytes. One terabyte of SSD storage weighs around 100 grams. The mass of the earth is 10²⁴ kilograms. Therefore, before you run this program, you will need to acquire 100 earth-sized planets and convert them all to SSDs.
|
# ? Jun 23, 2015 21:45 |
|
i think we can all related to people who ask for every possible piece of information so they can sort it out later: http://blogs.msdn.com/b/oldnewthing/archive/2012/02/02/10263027.aspx
|
# ? Jun 24, 2015 17:47 |
|
Why do operating system files still adhere to the old 8.3 naming convention?
|
# ? Jun 24, 2015 23:36 |
|
oh nice like that tv show
|
# ? Jun 25, 2015 07:11 |
|
ohhhh so his name is actually ll cool ☺
|
# ? Jun 25, 2015 07:45 |
|
Breakfast All Day posted:ohhhh so his name is actually ll cool ☺ lol
|
# ? Jun 25, 2015 10:50 |
|
Breakfast All Day posted:ohhhh so his name is actually ll cool ☺
|
# ? Jun 25, 2015 12:22 |
|
Breakfast All Day posted:ohhhh so his name is actually ll cool ☺ 😈
|
# ? Jun 25, 2015 13:09 |
|
Breakfast All Day posted:ohhhh so his name is actually ll cool ☺
|
# ? Jun 25, 2015 13:24 |
|
A Wheezy Steampunk posted:Video drivers will do anything to outdo their competition. Everybody knows that they cheat benchmarks, for example. I remember one driver that ran the DirectX "3D Tunnel" demonstration program extremely fast, demonstrating how totally awesome their video card is. Except that if you renamed TUNNEL.EXE to FUNNEL.EXE, it ran slow again. A time honored tradition that Samsung still adheres to in their phones
|
# ? Jun 25, 2015 15:08 |
|
|
# ? May 30, 2024 13:11 |
|
So the DirectDraw folks changed the way they queried for driver capabilities. One of the developers went into his boss's office, took a network card, extracted the MAC address, and then smashed the card with a hammer.
|
# ? Jun 27, 2015 19:00 |