| |
| |
| |
|
Page: 1 2 3 4 5 6 7 8 9 10 11
Comments:
<0> asriel: its for an embedded system
<1> windows... embedded...
<1> o_O
<0> its a on special controller that runs our application. User doesnt/must not access windows
<0> running CE 4.1
<1> i'd question your choice of os, but whatever.
<2> How do I include GLUT library in MinGW compiler ?
<3> why do you question it NineVolt ?
<3> what do you know about Windows CE?
<4> one of the params to SystemParametersInfo will do you
<4> SPI_SETMOU***XXX
<4> iirc
<4> dunno if it works on CE
<0> ill try it
<5> hmm
<5> the dr watson that ships on xp seems to barf on managed processes, not sure why
<5> i've configured the managed debugger to use dr watson, it pops up, saying it is generating an error report (configured it for full dump), the log file it generates is empty except for the dr. watson logo, and i get no dump
<5> i guess i'll write my own postmortem debugger
<5> not terribly difficult
<5> all i need is a full dump
<5> perhaps if i turn off logging maybe dr. watson will give me a dump
<5> it takes it like 15 seconds to go from a cancel button to an OK button, and still doesn't generate anything useful heh
<5> i also tried cdb but that appears to simply hang
<6> can i change visual studio 6.0 c++ compiler to the latest vc++ compiler? (free download)
<0> i think you can get vs2005 express version, but im not sure
<6> is it free?
<0> i believe so
<6> wow, with IDE ?
<0> the express version only or something like that
<0> yes but it does not offer everything and i have no clue whats missing
<5> dontdo, yes with a full IDE
<5> MFC/ATL missing is the most prominent
<0> peter, is "express" the right name?
<6> thanks guys.. thats a great news
<5> rog, yes
<0> thx
<6> oops, MFC is missing?
<5> express does not come with MFC or ATL, the Platform SDK (separate free download, easy to setup), or PGO (profile guided optimizations - a new link time code generation optimization technique; otherwise, the compiler is fully optimizing)
<7> vc2005 express is free
<5> it does some with full support for C++/CLI, a WinForms designer, a standard c++ implementation, and a pretty much feature complete IDE
<6> okay, thanks
<7> and a GREAT debugger
<8> greetings all
<7> howdies
<5> arg
<5> so close
<5> trying to make heads or tails of the SoS debugger extension
<5> enables CLR stack walking and heap dumping
<5> got a fulldump from dr. watson for a managed process, and i'm trying to determine the site of the exception
<5> as the stack was already unwound before the CLR threw a native (2nd chance) exception to crash the process, i have to rely on the exception object on the GC heap
<5> which i've found
<5> and can print the exception message
<5> hmm
<5> StackTrace is of type string
<5> yet SoS shows it as unsigned byte[]
<5> let me see if i can cast it to wchar_t*
<5> no good, i need to know the layout of Array and the offset to the actual data
<5> though, i'm looking at the field, not the property, perhaps the CallStack method walks the stack and creates a string, and then stores it
<5> time to look at the IL
<5> yep, that's exactly what it does
<5> System.Environment::GetStackTrace(System.Exception)
<5> which is marked ***embly
<5> fun stuff
<9> Welcome to #peterhulive
<5> pfft
<5> not like you clowns are using the channel
<9> true
<9> noones triaged your bug yet. too slow, MS!
<0> anyone ever messed with spi_setmouse ?
<5> RoG: i'm not familiar with it
<10> snif sniff
<11> crack?
<10> thought it was *****
<11> fishy smell?
<10> what would i know
<12> 22:23:06 up 459 days, 8:13, 4 users, load average: 0.00, 0.00, 0.00
<12> After that has been said, I suggest we drop that topic know.
<12> s/k//
<5> sigh
<5> SoS 2 has the !pe command
<5> i need that in 1.1
<5> those bastards
<5> jeez, they added a whole bunch of useful commands in 2
<5> and 2 doesn't not appear to work with .NET 1.0/1.1 ***emblies
<11> SoS = ?
<0> sos = . . . _ _ _ . . .
<5> son of strike, a debugger extension to debug managed code in a dump file
<9> so, shouldn't SoS 2 be Grandson of Strike?
<9> on that thought...
<5> apparently, the sos that ships with the Debugger Tools for Windows has a different !dumpobj that formats the _stackTrace, all i needed
<13> real men debug with printfs
<0> i cant even debug with printf :(
<5> pfft
<14> ahh.. don't remind me of that. /me spent the whole day by doing that, and failing to run gdb
<0> lol... have to debug on a com port
<13> that's because you don't believe in evolution RoG
<5> at least i know the layout of Array now, 4 byte size, followed by data
<8> refresh my memory
<15> hello, how can i get rand to give me negative values?
<8> I dont plan on inserting data frequently so I want to use which
<8> vector or list
<8> Im pretty sure its a vector
<13> http://www.joannejacobs.com/mtarchives/016331.html
<5> vector is slow on inserts that increase its size or anything that isn't inserted on the end (deque goes on further by allowing fast inserts at the front). vector has fast random access. list has constant time inserts, but no random access. so it's up to you.
<13> map
<13> the best of all worlds
<0> anyone knows how to force windows to reload its registry programatically?
<0> ive tried broadcasting message WM_SETTINGCHANGE but that does not seem to work
<11> uh
<11> you want the whole windows registry to be reloaded?
<0> well i want one setting to be relaoded (mouse sensitivity) but I think reloading everything would do the trick
<11> reload explorer.exe is a dirty way
<0> looks like its my last hope :\
<0> which is totally horrible.
<13> that can't be right.
<0> I'm really ****ed lol
<11> why? reloading explorer isn't so bad
<0> could be on my application
<0> i dont know
<11> what's your exact call to broadcast WM_SETTINGCHANGE?
<11> 'cause i remember using that a while ago and it worked
<0> SendMessageTimeout(HWND_BROADCAST,WM_SETTINGCHANGE,NULL,NULL,SMTO_ABORTIFHUNG|SMTO_NOTIMEOUTIFNOTHUNG,1000,&dwret);
<0> i tried that
<5> knowing the Array layout was the key, the !IP2MD command will translate an EIP value to a method descriptor, simple as pie
<11> RoG: and the function returns without errors?
<0> lemme check
<0> i get a timeout
<0> dwRet = 0, lastError = 0
<0> prolly a window couldnt handle it on time, which i believe doesnt matter
<0> any quick trick on how to reload explorer, im gonna try to see if that ****s everything
<0> i just kill it and restart it basically?
<11> yes
<11> that should work
<11> gotta go, hope everything works out
<7> samm, did you ever get an answer?
<16> I get
<16> ../../ircproxy/src/Expression.h:137: error: using-declaration for non-member at cl*** scope
<16> ../../ircproxy/src/Expression.h:137: error: expected ; before < token
<16> cl*** Bool {
<16> std::set<boost::shared_ptr<ExpressionEventServer> > M_event_servers; // <-- here
<16> any ideas?
<16> My guess would be that std::set is not defined, just declared.
<16> but... I do #include <set> :/
<16> maybe it's a problem with the PCH
<7> I'd guess that you can't have a set of boost::shared_ptr
<16> ah, indeed... #include <set> was missing in pch-source.h
<12> I think shared_ptr supports operator<, so set should work.
Return to
#c++
or
Go to some related
logs:
#mirc
#linux
#AllNiteCafe
London street is associated with Sweeney Todd
#linux
#networking
#chatzone
#chatzone
#linux
IRC pingout
|
|