whats wrong with .net debugging

Posted By: twoelfer | Apr 6th, 2004 @ 8:17 AM
page 1 of 1
Comments: 3 | Views: 2821
#Apr 6th, 2004 @ 8:17 AM
twoelfer
twoelfer

just stumbled accross this, so while i'm at it, here's another one...


the one single issue i hate a lot is that there's no way to make the clr 'run' 'as if' it was running on a different platform.

that is, i'd like to debug my app running inside the vs debugger ontop of xp pro while the underlying version of the net clr and framework thinks i'm running it ontop of w98.

here's why: not everything in the .net framework is supported on all platforms. and yes, this stuff is documented, but unsupported stuff still kinda turns up in the code... which means: you need to code on xp, and still test against installations on 98 (at least) nt and w2k as well.

it would be _much_ better to just tell the environment to behave as if one was running ontop of an 'older' platform.

WM_MY0.02$
thomas woelfer

#Apr 7th, 2004 @ 12:04 AM
bshankle
bshankle

1 word: Virtual PC (or is that two, nonetheless, it's all the rage for testing on multiple platforms)

#Apr 7th, 2004 @ 3:18 AM
twoelfer
twoelfer
i do have a copy of virtual pc for testing purposes. however, when debugging a .net app, simply telling the runtime to behave like it was running ontop of another platform would be a good start for testing _and_ make live a lot easier.

i haven't looked at rotor but i guess the runtime simply checks the windows version at startup time once and throws notSupportedExceptions when unsupported functions are called later.

thus, it shouldn't be so hard to add a configuration where the version info concerning the os isn't checked at runtime but at startup (or patched in later).

this is all about ease of use and an elegant way to test, not about checking out existing ways.

WM_THX
thomas woelfer
page 1 of 1
Comments: 3 | Views: 2821