Hi Joke_Bloke, You dreamed: << You cannot compare Mono on Linux and .NET developed by Microsoft for Windows. Mono was created with very limited specs and no help from Microsoft as a multiplatform version of .NET. >> I've posted my code over and over again, it's quite small, Ingo Molnar provided no source code for his claims, ...How bogus were his claims anyways, I have to wonder. Imagine if four guys ( five, if we could count Ingo ), generated a report like this: 2,008 Simultaneous threads acheived. 57 Seconds per 100,000 simultaneous threads spawned/completed. I'd find that interesting precisely because it was written in different languages, using diffent libraries and run on different kernels. You don't like that comparison ? What exactly do you want to see and why ? You mentioned only " Linux 2.4 kernel threads ", But Bailo is running 2.6 not 2.4 . Are you saying he needs to update his libaries or something ? Do you have these libaries ? Why don't you just run Ray's code then ? ( Even Bailo has done that... the only one to have run another's code ) You quoted Ingo: << ...on the default x86 kernel, it needed roughly 500 MB of RAM to do this test with the IRQ-stacks patch applied. >> So ? We don't have Ingo's code, do we ? He was using dual P4s with a custom kernel, I'm using off-the-shelf software ( Windows XP Home Edition ), My 200-buck box has only 256 Megs of RAM, the L2 cache is only 128 K, I'm also using about 20 megs of system-RAM for Direct Draw 7, ...Think about it. Re: Me not doing whatever it is that you want, You concluded: << If that wasn't obvious to you, as it seems from your post, you are much more stupid than I suspected or you are deliberately ignoring that fact to support your drivel. >> What " drivel " am I trying to support ? How to disintegrate the 2.6 kernel on demand ? What " brilliance " are you trying to support ? The notion that I can't read the cobwebs in your head ?