[Mono-dev] Architectural decisions behind mod_mono

[Tinco Andringa]

> Unfortunately Microsoft has a long history of confusion on this topic.
> I remember back when IIS5 was being developed I implored the IIS team
> make it possible to write filters as kernel modules.  I insisted that
> the extra kernel transitions needed to invoke my legacy encryption
> filter was going to kill performance.  Their reaction was 'No way! You
> might blue-screen the machine!'.  They never did understand that,
> whether it was in user space or the kernel, if there was a fatal bug in
> my filter the machine was no more useful than a doorstop.

Maybe but these kind of performance issues can usually be solved
through horizontal scaling, something which applications like Facebook
should have no problem with. Bluescreens on shared hosting boxes
however would be very bad, and the shared hosting market is very
important for webdevelopment frameworks to attract new developers.

That and maybe debugging issues, it's tough to debug something if it
keeps bluescreening on you. If a process dies because of a
segmentation fault, you just restart the process so it can resume work
while you figure out what happened. This is significantly less
downtime, which means especially much for hosting with uptime
guarantees.