Future of KSSL

[Jernej Kos]

>    KSSL does work fine.  It's apparently just missing a feature that
> you want.

No it does not work fine at all. If we for example take a look at the read 
method we can clearly see something like:

                        if (maxIters-- > 0) {
                                ::usleep(20000); // 20ms sleep
                                goto read_again;
                        }

This loop causes huge speed decreases since it just wastes cycles when it 
really should not block (that is the point of async connections). The same 
could be said about the connect and write methods, where socket polling 
should be done instead of headless looping.

Please note that I have already posted that to the list a couple of months ago 
and there were no replies.

Regards,
Jernej Kos.