While it might still take some time for the video to be released, here are my slides:
http://nimrod-code.org/talk01/slides.html
And here are some elaborations:
http://nimrod-code.org/talk01/corrections.html
I'd like to take the opportunity and thank the hosts for the Strange Loop event! It was great fun and very inspiring.
Very nice, thanks for the slides.
Just a little remark: the slide 8 triggered one of my 'annoyance point', it is written: 'it provides a realtime GC which supports max pause times of 1-2 miliseconds', the thing is: one number for realtime GC is meaningless, to define a realtime GC, you need to provide two numbers: we support a max pause time of 1-2 miliseconds every 100 miliseconds(for example).
Without the second number, every GC is a realtime GC, just break the 20ms pause time into ten 2ms pass scheduled back to back, easy!
No, not necessarily "every 100 miliseconds" but you need to give some kind of progress garantee like "supports X MB/s of allocations" or "doesn't simply grow the heap beyond limits".
The context here is a main game loop obviously where you run the GC on every loop iteration (at 60 FPS you have 16ms per iteration). Also I say explicitly in my talk that those times should be taken with a grain of salt until it has been used in production for such a setting.
It seems that the videos are available on the InfoQ website but only if you have attendeed the conference:
http://www.infoq.com/conferences/strangeloop2013/
"This is a restricted presentation that can only be viewed by Strange Loop 2013 attendees! The public release of this presentation will be sometime in the next 6 months - date to be determined."
Shiiiit.