Profile cover photo
Profile photo
Lars Bergstrom
Lars's posts

Post has attachment
Born May 17, 7lbs 11oz
Eric Torsten Bergstrom
7 Photos - View album

Post has attachment
We just posted our Mozilla Research internships (targeted at PhD students). While we have a number of suggested projects on the page, a substantial fraction of our interns propose and do their own thing. We're pretty open, as long as it's got something to do with the web as a platform...

Post has attachment
We just opened up a research engineer position to do compiler development on Rust here at Mozilla Research. Rust is a new systems programming language that feels like C++ meets Concurrent ML meets Cyclone. It is publicly available and we are using it internally to build a new parallel web browser, Servo.

Please contact me if you have any questions about this position, Rust, or even just Mozilla Research, if you're curious.

Post has attachment
I don't usually link to comics, but this one about nifty graphs got me. Since we switched our papers from using ploticus to a fancy pipeline of R generating LaTeX tikZ files, every paper has had at least one review that asked how we made those pretty charts...

Post has attachment
Wow, I didn't see that coming! Facebook using Haskell and having Simon Marlow... I'm wracking my brain for quotes about "too much power in the wrong hands."

Reflow-enhanced optimizations are now in and will soon be on by default in Manticore. The short summary is: we can now efficiently perform both inlining and copy propagation (replacing a variable with the original function name) for functions that have free variables where safe, even in fairly non-trivial scenarios. The speedup across our existing parallel benchmark suite on our 48~core machine is an average of 5%, depending on the benchmark, with a couple of outliers at 0% where no opportunities in inner loops were identified.

Many thanks to +Matthew Might+David Van Horn, and +David MacQueen, who all entertained quite a few correctness-related questions! It'll still be a few months before this makes it in to a paper submission and then a few more months after that to see light of day, if I'm lucky enough to have it accepted somewhere, but I wanted to follow up on the 6 month old comments I made about some ideas that we'd had about how to make it work. The graph-related trickery required for good performance was fairly subtle (basically, needed something much faster than Floyd-Warshall's O(n^3) for computing graph reachability tests), but once we had that and ironed out a few corner cases on correctness, the rest fell into place quickly.

This morning on my way to the conference center, I saw the usual string of bike commuters in suits. But, this time, one of them was drinking beer from a can. I couldn't decide if I was more impressed by the fact that it was a little before 9 AM or just the coordination required by such a feat during the rush-hour bike commute (which is Serious Business here in Copenhagen).

Post has attachment
More than 100 whisky labels, you say? I think I've found a third option to the "crime tour" vs. "canal tour" for my post-conference activities tomorrow...

Things I didn't get warned about for Copenhagen: if you order an Americano, > 50% of the time they'll just throw powdered/instant coffee into hot water rather than pulling an espresso shot and adding water. Now, if it were cheap, I wouldn't care, but 28kr ~= $4.25. Looks like espressos for me from here out...

Other than that, though, the city's been fantastic so far. Tomorrow starts the HOPE/ICFP/ML Workshop madness, though, so the rest of the city that I see in daylight will be through windows, if I'm lucky!

I love seeing Evernote push out new releases all the time, but I really wish they'd invest a little time and implement three-way, plain text merge during synchronization instead of just copying the wrong-base note into a "conflicts" folder and leaving me to do a baseless two-way text merge on a 3" cell phone screen.
Wait while more posts are being loaded