»
S
I
D
E
B
A
R
«
AusHac2010 Day 2 progress
July 17th, 2010 by axman6
Day 2 of AusHac2010 is coming to an end, and we’ve made a lot of progress:
Bernie Pope has been making great progress with a new MPI binding for Haskell
Ben Lippmeier, Erik de Castro Lopo and Ben Sinclair have been busily hacking on DDC, with 13 commits today alone
Stephen Blackheath has been working on some code using the Accelerate library that rasterises triangles for use in a commercial computer game.
Hamish Mackenzie, Jens Petersen and Matthew Sellers have been working on better Yi integration for Leksah, working on using Yi’s current configuration file, and improving “launch experience”, focusing on eliminating the requirement of creating an initial workspace file.
Lang Hames has been using his experience with LLVM from working at Apple as an intern to improve various low level problems in LLVM. His work should help resolve some of the problems the LLVM backend to GHC has, but should also be very beneficial to many other LLVM users. While doing this, he’s written a very nice tool that illustrates register liveness, with further work focusing on colouring the HTML output to show register pressure. The LLVM guys seem quite excited about this work, which is great.
Mark Wotton and Sohum Banerjea have been trying to extend Hubris, the Haskell-Ruby bridge, to work with polymorphic functions. Their heads are quite sore from all the head banging. Raphael Speyer has been working on an install script to make installation much easier for users.. but only if you use Ubuntu so far.
Ivan Miljenovic has been prematurely optimising his containers library, before finalising the API. This library is designed to let library writers leave the choice of which container data structure to output to the library consumer as well as making it easier to change which data structure you want to use in your code, with minimal code change. See his blog post for more details. **********
Trevor McDonell has been working on the CUDA backend to Accelerate, adding support for efficient nested tuple types, and other bug fixes. Sean Lee has been helping out with testing of this code, along with Manuel Chakravarty.
With one more full day to go, I think we’ll be getting a lot of awesome work done tomorrowQ!

Day 2 of AusHac2010 is coming to an end, and we’ve made a lot of progress:

Bernie Pope has been making great progress with a new MPI binding for Haskell

Ben Lippmeier, Erik de Castro Lopo and Ben Sinclair have been busily hacking on DDC, with 13 commits today alone

Stephen Blackheath has been working on some code using the Accelerate library that rasterises triangles for use in a commercial computer game.

Hamish Mackenzie, Jens Petersen and Matthew Sellers have been working on better Yi integration for Leksah, working on using Yi’s current configuration file, and improving “launch experience”, focusing on eliminating the requirement of creating an initial workspace file.

Lang Hames has been using his experience with LLVM from working at Apple as an intern to improve various low level problems in LLVM. His work should help resolve some of the problems the LLVM backend to GHC has, but should also be very beneficial to many other LLVM users. While doing this, he’s written a very nice tool that illustrates register liveness, with further work focusing on colouring the HTML output to show register pressure. The LLVM guys seem quite excited about this work, which is great.

Mark Wotton and Sohum Banerjea have been trying to extend Hubris, the Haskell-Ruby bridge, to work with polymorphic functions. Their heads are quite sore from all the head banging. Raphael Speyer has been working on an install script to make installation much easier for users.. but only if you use Ubuntu so far.

Ivan Miljenovic has been prematurely optimising his containers library, before finalising the API. This library is designed to let library writers leave the choice of which container data structure to output to the library consumer as well as making it easier to change which data structure you want to use in your code, with minimal code change. See his blog post for more details.

Trevor McDonell has been working on the CUDA backend to Accelerate, adding support for efficient nested tuple types, and other bug fixes. Sean Lee has been helping out with testing of this code, along with Manuel Chakravarty.

With one more full day to go, I think we’ll be getting a lot of awesome work done tomorrow!


Leave a Reply

»  Substance: WordPress   »  Style: Ahren Ahimsa
© Alex Mason (Axman6) 2009