Comments for Michael Cook's Place http://www.foobarsoft.com Bits of genius in a sea of mediocrity Sat, 21 Jan 2012 13:56:01 +0000 hourly 1 http://wordpress.org/?v=3.3.1 Comment on Second Best Clojure Bot, Postmortem by thobel http://www.foobarsoft.com/2011/12/second-best-clojure-bot-postmortem/#comment-307 thobel Sat, 21 Jan 2012 13:56:01 +0000 http://www.foobarsoft.com/?p=139#comment-307 Thanks for your very informative postmortem. It helped me a lot improving my programming style in Clojure. PS: without your kind debugging/visualizer tips my place in the top 500 would not be possible :) Thanks for your very informative postmortem. It helped me a lot improving my programming style in Clojure.
PS: without your kind debugging/visualizer tips my place in the top 500 would not be possible :)

]]> Comment on Trying PyPy with ReplicatorG by Michael Cook http://www.foobarsoft.com/2011/10/trying-pypy-with-replicatorg/#comment-13 Michael Cook Thu, 20 Oct 2011 01:02:12 +0000 http://www.foobarsoft.com/?p=12#comment-13 Since a few days after I did the full comparison I've been using PyPy exclusively. As I type this, I'm about 2 hours into a 3 hour print (part of that great differential that's been featured on Thingiverse). There is no external difference, since it's only tiny changes in the internal support structure. It's been working great, and the speed is much nicer. Since a few days after I did the full comparison I’ve been using PyPy exclusively. As I type this, I’m about 2 hours into a 3 hour print (part of that great differential that’s been featured on Thingiverse). There is no external difference, since it’s only tiny changes in the internal support structure. It’s been working great, and the speed is much nicer.

]]> Comment on Trying PyPy with ReplicatorG by Matt Smollinger http://www.foobarsoft.com/2011/10/trying-pypy-with-replicatorg/#comment-12 Matt Smollinger Thu, 20 Oct 2011 00:53:46 +0000 http://www.foobarsoft.com/?p=12#comment-12 Mmmmm PyPy. I just wish it was easier to get TkInter working on Windoze. Ez_setup doesn't work unfortunately. Have you tried printing the new pypy'd versions yet? I'm curious to see a side-by-side if they are usable or not. Mmmmm PyPy. I just wish it was easier to get TkInter working on Windoze. Ez_setup doesn’t work unfortunately.

Have you tried printing the new pypy’d versions yet? I’m curious to see a side-by-side if they are usable or not.

]]> Comment on Differences in PyPy generated gcode (Updated!) by Michael Cook http://www.foobarsoft.com/2011/10/differences-in-pypy-generated-gcode/#comment-11 Michael Cook Wed, 19 Oct 2011 13:30:16 +0000 http://www.foobarsoft.com/?p=24#comment-11 That's true, but I haven't been willing to go that far to test it. I've just been using the stock OS X 10.7 Python binary, which I'm guessing is only compiled to use the x87 instructions. That’s true, but I haven’t been willing to go that far to test it. I’ve just been using the stock OS X 10.7 Python binary, which I’m guessing is only compiled to use the x87 instructions.

]]> Comment on Differences in PyPy generated gcode (Updated!) by Carl Friedrich Bolz http://www.foobarsoft.com/2011/10/differences-in-pypy-generated-gcode/#comment-10 Carl Friedrich Bolz Wed, 19 Oct 2011 13:24:10 +0000 http://www.foobarsoft.com/?p=24#comment-10 Note that if it's really the float behavior you could get the same results that PyPy now produces in CPython compiled with SSE support. Note that if it’s really the float behavior you could get the same results that PyPy now produces in CPython compiled with SSE support.

]]> Comment on Trying PyPy with ReplicatorG by Michael Cook http://www.foobarsoft.com/2011/10/trying-pypy-with-replicatorg/#comment-9 Michael Cook Wed, 19 Oct 2011 13:21:22 +0000 http://www.foobarsoft.com/?p=12#comment-9 You're right, and that did explain one of the two kinds of changes that I saw in the generated GCode. The other seems to be due to the higher precision floating point operations PyPy uses. Both are mentioned (with a note from a PyPy developer) in the <a href="http://www.foobarsoft.com/2011/10/differences-in-pypy-generated-gcode/" rel="nofollow">follow-up post</a>. You’re right, and that did explain one of the two kinds of changes that I saw in the generated GCode. The other seems to be due to the higher precision floating point operations PyPy uses. Both are mentioned (with a note from a PyPy developer) in the follow-up post.

]]> Comment on Trying PyPy with ReplicatorG by Jared Wigmore http://www.foobarsoft.com/2011/10/trying-pypy-with-replicatorg/#comment-7 Jared Wigmore Wed, 19 Oct 2011 09:24:51 +0000 http://www.foobarsoft.com/?p=12#comment-7 Sounds cool. I know that PyPy doesn't guarantee that dictionaries will be iterated in the same (arbitrary) order as CPython, could it be that? Sounds cool. I know that PyPy doesn’t guarantee that dictionaries will be iterated in the same (arbitrary) order as CPython, could it be that?

]]>