http://www.nobugs.org/blog/archives/2003/06/01/strongtalk/ Thoughts of a software engineer Sat, 19 Nov 2011 18:48:17 +0000 hourly 1 http://wordpress.org/?v=3.1.1 http://www.nobugs.org/blog/archives/2003/06/01/strongtalk/comment-page-1/#comment-14 alan knight Thu, 01 Jan 1970 00:00:00 +0000 http://www.nobugs.org/blog/archives/2003/06/01/strongtalk/#comment-14 Actually, it won't generate faster code. The Strongtalk system had two pieces - one was a static typechecker, one was an optimizing compiler. The two were completely unrelated, in that the optimizer made no use at all of the static type information, and was purely a run-time dyanmic optimizer, based on feedback from the running program. The Strongtalk papers discuss this, but I can't come up with a reference off the top of my head. Actually, it won’t generate faster code. The Strongtalk system had two pieces – one was a static typechecker, one was an optimizing compiler. The two were completely unrelated, in that the optimizer made no use at all of the static type information, and was purely a run-time dyanmic optimizer, based on feedback from the running program. The Strongtalk papers discuss this, but I can’t come up with a reference off the top of my head.

]]> http://www.nobugs.org/blog/archives/2003/06/01/strongtalk/comment-page-1/#comment-15 Andrew Thu, 01 Jan 1970 00:00:00 +0000 http://www.nobugs.org/blog/archives/2003/06/01/strongtalk/#comment-15 Ah, that's interesting. My C/C++ background meant that I saw static types, and leaped to the (incorrect!) conclusion they'd be used to optimize message sends. I've now read through the entire strongtalk tour which comes with the system, and I now understand how the optimizer works. :-) Ah, that’s interesting. My C/C++ background meant that I saw static types, and leaped to the (incorrect!) conclusion they’d be used to optimize message sends. I’ve now read through the entire strongtalk tour which comes with the system, and I now understand how the optimizer works.

]]>