#pentagram@irc.freenode.net logs for 20 Jan 2009 (GMT)

Archive Today Yesterday Tomorrow
Pentagram homepage


[00:11:10] <-- Fingolfin has left IRC ()
[00:26:48] --> watt has joined #pentagram
[00:26:48] --- ChanServ gives channel operator status to watt
[00:51:13] <-- watt has left IRC ()
[03:26:10] --> watt has joined #pentagram
[03:26:10] --- ChanServ gives channel operator status to watt
[03:38:43] <-- Colourless has left IRC (Read error: 148 (No route to host))
[07:08:08] --> Colourless has joined #Pentagram
[12:49:04] <-- Kirben has left IRC (Read error: 145 (Connection timed out))
[14:23:09] <-- watt has left IRC ()
[14:54:30] --> watt has joined #pentagram
[14:54:30] --- ChanServ gives channel operator status to watt
[15:56:20] --> pupnik_ has joined #pentagram
[15:58:09] <-- pupnik has left IRC (Read error: 60 (Operation timed out))
[19:52:55] <-- watt has left IRC (kornbluth.freenode.net irc.freenode.net)
[20:10:26] --> watt has joined #pentagram
[21:50:05] <watt> hmm... gee, I sure wish I would've know about the placement new operator in C++ back when I was messing with memory management
[21:50:08] <watt> http://en.wikipedia.org/wiki/Placement_new
[21:50:32] <watt> Never encountered that before now. Sigh.
[21:51:41] <wjp> hi watt
[21:52:02] <wjp> you didn't use that in the memory manager?
[21:54:06] <watt> I overloaded the tradition new and delete operators.
[21:57:02] <watt> And per class that I cared about too. Thinking I could've gotten away with something a bit cleaner now.
[22:02:33] <watt> Well, in any case, it was probably silly code to add in the first place - would have been better off attempting to cleanup memory intensive code rather than avoid the cost of allocation.
[22:08:56] --> Kirben has joined #pentagram
[22:08:56] --- ChanServ gives channel operator status to Kirben
[23:00:10] <watt> To be specific on what I found interesting, this is a good example of what llvm's clang project is doing: http://lists.cs.uiuc.edu/pipermail/cfe-commits/Week-of-Mon-20090119/011064.html
[23:15:00] <watt> Ah.. well, that's cute. They added a perl script to do something similar to that uberlipo script I made to make universal packages - http://llvm.org/viewvc/llvm-project/cfe/tags/Apple/clang/merge-lipo?revision=62628&view=markup&pathrev=62628
[23:16:28] * Colourless looks at it then thinks why am I looking at that
[23:16:44] <watt> hehe