search of the:   for: 
code logs -> 2009 -> Wed, 11 Nov 2009< code.20091110.log - code.20091112.log >
--- Log opened Wed Nov 11 00:00:00 2009
00:03 * Vornicus enjoys make -j 6
00:05
< MyCatVerbs>
Namegduf: No, just at least sufficiently many in excess that the CPUs never have to wait for the disks. :)
00:06
< Namegduf>
MyCatVerbs: Yeah, I was copying the old advice I read documented somewhere of 2n + 1 where n is the cores.
00:06
< Namegduf>
Which is an awesome way to increase load average.
00:06
< MyCatVerbs>
Whether that's good advice still or not requries profiling.
00:06
< McMartin>
I'm pretty sure I learned it as n+1.
00:07
< Namegduf>
I read n+1, unless you had <something which newer systems including mine had>
00:07
< MyCatVerbs>
I'm pretty sure I learned that it doesn't matter much so long as there's always something going on on every CPU.
00:07
< Namegduf>
Which is about as accurate as my memory goes.
00:07
< MyCatVerbs>
Namegduf: SMT? Hyperthreading, on Intel chips?
00:07
< Namegduf>
Yeah, it was accompanied by benchmarking.
00:07
< Namegduf>
MyCatVerbs: First one sounds right.
00:08
< Namegduf>
The benchmarking showed the last "n" of them shaved off about 10 seconds or something.
00:08
< Namegduf>
Or n-1, possibly.
00:08
< Namegduf>
(This was probably over a year ago)
00:09
< Namegduf>
I was joking, anyway, about how he should ruin the performance of everything else for the compiling.
00:09
< Namegduf>
n+1 is probably optimal "enough".
00:09
< Namegduf>
And still quite sufficient to do that.
00:10 You're now known as TheWatcher[T-2]
00:11
< gnolam>
http://www.lysator.liu.se/~gnolam/temp/Ocean_2009-11-10_wave_movement.avi [XviD, ~11 MiB] <- everything looks better in motion ^-^
00:12
< gnolam>
http://www.lysator.liu.se/~gnolam/temp/Ocean_2009-11-10_time_of_day.avi [XviD, 6 MiB]
00:12
< Derakon[work]>
Pretty snazzy.
00:13
< gnolam>
(Sorry about the quality. I suck at video encoding. And yes, I'm hitting the "accelerate time" button in the Time of Day video.)
00:13 You're now known as TheWatcher[zZzZ]
00:18
< gnolam>
My backup plan if I can't get the project itself to work on the System of Doom tomorrow is to demo those videos instead. :P
00:24 Derakon[work] [Derakon@Nightstar-d44d635e.ucsf.edu] has quit [[NS] Quit: Leaving]
01:03 Derakon[AFK] is now known as Derakon
01:14 AnnoDomini [farkoff@Nightstar-9a0feb34.adsl.tpnet.pl] has quit [[NS] Quit: God created man, God created woman, but Samuel Colt made them equal.]
02:16 gnolam [lenin@Nightstar-38637aa0.priv.bahnhof.se] has quit [[NS] Quit: Z?]
03:05 Attilla [The.Attilla@FBC920.480E8C.3E679A.E258A8] has quit [Connection reset by peer]
05:47
<@Derakon>
Google releases a new programming language, Go: http://arstechnica.com/open-source/news/2009/11/go-new-open-source-programming-l anguage-from-google.ars
05:50
< Namegduf>
Interesting.
05:50
< Namegduf>
I find goroutines intriguing.
05:52
< Namegduf>
"garbage collected" is possibly the most important part.
05:52 * simon` wants a regular expression tester for a simple set of regex operators.
05:52
< Namegduf>
A garbage collected, runs as fast as C language, C-like syntax with some evolution, emphasis on compile-time type checking and concurrency.
05:52
<@Derakon>
Sounds like just your kind of language. :)
05:52
< Namegduf>
Yeah.
05:53
< simon`>
undernet's #regex has a pretty cool bot with a number of levels that'll tell you what you might've done wrong by doing a whole bunch of tests.
05:53
< McMartin>
And a syntax apparently designed by throwing C, Python, and Perl into a blender and then throwing it against a wall
05:53
< Namegduf>
Who put Perl in and why?
05:53
< Namegduf>
But I kid.
05:53
< Namegduf>
It's what I'd consider a proper high-level language, heh.
05:53
< Namegduf>
Oh, no pointer arithmetic.
05:53
< Namegduf>
That must be helpful.
05:54
< jerith>
IS that the Rob Pike of Unix fame?
05:54
< McMartin>
I'd like to see someone do a C-like without pointer arithmetic to see if the various whole-program alias analyzers that were developed for CLR and the JVM could get you a C-like that's reliably faster than C
05:54
< McMartin>
Probably. Pike works at Google.
05:55
< Namegduf>
McMartin: Look at GO
05:55
< Namegduf>
Er, "Go".
05:55
< McMartin>
I have been~
05:55
< Namegduf>
It seems to be that.
05:55
< McMartin>
I'm not convinced they've done the full optimizer on it, becuase in practice they aren't going to need to.
05:55
< Namegduf>
Probably not yet, but it looks like it's actually conductive to being optimised.
05:55
< McMartin>
(Also, I have a distinct feeling that I got to play with a prototype of this a few years back in my Topics in Compilers class. The language looks unfamiliar but the design goals are Very Familiar Indeed.)
05:56
< Namegduf>
The speed-building goal is curious.
05:56
< McMartin>
And, uh, unimpressive, tbh
05:56
< McMartin>
gcc takes forever to build because it's bad at this.
05:57
< McMartin>
(Well, and because it's doing a bunch of insanely tedious bootstrapping effects that involve it throwing away 98% of the build products something like three times)
05:57
< Namegduf>
"Go does not support implicit type conversion. Operations that mix different types require casts (called conversions in Go)." <-- Oh, wow, awesome.
05:58
<@Derakon>
...?
05:58
<@Derakon>
Requiring explicit type conversion is a notable feature now?
05:58
< Namegduf>
In all cases? Yes.
05:58
< Namegduf>
I hate implicit conversions of any kind that may ever lose information.
05:58
< Namegduf>
Assigning of signed to unsigned, or visa versa, for example.
05:58
< McMartin>
Yay, it's +. all over again
05:58
<@Derakon>
Causes at the very least compiler warnings in all compilers I've used.
05:59
< Namegduf>
Not in C++.
05:59
<@Derakon>
C++ is a language, not a compiler.
05:59
<@Derakon>
Perhaps you should check your compiler warn settings.
05:59
< Namegduf>
Not in C++ with G++-what-goddamn-compiler-did-you-think-I-was-using.
05:59
< McMartin>
MSVC++, in fact, does warn on this.
05:59
< Namegduf>
I *think* it's with -Wall
05:59
< Namegduf>
Not 100% on that, but the warnings are set quite high.
05:59
< McMartin>
Yeah, -Wall is in fact -Wnot-really-all-ha-ha
06:00
< Namegduf>
And I think it should be in source, yeah.
06:00
< Namegduf>
As in, explicitly writtten, always.
06:00
< Namegduf>
It's a significant operation with major impact on the acceptable range of values in the starting variable.
06:00
< Namegduf>
Having this be hidden when viewing the source is quite terrible.
06:01
< Namegduf>
Hmm, no const pointers, Go?
06:01
< Namegduf>
That's "odd".
06:01
< Namegduf>
Er, qualifiers.
06:01
< Namegduf>
(Although pointers are the big helpful case.
06:01
< McMartin>
Maybe they're relying on program analysis for that.
06:01
< Namegduf>
Maybe.
06:01
< McMartin>
If you don't have pointer arithmetic, consting references gets a lot less important
06:01
< Namegduf>
I like guaranteeing that functions won't alter things they're given pointers to.
06:01
< Namegduf>
Ah.
06:02
< Namegduf>
No function overloading is odd.
06:02
< Namegduf>
I can understand why no user-defined operations; redefining the language is a bad idea.
06:02
< McMartin>
Stepping back a bit, -Wconversion is not in fact one of the -Wall options, and it looks like it does do what you want.
06:03
< Namegduf>
That's weird and poor of GCC.
06:03
< McMartin>
GCC's only true advantage is its ubiquity, not kidding
06:03
< McMartin>
I look forward to it getting an LLVM backend that actually works
06:04
< McMartin>
I also look forward to gdb being able to work with Mach-O binaries >_<
06:04
< McMartin>
Since right now nothing can.
06:04
< McMartin>
(In a multithreaded OS X program, "step" in gdb is synonymous with "continue")
06:04
< McMartin>
(You have to manually breakpoint every source line.)
06:05
< Namegduf>
...defer.
06:05
< simon`>
McMartin, there's a guy at my uni who is working on a port of Moscow ML to an LLVM backend.
06:05
< Namegduf>
That is possibly one of the best things I've ever wanted to have; roughly usable as a deconstructor for a function call.
06:06
< McMartin>
Namegduf: As in, arbitrary paired calls within a scope?
06:06
< Namegduf>
"The defer statement may be used to call a function after the function containing the defer statement returns."
06:06
< McMartin>
Yeah, that.
06:06
< Namegduf>
So, yeah.
06:06
< McMartin>
That's old, but nobody fucking implements it rargh
06:06
< McMartin>
The fact that C# has something like it and Java doesn't is at least 12 reasons C# is better than Java
06:07 * Namegduf will not let mono anywhere near his system, despite using WINE
06:07 * simon` read that some people had made Concurrent C#
06:07
< McMartin>
IIRC, Delphi has something like that directly, and you can fake it reasonably effectively in C# (which is Delphi descended, so no surprise there), Python (as of 2.6: the 'with' clause), and C++ (stateless stack objects).
06:07
< simon`>
syntactically it's not that bad... they add an 'async' keyword and some more.
06:09
< Namegduf>
Hmm, so goroutines are basically "cheap threads"
06:10
< McMartin>
I'm not sure I've ever seen coroutines defined the same way twice, but yeah, that's what it boils down to.
06:10
< McMartin>
I tend to think of it as co-op threading that doesn't pretend to be real threading.
06:10
< Namegduf>
Yeah; what's different is that it does real threading to distribute them.
06:10
< McMartin>
That wasn't really my field so I don't know if that makes it easier to prove optimizations on it.
06:10
< Namegduf>
So it's co-op threading that gets the scalability bonuses of real threading, with the cheapness of cheap threading.
06:11
< McMartin>
I hope they aren't using usermode threads >_<
06:12
< ToxicFrog>
Where are you getting this information on goroutines? The article doesn't actually say whether they're OS threads, green threads, or coroutines
06:12
< ToxicFrog>
(or something else entirely)
06:12
< McMartin>
TF: I'm actually getting it from MUD gossip~
06:12
< Namegduf>
http://golang.org/doc/go_for_cpp_programmers.html#Goroutines
06:13
< Namegduf>
"Internally, goroutines act like coroutines that are multiplexed among multiple operating system threads. You do not have to worry about these details."
06:13
< McMartin>
I hope they're not green threads though because green threads are quite possibly the worst idea in the history of ever.
06:13
< McMartin>
If you don't in fact have to worry, they aren't green threads.
06:13
< McMartin>
Since the defining feature of green threads is that if one thread blocks on I/O all the threads are slept
06:16
< Namegduf>
Ah, ew.
06:25
< Namegduf>
Huh. It always initialises to zero.
06:25
< Namegduf>
If an initial value isn't provided.
06:26
< Namegduf>
I suppose if the compiler can guarantee that something is always written before it's read, it can optimise that out?
06:44 Orthia [Orthianz@Nightstar-abe70ff9.xnet.co.nz] has joined #code
06:44
< Orthia>
I have a really silly question that is nonetheless tripping my brain.
06:45
< Orthia>
You have something that has a cooldown of 100 seconds. You have a trick that lets you reduce the cooldown by 5 seconds every 2.
06:45
< Orthia>
How long does it take with the modifier going?
06:46
< Namegduf>
Is that in addition to its normal cooling down within the five?
06:46
< Orthia>
It's cooling down naturally during that too.
06:47
< Namegduf>
(100/7)*2 is my immediate thought, but I need to look for gotchas, aside possibly needing to always round up to the nearest 2.
06:47
< Orthia>
It's always gotta be 2 seconds between triggering.
06:47
< jerith>
Every two seconds you're getting seven seconds of cooldown.
06:47
< Namegduf>
Yeah, that'd cause the roundup unless it was 'very small'
06:47
< Orthia>
And the cooldown rate itself is variable (Up to 50 second base recharge)
06:47
< jerith>
So it would be ceil(100*2/7)
06:48
< Namegduf>
ceil(100/7)*2
06:48
< Namegduf>
I think.
06:48
< Namegduf>
Round to the nearest two, not the nearest one.
06:48
< jerith>
True. Namegduf's is right.
06:49
< Namegduf>
That's not technically true; if it's got only 1/7 left, it only takes 1 instead of 2 the last time.
06:49
< Namegduf>
(In general, for <=2, it takes that time, not the full 2)
06:49
< Namegduf>
Not sure how to easily represent that, though.
06:49
< jerith>
Consider 15.
06:50
< jerith>
Um, 13.
06:50
< jerith>
2 seconds drop that to 6.
06:50
< jerith>
Another 2 drops that to 0.
06:50
< jerith>
Now consider 15
06:50
< jerith>
4 seconds drops that to 1.
06:51
< jerith>
For a total of 5s.
06:51
< jerith>
My head hurts now. :-(
06:52
< Namegduf>
Yeah.
06:55
< Namegduf>
time = 0; for (cooldown = 100; cooldown > 0; cooldown -= 2) { if (cooldown > 2) { cooldown -= 5; time += 2; } else time += cooldown; }
06:55
< Namegduf>
In C-without-types, the worst language in history.
06:55
< jerith>
(C div 7)*2 + min(2, C rem 7)
06:55
< Namegduf>
Haha.
06:55
< Vornicus>
30 seconds. you get 14 cooldown reducers.
06:56
< Namegduf>
Yeah.
06:56
< Namegduf>
(I love how we worked out how to calculate it generically instead of answering)
06:56
< jerith>
The general case is more interesting.
06:56
< Namegduf>
It is!
06:57
< jerith>
Because we can implement it and return the original solution as a special case.
06:57
< jerith>
08:47 < Orthia> And the cooldown rate itself is variable (Up to 50 second base recharge)
06:57
< jerith>
And that implies that a general solution would be useful anyway. :-)
07:12
< Orthia>
The general case is indeed most interesting.
07:14 Derakon is now known as Derakon[AFK]
07:17
< simon`>
I'm using this algorithm to convert an NFA into a DFA. the first part is determining the set of epsilon-transitions. http://stuff.anarcho.dk/epsilon.png
07:22
< simon`>
rather, the minimal set of epsilon-transitions, so ?\in F(?)\in F(F(?))\in ... until F(...(?)...) = F(...(F(?))...) (i.e. no more transitions)
07:24
< simon`>
F(?) = {1}, since there is an epsilon-transition to state 1.
07:25
< simon`>
F({1}) = {1}u{2,3}, since there are e-transitions to those two
07:26
< simon`>
F({1,2,3}) = {1,2,3}, I think, since there are no direct e-transitions from 2 or 3. so even though there are e-transitions in the rest of the NFA, I don't think they're part of the solution to this algorithm.
07:26
< simon`>
*sigh*
07:29
< simon`>
ah yes. that's it.
07:54 Orthia [Orthianz@Nightstar-abe70ff9.xnet.co.nz] has quit [Client closed the connection]
08:02 Orthia [Orthianz@Nightstar-abe70ff9.xnet.co.nz] has joined #code
08:18 gnolam [lenin@Nightstar-38637aa0.priv.bahnhof.se] has joined #code
08:51 You're now known as TheWatcher
08:59 Rhamphoryncus [rhamph@Nightstar-a62bd960.abhsia.telus.net] has quit [Client exited]
09:07
< gnolam>
#ifdef LEPPARD
09:07
< gnolam>
#endif
09:10 * Vornicus fires gnolam out of a cannon into a volcano.
09:14 Vornicus is now known as Vornicus-Latens
09:55
< simon`>
does emacs regex have the equivalent to pcre's (?!...)?
09:56
< simon`>
that is, grouping that doesn't backref.
09:56
< simon`>
I can't find it on any webpages, so I'm guessing not.
10:34 EvilDarkLord [jjlehto3@Nightstar-f1ccbb45.hut.fi] has quit [Ping timeout: 121 seconds]
10:35 EvilDarkLord [jjlehto3@Nightstar-f1ccbb45.hut.fi] has joined #code
10:58 Attilla [The.Attilla@FBC920.480E8C.3E679A.E258A8] has joined #code
10:58 mode/#code [+o Attilla] by ChanServ
11:00 AnnoDomini [farkoff@Nightstar-9a0feb34.adsl.tpnet.pl] has joined #code
11:07 EvilDarkLord [jjlehto3@Nightstar-f1ccbb45.hut.fi] has quit [Ping timeout: 121 seconds]
11:07 EvilDarkLord [jjlehto3@Nightstar-f1ccbb45.hut.fi] has joined #code
11:13 AnnoDomini [farkoff@Nightstar-9a0feb34.adsl.tpnet.pl] has quit [Ping timeout: 121 seconds]
11:18 AnnoDomini [farkoff@Nightstar-30729f3d.adsl.tpnet.pl] has joined #code
14:12 gnolam [lenin@Nightstar-38637aa0.priv.bahnhof.se] has quit [[NS] Quit: Das Reboot]
14:17 gnolam [lenin@Nightstar-38637aa0.priv.bahnhof.se] has joined #code
15:15
< AnnoDomini>
Pfew. Almost thought I'd actually have to install IE to run this ASP.NET thingie.
15:23
< AnnoDomini>
Hmmmm. Decreasing colour depth does help with JPG splotches. If I were to combine a simple odd-dot-removal filter of some sort with an algorithm that figures out what colours are necessary to render the image (with some epsilon>0 for accuracy), I could build something that would work.
16:31 Derakon[AFK] is now known as Derakon
17:01 Rhamphoryncus [rhamph@Nightstar-a62bd960.abhsia.telus.net] has joined #code
17:31 Pinkhair [none@Nightstar-a7d2ccfd.dyn.optonline.net] has quit [Client closed the connection]
17:37 AgentIchiro [Smith@Nightstar-6450f657.dsl.teksavvy.com] has joined #code
17:38 AgentIchiro [Smith@Nightstar-6450f657.dsl.teksavvy.com] has quit [[NS] Quit: Leaving]
17:39 SmithKurosaki [Smith@Nightstar-6450f657.dsl.teksavvy.com] has quit [Connection closed]
17:40 SmithKurosaki [Smith@Nightstar-6450f657.dsl.teksavvy.com] has joined #code
18:11
< gnolam>
Guess if I could get the ocean renderer to work on the System of Doom? :P
18:11
< gnolam>
Thank Eris I made those videos.
18:22 * gnolam ponders making a Carcass song title generator.
18:22
< gnolam>
Either that or use their titles as RPG monsters/spells.
18:24 Pinkhair [none@Nightstar-a7d2ccfd.dyn.optonline.net] has joined #code
18:27
< gnolam>
"You cast Magic Missile. The Cadaveric Incubator of Endoparasites loses 30 HP."
18:56 You're now known as TheWatcher[afk]
19:00 crem [moo@Nightstar-8ca3eea7.adsl.mgts.by] has quit [Ping timeout: 121 seconds]
19:06 Pinkhair [none@Nightstar-a7d2ccfd.dyn.optonline.net] has quit [[NS] Quit: ]
19:10
< dmlandrum>
Heh
19:11
< dmlandrum>
You could probably do the same with Sepultura.
19:51 You're now known as TheWatcher
20:06 Vornicus-Latens is now known as Vornicus
20:37
< McMartin>
Hm
20:38
< McMartin>
I bet my Ultima IV save-hacking program would be a good test of Go.
20:54
< gnolam>
Whyso?
21:14 Reivles [reaverta@ServerAdministrator.Nightstar.Net] has quit [Ping timeout: 121 seconds]
21:26 * gnolam hmms.
21:27
< McMartin>
gnolam: Because it's small, does something that's currently handy for me, and is a commandline utility.
21:28
< McMartin>
Heh. "By Google. Compiles into machine code. Garbage collected. Type inference. Not quite object oriented. A bit Plan-9-ish. Kinda ugly."
21:28
< McMartin>
That pretty much sums it up
21:29
< McMartin>
I haven't been able to work out if it still has true dynamic typing and is just being smart with type inference to make errors happen sooner, or if it's statically typed and the dynamic tricks actually don't work as a result.
21:31
< McMartin>
Binary struct unpacking (which a save-hacker is) is also a good test of collection heterogeneity and dynamic typing as a whole.
21:31
< McMartin>
So all in all, nice, short, and reasonably comprehensive.
21:31
< McMartin>
And I'm currently playing Ultima IV so having the save hacker is nice~
21:41
< AnnoDomini>
Say, would anyone know off-hand how to make dynamic objects in C#? Trying to do it exactly like in Java yields error highlights.
21:41
< AnnoDomini>
I want a new Label, for instance.
21:43
< McMartin>
Sorry; I haven't done any C# GUI work, just console stuff and not much of that.
21:43
< McMartin>
What do you mean by 'dynamic objects' here, though?
21:43
< AnnoDomini>
Label l1 = new Label(); // this, in Java
21:44
< McMartin>
IIRC, C# has stack objects, so you may have to mark l1 as being boxed. Let me check something real quick.
21:45
< McMartin>
No, that should be fine, unless Label has no default constructor.
21:46
< McMartin>
You may need to do Label l1 = new Label("blah blah blah"), and you may have to set your using clauses so that it's the right one
21:46
< McMartin>
Or do Windows.Forms.Label or whatever they call it.
21:46
< AnnoDomini>
Hmm. It works now, for some reason. Earlier, it underlined it as faulty (expecting ;).
21:46
< McMartin>
http://www.csharp-station.com/Tutorials/Lesson07.aspx
21:46
< McMartin>
Perhaps it was missing a ";" at that time.
21:46
< AnnoDomini>
It was not.
21:50
< McMartin>
"That time" was "the last time Intellisense looked at it".
21:52
< AnnoDomini>
I seep.
21:53 Pinkhair [pinkhair3d@Nightstar-a7d2ccfd.dyn.optonline.net] has joined #code
21:54
< gnolam>
AnnoDomini: I hear you can buy protection for that sort of thing.
21:54
< AnnoDomini>
Intellisense or seepage?
22:14 Vornicus [vorn@ServerAdministrator.Nightstar.Net] has quit [[NS] Quit: ]
22:19 Vornicus [vorn@ServerAdministrator.Nightstar.Net] has joined #code
22:26
< AnnoDomini>
Ah, I see. C# is case-sensitive. "New" is not "new" and it interprets it somehow as a missing semicolon.
22:27
< McMartin>
Read it as a type declarator and then another declaration.
22:59 Orthia [Orthianz@Nightstar-abe70ff9.xnet.co.nz] has quit [Ping timeout: 121 seconds]
23:00 Reivles [reaverta@ServerAdministrator.Nightstar.Net] has joined #code
23:01 Orthia [Orthianz@Nightstar-abe70ff9.xnet.co.nz] has joined #code
23:07 Orthia [Orthianz@Nightstar-abe70ff9.xnet.co.nz] has quit [Ping timeout: 121 seconds]
23:19 Orthia [Orthianz@Nightstar-beaca266.xnet.co.nz] has joined #code
23:31
< McMartin>
In other news, Go does not in fact compile in under 10 seconds~
23:40
< McMartin>
Hello world: approx 780k.
23:40
< McMartin>
Er. 580k.
23:40
< gnolam>
k what?
23:41
< gnolam>
I'm hoping it's not kB or kLOC of source code. :)
23:41
< AnnoDomini>
God damn it. This last task was obviously not checked for completeness and sense.
23:41
< AnnoDomini>
I can infer some things, but purely by trying to think like whoever wrote this, not from face value.
23:42
< McMartin>
kB of binary.
23:43 * McMartin assumes that this is because, like in Haskell, the complete runtime is being statically linked.
23:44
< AnnoDomini>
I remember fondly my External Devices class, where we used x86 assembly for HelloWorld.exe - the size was measured in BYTES.
23:44
< McMartin>
Well.
23:45
< gnolam>
I fail to see how a class using x86 assembly can be remembered fondly. :o
23:45
< McMartin>
You were actually calling out to the OS to do the output for you.
23:45
< McMartin>
This like like my Windows program that had a mouse-driven UI and was 8k.
23:46
< AnnoDomini>
gnolam: It was one of the few classes where our efforts had measurable effect.
23:46
< Tarinaky>
I have a computing workshop tomorrow. :/
23:48
< gnolam>
That's why I like computer graphics. Instant gratification.
23:53
< Namegduf>
McMartin: Go's own... library things in Go do.
23:53
< Namegduf>
McMartin: Go itself contains a lot of C, though, so...
23:57
< McMartin>
mcmartin@spiff:~/devel/gotest$ ldd hello statically linked
23:59
< Namegduf>
I wonder if it can dynamically link as an option, becoming dependent on Go's runtime library as a result.
--- Log closed Thu Nov 12 00:00:14 2009
code logs -> 2009 -> Wed, 11 Nov 2009< code.20091110.log - code.20091112.log >
 search of the:   for: