IRC Logs

Network: freenode Channel: #csharp

	Enter your search terms Submit search form
	Web www.irclog.org

[14:57]<rnmgxjsv>yeah im sorry java is crap
[14:57]<lyfuym-wjzc>java is not crap
[14:57]<rnmgxjsv>my friend works for sun, and he wouldnt code in java to save his life
[14:57]<lyfuym-wjzc>the virtual machine is crap
[14:57]<mjzymygo>Java isn't crap
[14:57]<mjzymygo>just it's VM
[14:58]<lyfuym-wjzc>Moridin8 *snap*
[14:58]<mjzymygo>... well... and the backwards compat it has to deal with... but hey...
[14:58]<zyccus>http://www.tommti-systems.de/main-Dateien/reviews/languages/benchmarks.html
[14:58]<rnmgxjsv>ok maybe "crap" is a bit harsh
[14:58]<rnmgxjsv>but its far from good
[14:58]<rnmgxjsv>woah @ the java memory usage
[14:59]<zyccus>C++ 'wins' but java and c# are faster in some cases
[14:59]<rnmgxjsv>it says java beats C#, but ive only had bad experiences with java apps
[15:00]<zyccus>what kind of bad experiences?
[15:00]<rnmgxjsv>slow ass applications
[15:00]<mjzymygo>rikkus: I can tell you quite categoriically that those benchmarks are incorrect when you place C++/Java against .NET 2 based c#
[15:00]<zyccus>where does it say that java beats c#?
[15:00]<mjzymygo>C# kinda rips ass
[15:00]<rnmgxjsv>oh
[15:00]<zyccus>Moridin8: I'm sure... I know it got some optimisations
[15:00]<rnmgxjsv>i read it wrong
[15:00]<rnmgxjsv>java gets 5 wins
[15:00]<rnmgxjsv>but C# 9
[15:00]<mjzymygo>yeh... less boxing and Generics
[15:01]<mjzymygo>upto 3 times faster
[15:01]<rnmgxjsv>oh they even used the Intel C/C++ compiler
[15:01]<rnmgxjsv>outperforms msvs in alot of cases (not all)
[15:01]<y-gz>of course it should
[15:01]<y-gz>intel designed the cpus lol
[15:02]<rnmgxjsv>haha
[15:02]<sdnffdxdlld>java is improving alot over previous versions too (because .net was whooping its arse)
[15:02]<y-gz>they really *should know* which instruction takes less clocks and also how to optimize a sequance
[15:02]<mjzymygo>JEE5 is that attempt smellyhippy
[15:02]<mjzymygo>but it falls well short
[15:03]<mjzymygo>still has to cater for older libraries and code
[15:03]<rnmgxjsv>eh nowadays the clocks are very similar its more bottlenecks and pipelines and etc
[15:03]<sdnffdxdlld>theres considerable improvment in some areas ... .net is still better of course ^_^ but it has improved
[15:03]<rnmgxjsv>but yeah i agree intel knows their stuff
[15:03]<zyccus>last time I tried Intel's compiler versus GCC, GCC won by a mile... that was on Linux, and a long time ago, though.
[15:04]<lyfuym-wjzc>y0 rikkus
[15:04]<rnmgxjsv>are you sure you benchmarked properly
[15:04]<rnmgxjsv>GCC doesnt even beat the MSVS compiler
[15:04]<zyccus>RedGhost: yes
[15:05]<rnmgxjsv>odd
[15:05]<zyccus>I didn't optimise for either compiler
[15:05]<lyfuym-wjzc>GCC owns
[15:05]<zyccus>I think I used some 'real world' app, like the vorbis encoder
[15:05]<rnmgxjsv>ugh @ gcc owns
[15:05]<rnmgxjsv>the inline asm is terrible
[15:05]<y-gz>lol
[15:05]<y-gz>you can switch to intel syntax
[15:06]<lxzjsvd>gcc's inline asm has bad docs
[15:06]<y-gz>i dislike the at&t one too
[15:06]<rnmgxjsv>yeah i found out about that a few months ago
[15:06]<rnmgxjsv>but it was too late, i had already given up on GCC
[15:07]<rnmgxjsv>maybe it's my subconscious hate for linux fanboys
[15:08]<rnmgxjsv>who knows
[15:09]<zyccus>gcc is pretty horrible
[15:09]<y-gz>in what sense?
[15:09]<zyccus>in terms of how long it takes to compile
[15:09]<y-gz>hm
[15:09]<zyccus>apart from that, it's fine
[15:09]<y-gz>isn't that because of automake and co
[15:09]<zyccus>no, but they make matters worse
[15:10]<y-gz>plus are you talking about the 4.x.x
[15:10]<zyccus>it's always been horrifically slow
[15:10]<y-gz>:/
[15:10]<zyccus>I had to buy a very expensive computer just because gcc took so long to compile KDE
[15:10]<zyccus>compiling the same code with MSVC was about 20x faster
[15:11]<rnmgxjsv>you opted to buy a new computer rather than use MSVC? lol
[15:11]<zyccus>you can't really use MSVC to compile code on Linux
[15:12]<rnmgxjsv>wine ^
[15:12]<zyccus>doesn't work that way
[15:12]<zyccus>you can get the compiler to run... but you need to produce ELF binaries
[15:12]<rnmgxjsv>yeah i know im just teasing
[15:13]<zyccus>anyway, MSVC at the time (6) couldn't cope with bits of C++
[15:13]<rnmgxjsv>i think Open Watcom can compile for *nix( correct me? ), and there is always the intel compiler
[15:13]<zyccus>the intel one was about the same speed, though it did _work_
[15:13]<zyccus>yes I tried some watcom thing - never got it to work properly
[15:14]<rnmgxjsv>i used MSVC 6.0 back when i coded in C++
[15:14]<zyccus>your choice is basically gcc or icc
[15:14]<rnmgxjsv>and for me lcc :p
[15:14]<rnmgxjsv>PellesC@win32 ftw
[15:16]<rnmgxjsv>eh yeah its 5:14 am, maybe time for sleep :\
[15:28]<gu_ggj>what is the corespondent type for stringtokenizer in c# ?
[15:29]<gu_ggj>pls
[15:29]<sdnffdxdlld>string[] stringTokens = string.Split(' ')?
[15:29]<sdnffdxdlld>if you want to split on a space
[15:29]<gu_ggj>tx
[15:30]<sdnffdxdlld>np
[15:30]<sdnffdxdlld>SpikeLite google string.split
[15:30]<slycnlyvn>Result: String (Java 2 Platform SE v1.4.2) : http://java.sun.com/j2se/1.4.2/docs/api/java/lang/String.html
[15:30]<sdnffdxdlld>opps
[15:31]<sdnffdxdlld>SpikeLite msdn string.split
[15:31]<slycnlyvn>Google service is currently b00rked, please try again in a bit.
[15:31]<sdnffdxdlld>lol
[15:31]<sdnffdxdlld>SpikeLite msdn string.split
[15:31]<slycnlyvn>smellyhippy, String.Split Method (Char[]) (System): http://msdn2.microsoft.com/en-us/library/b873y76a.aspx
[15:31]<sdnffdxdlld>there ya go ... that has info on what you can pass and stuff

Back to #csharp
Or go to some related logs:
#gentoo
#python
#gentoo
#gentoo
#gentoo