AppletTalk.com Java discussions newsgroups   Archives   FAQ   Search   Memberlist   Usergroups   Register   Profile   Log in to check your private messages   Log in  Why my java code is THAT slow compared too C++? Goto page Previous  1, 2, 3, 4 ... 328, 329, 330  Next          AppletTalk.com Forum Index -> Java Language Programming View previous topic :: View next topic   Author Message ossie.moore@gmail.com Guest Posted: Sun Feb 12, 2006 6:35 am    Post subject: Re: How to deploy Java Applications To People W/O Having To just the jar. Infact, you'll note that my example only updates the jar, not the jnlp file (see recompile.sh content in the index.html page) Back to top ossie.moore@gmail.com Guest Posted: Sun Feb 12, 2006 6:35 am    Post subject: Re: How to deploy Java Applications To People W/O Having To I'm doing exactly what you state. I mentiont that i'll stop it runnign simply because i don't want to leave a script compiling a source file once every two minutes forever. Back to top ossie.moore@gmail.com Guest Posted: Sun Feb 12, 2006 6:35 am    Post subject: Re: log4j two levels at once You can do that and so much more. Example.. for both the default logger and for any particular class, you can setup multiple loggers, at different levels. Example... Say you have a program doing something odd and you want to debug a particular class. You could set the default logger to "INFO" for Console, "WARN" for file. Then, add an entry for the suspect class to log at "DEBUG" for Console. This way, you'd see DEBUG messages on the console for that particular class. Heck, you could even send DEBUG level logging for a particular class to say syslog while having another class send DEBUG to another file. You can also vary the format of the log messages from appender to appender. Back to top Luc The Perverse Guest Posted: Sun Feb 12, 2006 6:35 am    Post subject: Re: Windows or Linux? "Chris Smith" <cdsmith (AT) twu (DOT) net> wrote in message news:MPG.1e572699ccd42c92989d9d (AT) news (DOT) altopia.net... Quote: Luc The Perverse <sll_noSpamlicious_z_XXX_m (AT) cc (DOT) usu.edu> wrote: It is clear that you are new to programming. Try not using a path at all, and see where files go. You should be able to use this "working directory" to keep your files in. Sorry, Luc, but I think that's really bad advice. In a GUI environment, the current working directory is almost always an unclear concept to the user, and it's not defined by any relevant specification. Applications should not make use of the current working directory unless it's clearly the intent of the user for it to do so, which implies that it's probably a command line utility. Command line utilities are among the rarest uses of Java applications, and in the few cases where they are written, the author already knows, as a fact, that the CWD is required. He;s tring to learn very basic programming concepts - my suggestion was to get everything else working before worrying about platform independant pathing. He is obviously on a windows box right now using a hard coded path -- LTP :) Back to top Luc The Perverse Guest Posted: Sun Feb 12, 2006 6:35 am    Post subject: Re: Why my java code is THAT slow compared too C++? "Thomas Hawtin" <usenet (AT) tackline (DOT) plus.com> wrote in message news:43ee4358$0$9225$ed2619ec@ptn-nntp-reader01.plus.net... Quote: That is not necessarily why the C code is faster. My initial gut reaction, and his solution both seem to hint at the idea that he was using an inefficient algorithm. I have coded a word frequency calculator in C++ before, and used N^2 efficiency sorting algorithms and they were VERY slow. So I disagree - he changed one thing and now they run in about the same amount of time as opposed to the java running more than double. To me that seems pretty conclusive. -- LTP :) Back to top ossie.moore@gmail.com Guest Posted: Sun Feb 12, 2006 6:35 am    Post subject: Re: max heap size on 32 bit linux ? This has to do with the max memory a particular process can occupy. I believe the *default* value most vendors us in workstation editions of Linux is 2GB of ram per process (minus some overhead). You can reconfigure your kernel manually to allow more memory, or, if you are not comfortable doing that, purchase an enterprise edition support package. Here's an old post i found by googling for "linux max memory per process"... http://lists.samba.org/archive/linux/2005-November/014552.html Back to top priyom Guest Posted: Sun Feb 12, 2006 9:12 am    Post subject: Re: javamail send error "The fact that your address is valid is not sufficient, since a correctly configured mail server will only accept mail to or from its own mail domain." -What I meant was it's not an address problem. I am in the same domain, connecting through LAN and trying to send a mail to myself, which I presume would be in the same domain (I do have an ID created in the domain with send capability) Back to top Jon Martin Solaas Guest Posted: Sun Feb 12, 2006 10:12 am    Post subject: Re: cpu intesive jave program puzzlecracker wrote: Quote: Basically, you need some way of hooking into the scheduler, or know exactly what the other processes on the system are doing. do i really need know what other processes are doing? Let's say I am not concerned with current percentage, but only with a value I want it to boost to. ex. currently cpu%=50 I want to to raise it by 20%, and that is fine if some other processes raise it by some value n during interupts.... All I care is to know that my call can raise it if there is a room. That is all. What exactly are you after? Boosting the jvm process scheduler priority? You can use System.execute to run the renice command, but this won't set a fixed cpu utilization, if the process is given low priority it'll still consume 100% cpu if there aren't other processes in need of cpu-cycles. Likewise, your process may be slowed down if it's io-intensive, and won't be able to consume all cpu-cycles available. Maybe you should look into a platform with realtime features, like RT-Linux. And if you describe exactly what you need to do I'm sure someone will be able to come up with much better answers than this :-) -- jon martin solaas Back to top Gordon Beaton Guest Posted: Sun Feb 12, 2006 10:12 am    Post subject: Re: Null Webserver On Sun, 12 Feb 2006 02:09:25 -0700, Luc The Perverse wrote: Quote: You see though, that's the thing - I don't know where to look for the simplest protocal instructions. I search for HTML specification and find a W3C subsite that is at least hundreds of pages long - and it still doesn't answer the most basic questions. Then I realize HTML specification isn't what I'm looking for, so I search for web server specifications and get nothing related to what I need. HTTP is described in rfc2616. Some extensions are described in other documents but rfc2616 is all you need to write a basic web server. /gordon -- [ do not email me copies of your followups ] g o r d o n + n e w s @ b a l d e r 1 3 . s e Back to top Luc The Perverse Guest Posted: Sun Feb 12, 2006 10:12 am    Post subject: Re: Null Webserver "Chris Uppal" <chris.uppal (AT) metagnostic (DOT) REMOVE-THIS.org> wrote in message news:43ee01d5$1$1171$bed64819 (AT) news (DOT) gradwell.net... Quote: Anyway, it would be a reasonable starting point if you prefer to work by reading and modifying existing code. If not, then creating a simple HTTP server from scratch would be an interesting and informative exercise, one I suspect you would enjoy -- the Sun code gives you an idea of the scale of the problem You see though, that's the thing - I don't know where to look for the simplest protocal instructions. I search for HTML specification and find a W3C subsite that is at least hundreds of pages long - and it still doesn't answer the most basic questions. Then I realize HTML specification isn't what I'm looking for, so I search for web server specifications and get nothing related to what I need. I assume it's a TCP connection, but I'm not positive. I could mindlessly hack away at the example code, but I don't think that would do much. Or I could start at the other end and use a packet sniffer and try to determine what it is doing. This would be a good learning experience, but I'm not sure I'm that bored. What I'd like is a build your own webserver tutorial - or at least an incremental approach to understanding webserver communications. If that doesn't exist, where can I find a mature full specification of the communication? -- LTP :) Back to top Gordon Beaton Guest Posted: Sun Feb 12, 2006 11:12 am    Post subject: Re: Null Webserver On Fri, 10 Feb 2006 16:27:49 -0700, Luc The Perverse wrote: Quote: To reduce ads that I see, I have downloaded a 1/2 mb hosts file which routes most DNS to localhost 127.0.0.1 Problem is, sometimes it displays "cannot find page" and it tries and waits for a picture to load. I'd like to fill the void, with a webserver that serves nothing but blank pages, and null media. (1x1 transparent gif, 1x1 JPG, Empty do nothing Flash etc.) I've done something similar to generate random pages. All that's necessary is that you read (and ignore) the request, then respond with whatever content you like. Run it from inetd or similar and you don't even need to deal with the socket connections, you can read from System.in and write to System.out. It really doesn't need to be more complicated than this (this code is untested): // get request br = new BufferedReader(...(System.in)); while (((line = br.readLine()) != null) && (line.length != 0)); // response header System.out.println("HTTP/1.1 200 OK"); System.out.println("Date: %s" + rfcDate()); System.out.println("Server: Apache/2.0.50 (Unix) DAV/2"); System.out.println("Connection: close"); System.out.println("Content-Type: text/html; charset=iso-8859-1"); System.out.println(); // response body System.out.println("<!DOCTYPE HTML PUBLIC \"-//IETF//DTD HTML 2.0//EN\">"); // add your null content here System.out.println(getNullContent()); /gordon -- [ do not email me copies of your followups ] g o r d o n + n e w s @ b a l d e r 1 3 . s e Back to top Gordon Beaton Guest Posted: Sun Feb 12, 2006 11:12 am    Post subject: Re: Null Webserver On 12 Feb 2006 11:30:31 +0100, Gordon Beaton wrote: Quote: System.out.println("Date: %s" + rfcDate()); My C is showing. Should have been: System.out.println("Date: " + rfcDate()); /gordon -- [ do not email me copies of your followups ] g o r d o n + n e w s @ b a l d e r 1 3 . s e Back to top Thomas Hawtin Guest Posted: Sun Feb 12, 2006 11:12 am    Post subject: Re: Why my java code is THAT slow compared too C++? Luc The Perverse wrote: Quote: "Thomas Hawtin" <usenet (AT) tackline (DOT) plus.com> wrote in message news:43ee4358$0$9225$ed2619ec@ptn-nntp-reader01.plus.net... That is not necessarily why the C code is faster. My initial gut reaction, and his solution both seem to hint at the idea that he was using an inefficient algorithm. It is vastly more important that the hash code produce a good answer than that it be ultra quick. With two multiplies, the hash code computation wasn't as cheap as it might have been. Sun's JVM seems poor at replacing constant multiplies with shifts and adds/subtracts, although that is less important with faster multipliers. If you replace v = v*31; with v = (v<<5) - v; and perhaps look at two together (v = (v<<10) - 2*(v<<5) + v you might understand that the low order bits are not going to be fairly distributed. Quote: I have coded a word frequency calculator in C++ before, and used N^2 efficiency sorting algorithms and they were VERY slow. Doing n operation on a hash map take O(n) time, apparently. Screw your hash code up, and it becomes O(n^2) and your caches burn for an extra constant time slow down. Quote: So I disagree - he changed one thing and now they run in about the same amount of time as opposed to the java running more than double. To me that seems pretty conclusive. How the hash code was calculation was performed changed (almost certainly to a slower technique). Checking the implementation of HashMap, it will only calculate the hash map once when inserting (or retrieving) an entry. There is no way that the time taken to calculate the hash code is going to make any impact on performance. The original poster didn't know about the embedded hash code at the time of writing the original code. It seems unlikely that the file format happened to get the same hash value. So that was a separate change. Given the likelihood of collisions with the first algorithm, that is a very good candidate for causing the increased performance. Tom Hawtin -- Unemployed English Java programmer http://jroller.com/page/tackline/ Back to top Chris Uppal Guest Posted: Sun Feb 12, 2006 12:12 pm    Post subject: Re: Why my java code is THAT slow compared too C++? Thomas Hawtin wrote: Quote: The original poster didn't know about the embedded hash code at the time of writing the original code. It seems unlikely that the file format happened to get the same hash value. So that was a separate change. Given the likelihood of collisions with the first algorithm, that is a very good candidate for causing the increased performance. There's also the marginal effect that the OP was probably interpreting the hash code strings as words and adding them to the table too. Not only would that double the number of words in the table, the words themselves would have a much reduced distribution (all decimal digits, or some such) which might further exacerbate any problem caused by weak hashing. -- chris Back to top JavaEnquirer Guest Posted: Sun Feb 12, 2006 12:12 pm    Post subject: Re: Generics and getClass Quote: The solution I've most commonly seen is to pass in an instance of Class<T>. E.g. String foo = makeMeThisObject(String.class); - Oliver Cheers Oliver, I feared that I'd have to do that, though, fear is too strong a word!!! Back to top Display posts from previous: All Posts1 Day7 Days2 Weeks1 Month3 Months6 Months1 Year Oldest FirstNewest First         AppletTalk.com Forum Index -> Java Language Programming All times are GMT Goto page Previous  1, 2, 3, 4 ... 328, 329, 330  Next Page 3 of 330   Jump to: Select a forum Java Newsgroups----------------Java DiscussionsWeb-based java games comp.lang.java 3D Graphics API's for Java Java: Support and criticism Java System Announcements comp.lang.java.api JavaBeans Java and CORBA Java and Databases comp.lang.java.developer Java GUI Toolkits Java Help comp.lang.java.javascript JVM, native methods and hardware comp.lang.java.misc Java Language Programming Security and Java comp.lang.java.setup Java Software Tools comp.lang.java.tech ibm.software.java.linux  Non-English Java Newsgroups----------------Java Language (German)Language Java (Français)Java (Italian)  You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum Powered by phpBB © 2001, 2006 phpBB Group SEO toolkit © 2004-2006 webmedic.