heya folks, trying to install jdk with "sh jdk-6u7-linux-i586.bin", i get "jdk-6u7-linux-i586.bin: line 480: ./install.sfx.26481: No such file or directory" .... any ideas? google was not helpful

sounds like a trashed download.

hrm, i will try again. though i had 2 copies of the file with the exact same size

no wait, that was the x64 one....*tries again*

Hi, when I have "HashMap<GameMap,Integer> votedMaps", and I want to search for a key, votedMaps.containsKey(myGameMap), it doesn't find the key, but it does exist

I overrided the equals method in GameMap, and placed a break point inside of it, but it never reaches that when it does containskey

this totally doesn't make sense. why would even getString() throw that as an error.

dmlloyd:

~be dmlloyd

JDK5+ or piss off!

that's so old

hehe

BLEAH. well it's working, but all i'm doing is not touching the timestamp column if another condition is set. I'll hax0r on that later.

nope, wasn't a corrupt download

/me touches eidolon's timestamp.

aieee!

eidolon is updated to now()

'sup owen, how goes the fight?

Kill me now. Then hire my corpse.

mm, productivity in the afterlife.

FWIW, my 'No such file or directory' installation error was due to missing ia32-libs package

is there a way in a generic class to determine what type(s) someone inheriting from your class has specified for type variables?

why?

is there any Java Spanish Channel?

why is that EVERYTHING feels like a global in java? am I using it wrong? Everything is a pointer and there is no const: so how are you supposed to expose class internals without letting a user edit them in bad ways?

getter/setter methods.

not everything is a pointer.

and, yes, it sounds like you're doing it wrong.

I know about built in types...but that's a different story

what if a class you make is small but includes many huge classes? do you have to write getter/setter for all?

is it the only way?

hm

it depends on what you want to expose

I'm trying to think of a concrete example...

if i define an anonymous object -- Object foo = new Object() { public void bar() { } } - i get a compile warning saying the method bar() is never used locally... just out of curiousity, how would one use it locally?

you can't

cheeser, ok, say you have a class with a data array, you want to expose that array but not let people edit it.

that's the point of the warning :)

that doesn't really make sense to do that adante

unless you use bar() from inside a method that was specified on Object and overridden in your anon subclass

right

or use reflection

right, just figured that out, cheers

otherwise you're declaring a method that'll never get used

thanks.. been messing with groovy the past few days, i probably have a bit of an evil/dirty mindset at the moment

TheLorax: you would restrict the access of that array using just methods (ie: getAtIndex(int index)), or to be fancier, make your own immutable List

Algonquian, if you use getAtIndex and your array is non-built-in then you return pointers that can be changed anyway

not to mention, you add a function call

your notion of "can be changed" is hazy

whether or not an object is mutable is part of the contract of that object

e.g. String, BigInteger, etc are immutable

so returning copies of the reference is no big deal

if you don't want someone changing an object, you should not make that object changeable in the first place

failing that, use a wrapper object (Collections.unmodifiableList() and friends are a good example)

can anyone tell me the correct channel to ask a question about the SWT? Got an problem designing a GUI.

~swt

dmlloyd, swt is the Standard Widget Toolkit - see http://dev.eclipse.org/viewcvs/index.cgi/%7Echeckout%7E/.../main.html

no idea

well, String is *almost* immuatable

P^)=

that'S the problem. i'm the only on in there ;)

dmlloyd, the question is about creating objects of existing changeable objects but not letting the use of your new object edit the internals of your new object in a bad way. IE myobject.getinternal.unlockmutex.

the user*

creating objects of objects?

~tell TheLorax about google law of demeter

TheLorax, http://www.google.com/search?q=law+of+demeter

How can you change a string cheeser?

dmead, as in an object that has members of other objects

your sentence completely did not parse

dmlloyd, ^

like a collection you mean?

because java already has those

SJrX: myString.hashCode()

Collection, List, Set, etc

class myNewObject {private otherObject;};

SJrX: STrings are immutable.

svm_invictvs: almost

but for what most people care about, yes, it is.

You mean override the hashCode method on a string?

they are immutable in concept

SJrX: no, i mean call it.

cheeser: how arent' they? With reflections?

What will happen?

no, he means that hashCode can change the state of the object

oh

Really?

iow, it updates the cached hashCode field

Why can hashcode change the state of an object?

the hash code is lazy init'd

Ah

Oh

I see

8^)=

Ee

It only generates it and then caches it when is called

neat-o

cheeser: so it's like the mutable keyword in C++. There for "innoculous changes"

no

there's no compile time assertion of mutability or lack thereof in java

or runtime

But it's along those lines.

not really

Hmmmmm I have a ServerSocket and get a connection, when I look at the connection RemoteAddr() the address is always 0.0.0.0 I don't understand

the existence of "mutable" in C++ should give you an idea of how good an idea it is :)

how I can get the RemoteAddr()

class String { int hashCode = 0; ... public int hashCode() { if (!hashCode) { hashCode = calculateHashCode(); } return hashCode; } }

SJrX: you get it on the connected Socket, not the ServerSocket

Calculating and caching the hash code of the String would be an "innoculous change"

dmlloyd that's what it is on

what does "innoculous" mean

actually let me check that

protects you against polio?

I think it means small

innocuous

and inconsequential

Sorry

anyway an innocuous change is totally subjective

I can't spell. But the word, "innocuous" means, "not harmful or injurious"

dmlloyd, yes, and the programmer should be able to define what an innocuous change is.

I know what the word means, I was just needling you for your typo :)

thanks dmlloyd

~~svm_invictvs dmlloyd

svm_invictvs, dmlloyd is here to help. Well, not really.

~svm_invictvs

svm_invictvs, I have no idea what svm_invictvs is.

TheLorax: fine for them. There's still no such keyword in java.

you can, however, declare a field to be unchangable

using "final"

dmlloyd, will look up. thanks

dmlloyd: as long as that field is a primitive

dmlloyd you are correct thanks

the field itself is unchangable, not what it may point to :)

cheeser: er

what svm_invictvs said

final ConstInterfaceOfObject foo;

like the difference between const char* and const char* const.

;)

again let's not confuse references and objects

objects cannot be declared immutable

period

Next question, In my TCP/IP application, I have a Scanner on an input Stream. I need to implement timeouts. Do I a) (Have the master thread every so often see if the thread hasn't done anything, and if so interrupt it?, or is there a way, and if so how, to have my call to .nextLine() block only for a certain amount of time then have a timeout.

you can make them act that way, and final fields are a handy tool for that

dmlloyd: not as a compile time assertion. But certainly through an interface.

svm_invictvs: not really

at best, you'd have a convention for your project.

cheeser: How so?

SJrX: you can set the socket timeout

cheeser: What about the const interface idiom

what about it?

final ConstInterface foo...

i could still cast the reference to the underlying impl and call a setter.

Pretty much makes that object immutable.

nope

dmlloyd in theory I'm responsible for still sending traffic to the client, if it is not responding.

cheeser: As you could const_cast in c++

cheeser: or use reflection

If I detect a timeout I need an open socket connection

this isn't c++ so ...

SJrX: then you'll have to use another thread, or use non-blocking I/O

cheeser: but doing so would be an obvious abuse of the interface.

you could also look at xnio

~~SJrX xnio

SJrX, xnio is an NIO replacement framework, which keeps Channels but does away with Selectors, developed by ##java's very own dmlloyd. The project can be found at http://www.jboss.org/xnio - have fun!

cheeser: just like using a const_cast.

hi

Does anyone know where to get Logback support on IRC? #logback is empty for some reason (the authors used to wait there)

Cow_woC: google?

dmlloyd is this a drop in replacement for the existing stuff, or something seperate. This project is a learning project for the Java API's, etc... So I'm hesitant on trying non standard stuff.

it's standard to me :)

fair enough

you could use NIO directly

Hmmmmm

write your own wrapper InputStream for example, which throws an IOException if Selector.select() times out

but I guarantee you it will be a lot of work

otherwise, Thread.interrupt() is it

Hmmmmm

and that's a bit hairy too

Why is that hairy?

because now you've got multiple threads

Oh no I already had multiple threads

and if you think that concurrency is a simple thing, you don't understand concurrency :)

I think in this application it is fairly simple :P

dmlloyd: eh?

It's an SMTP server :P

dmlloyd: it's not as hard as people make it out to be *if* you code it properly

dmlloyd: You just have to be leet with concurrency :)

dmlloyd: the keyword there being *if*

um, an SMTP server probably wants XNIO or at least NIO if you care at all about scalability

Perhaps the worst thing is trying to retrofit multithreaded code into code that wasn't designed for it.

I don't care about scalability

No buffered input streams, and reading lines off of them from Sockets?

~~SJrX nio

SJrX, nio is http://java.sun.com/javase/6/docs/technotes/guides/io/example/index.html and http://gee.cs.oswego.edu/dl/cpjslides/nio.pdf

I stand by my statement

which statement?

<dmlloyd> and if you think that concurrency is a simple thing, you don't understand concurrency :)

Hmmmmm this is a learning project to replace my own personal mail server

well, make sure you've read this first:

It handles one user maybe 200 messages a day on a system with a load average of 0,0,0. Any complex concurrency stuff will be handled by MySQL

~~SJrX jcip

SJrX, jcip is Java Concurrency In Practice, a book focused on implementing threaded and concurrent applications in Java. http://jcip.net/

and/or this:

~jmm

dmlloyd, jmm is the Java Memory Model. See http://tinyurl.com/65v87y (JSR 133), http://tinyurl.com/5k6x27 (FAQ), http://tinyurl.com/ccx52 (Bill Pugh's JMM page), and http://tinyurl.com/26md5p (Implementation Guide for the extra-curious)

That's actually an excellent book. And if you can learn from that book, you can take the techniques in that book to other languages.

the FAQ in particular

Do you really think that that is necessary for this project?

I just read an RFC and about to read another

best to learn the right way first

><

my god damned Xbox keeps red-ringing ><

Fair enough, but I learn by doing, and no matter what I do for this project there will almost always be a book to read that will tell me how to do things better.

I wish I could figure out what causes that ><

well I hope your mail server doesn't have multiple CPUs in that case :)

at least read that FAQ

No it doesn't and if your refering to the fact that the locks have to be gained and released for information between threads to be syncronized. I know that.

This faq: http://tinyurl.com/5k6x27 ?

yes

was that what you were referring to with the multi-core thing?

yes, memory visibility is the #1 problem that appears on SMP when running a program that "works" on 1 CPU

but there are others as well, like race conditions due to assumptions about scheduling and so forth

Yeah but all that is done in MySQL

barriers!

all what is done in MySQL?

you said your app has multiple threads

Yeah

svm_invictvs: sorry for the delay, but that's a project convention and not enforceable by the VM

so... you have to arbitrate between them, regardless of the DB

It does and each thread really just reads in a piece of mail, and throws it into the DB.

also, that type of job is very well-suited to Executors as well

cheeser: barriers?

A connection comes in, the connection is completetly indepodent from the rest of the system at that point, no shared objects or memory, except possibly the socket in the master thread, but even that can be removed.

~javadoc Executor

dmlloyd, please see java.util.concurrent.Executor: http://java.sun.com/javase/6/docs/api/java/util/concurrent/Executor.html

svm_invictvs: your const interface bit

cheeser: oh, yeah.

~javadoc Executors

dmlloyd, please see java.util.concurrent.Executors: http://java.sun.com/javase/6/docs/api/java/util/concurrent/Executors.html

both of those you'll find handy

Thanks

~wiki

the ##java channel wiki lives at: http://www.javachannel.net

more red rings.

><

If I am writing an object that I want to make multiple instances of, can it have static variables in it?

~~antwaungomez static

antwaungomez, static is http://java.sun.com/docs/books/tutorial/java/javaOO/classvars.html and also see http://mindprod.com/jgloss/static.html

sure

Algonquian: Ok, I have run into a problem, I just tested it out and realized that, if I have multiple instances of it, the static variables are the same for all of them. I don't want that. Unfortunately, I cannot call the paintComponents function from awt when accessing dynamic variables. What do I do?

fix your code to work with instance variables... it sounds like you're accessing them directly too, you should look at setter/getter methods

you can pastebin example code

~pastebin

http://pastebin.stonekeep.com

I've got a keylistener and I want it to pick up key events no matter where they happen. What's the best way to make that happen?

Algonquian, thank you so much. You inspired me. I think I have it under control now.

sheesh

in LinkedBlockingQueue, is there any way to do a peek() which waits if the list is empty :)

It doesn't do that already?

Why is it called a "blocking" queue, then?

hey guys, is there a possibility to calculate the width of a string if it is been printed by drawString(...) ?

Oh yes, there is. You use FontMetrics.

lilwik: ahh .. nice thx :)

another short question .. do anyone knows how to reduce the used characters for a double number i.e. using scientific notation and rounding the coeffizient

?

aehh .. of course if i want to print the double value as a strring :)

~~sox javadoc DecimalFormat

The user sox is not on ##java

~~s0x javadoc DecimalFormat

s0x, please see java.text.DecimalFormat: http://java.sun.com/javase/6/docs/api/java/text/DecimalFormat.html

ups ... forget my question ^^

Anyone have a handy Email Address Validation Class for Java, I have been poring over google but the ones I have found are EE.

Ive got one somewhere I wrote validates the mx server as well

Or were you just looking for a regex to validate the email?

mornin`

~email regex

r0bby, email regex is http://www.ex-parrot.com/~pdw/Mail-RFC822-Address.html

Both would be cool

Xeese

SJrX: rip that regex :)

afaik the actual official RFC 822, is actually not followed and allows things like comments in email addresses etc...

Doesn't commons have something

Doesn't commons?

errr commons?

No I mean in email addresses you can have comments

~commons

Apache Commons is a set of libraries built for reusability and filling the gaps in the standard library. It can be found at http://commons.apache.org/

I don't have my book here, but I think it's in PHP Security, that it talks about validating email addresses, and that the actual specification isn't really real world usage

SJrX: so weed out what you dont need from that regex

:)

Amazingly Regexp is a skill I have been able to avoid learning for 4 years now :P

http://www.white-hat-web-design.co.uk/articles/js-validation.php

steal the regex

~next

Another satisfied customer. Next!

Hehe.. regexp is fairly straight forward until capturing groups

pgib: pfft if you know what you're parsing it's piss easy

I mean in a brain-warping sense. since regexs are such shorthand

I hate string parsing :P

oh god

I think probably everyone does, but I will learn it eventually

http://forums.sun.com/thread.jspa?threadID=5165170&messageID=9631380

pgib: it was at first

r0bby so just take the other link you just posted, and that link and replace the lowered.matches with the other regexp and that's it?

but once i learned what they were I figured out how to read them :)

SJrX: just take that regex, better yet do email.matches(yourEmailRegex)

it's as easy as that

Yeah, I guess. I don't really use them _too_ often, so, there is usually some relearning required, for me, when I want to do something complex

the one in the js-validation one right

yes

i also found this http://www.codetoad.com/asp_email_reg_exp.asp

http://saloon.javaranch.com/.../ultimatebb.cgi?ubb=get_topic&f=1&t=020629

you didnt look hard enough they gave you solutions

Sorry I saw some regexp ones but didn't know how to use them.

heh. I wonder how much all this stuff is gonna change since IANAL is opening up every possible top-level domain

I mean: how many apps will break

lol, not IANAL

SJrX: pussy

regex is powerful

:)

and useful for this exact case :_

:)

ICANN

Yeah I know it is

hmmmm even after I escape the javascript one, my own email address fails the check

do I have to escape more than just the \ to \\

I don't think, unless it is in a string. The \ is the delimiter

return address.matches("/^([A-Za-z0-9_\\-\\.])+\\@([A-Za-z0-9_\\-\\.])+\\.([A-Za-z]{2,4})$/");

you have to double escape regex in java

So

yeah, but he said jscript?

Which is not java...

return address.matches("/^([A-Za-z0-9_\\\\-\\\\.])+\\\\@([A-Za-z0-9_\\\\-\\\\.])+\\\\.([A-Za-z]{2,4})$/");

you shouldn't rely on JS dude

I can disable js and you'd be screwed

um

<r0bby> http://www.white-hat-web-design.co.uk/articles/js-validation.php

<r0bby> steal the regex

I agree, always need SS validation as well

steal the _REGEX_

I thought he was testing the regex in JS

not the code itself...

my misunderstanding

I am stealing the REGEX

and it's not working

take out the starting and trailing "/"

hmmmm well that worked for my email address, was that a guess or did you know that

SJrX just wants to know if he should test if other things pass/fail

SJrX: grab beanshell

hmmmm it's working

what's beanshell?

~beanshell

r0bby, beanshell is a small, free, embeddable, Java source interpreter with object scripting language features, written in Java. It can be found at http://www.beanshell.org.

i use groovyConsole for testing regexps

ah. I use python

thanks

SJrX

use javamail

http://java.sun.com/products/javamail/javadocs/.../InternetAddress.html#validate()

:>

http://java.sun.com/products/javamail/

Isn't that EE?

fubar!harvard!uunet.uu.net!overload!user

SJrX: it's an api..

you can use it in desktop

Isn't that API only available in EE?

joed: mail.jar is the javamail api correct?

When I hold down a key, why do I get an endless stream of KeyEvents for pressing it? Shouldn't I get just one until I let go?

It is only that jar yeah

SJrX: include that in your classpath

use the method i gave you

~next

Another satisfied customer. Next!

I'm trying to create a class that records which keys are being pressed, but it's not working.

oh god

Ah

lol

thanks r0bby

this smells of malice

should I use the JavaMail API or the connections API you gave earlier for SMTP sending?

JavaMail

I don't mean something that keeps a long-term record, I just want to be able to look up if a certain key is down or up at the moment.

so?

Java's event-based system is fighting me. I'm getting an endless stream of key events when I hold down a key.

no shit sherlock

:P

Well, that's wrong isn't it?

btw of 11605 emails 11436 passed with that regex

no, it's probably catching your events

by the way*

keydown

keyListner?

Oh whoa, I think it just started working! I guess I was doing it right after all.

When I stop my program from executing via the STOP button in eclipse or the OS requests that it shutsdown does my application throw a InterruptedException where this happens?

or is that just internal for threads

sup guys

what am I doing wrong here? System.out.println((1-(investHist.getVolume()/storageUnit.getAvgVol()))*100); investHist.getVolume() is double = 236000 .... and storageUnit.getAvgVol() is double = 400120

A lock object just has to reference the same place in all locations correct, it doesn't have to be a member of the object?

but I get result = 0

i.e. I'd like to have a concurrency lock on a Set, there

Do I need to extend the Set class and have a lock object or can I put any ol' lock object there

trend: you are not doing anything wrong, you get the expected results

trend: int division

i'm sorry , what do you meanby that?

trend: convert some of the values to double

SJrX, there's no such thing as a lock object.

ReentrantLock

ohhh, ok.. thank you :)

trend 4/3 is 0

SJrX: no, it is 1

1.333333

heh

SJrX, different thing. So, you want to hold a lock in a compound action of a Set?

errr whoops

Yeah

SJrX, or a single operation on the set?

:) heh

Hmmmmm

trend: 1/2 = 0, 1.0/2 = 0.5

well not sure, I've been away for 3 months.

I have one thread that is adding stuff to the set, every so often the other thread will wake up and iterate over members of the set.

ohhh, I see what you are saying now

Adding and removing stuff to the set, happens randomly

SJrX, use any use a CopyOnWriteArraySet

SJrX, Concurrent collections doesn't have fail-fast iterators anymore

fail-fast?

hmmm

SJrX, have you read the documentation on iterations and iteration over a collection while shared on different threads>

double avgVolP = (1.00000 - (investHist.getVolume()/storageUnit.getAvgVol()))*100.0000; should work?

actually .. both of the vars are int s

no I haven't

SJrX, if you're not using a concurrent collection, then you have to hold a lock before you enter to the iteration.

Yeah that's what I was thinking

SJrX, you make sure you lock that region and that only one thread a time does that.

but I think this CopyOnWriteArraySet() may do what I want

SJrX, but seriously you want a CopyOnWriteArraySet

So with this, I don't have to do anything special really

SJrX, define "anything special".

SJrX, if that's the only thin you wanna do, that's it. But read the docs, make sure the implementation doeos what you need.

Hmmmmm

The sample code there is of no help

SJrX, huh?

the basic operations seem to imply it does what I want.

http://java.sun.com/j2se/1.5.0/docs/api/java/util/.../CopyOnWriteArraySet.html

So from what I can tell, whenever one thread adds something to the set it will be slow (ok), all instances will be updated. When I grab an iterator off of it, it will then have a copy at the time it was created.

crap.. no go w/: double avgVolP = (1.00000 - (investHist.getVolume()/storageUnit.getAvgVol()))*100.0000; I get 0.0 as the result

Even if the other thread removes an element from the set, it will still be in the iterated version, and hence will always exist

it will be slow?

that's what it says, well expensive

hmmmm

it's cheaper than doing a Collections.synchronizedSet(set)

since it doesn't work with synchronized, it uses a CAS algorithm.

SJrX if that's not your solution, then rely on what I posted above

SJrX, and synchronize over the collection before you iterate

Hmmmmm yeah I think that works

I'm just trying to get it in my head

SJrX, it's easy, thinkg about "serial access"

just as a sequence :)

My previous question that no one asked, when my program terminates from CTRL+C, hitting STOP or the OS, is that just an IterruptedException getting tossed in my program somewhere, or is that just for threads.

I'd like to do some shutdown cleanup

anyone know what's up w/ my math problem?

System.out.println((1.0000-(investHist.getVolume()/storageUnit.getAvgVol()))*100.00000); gets 0.0

SJrX, InterruptedException is just that, whenever a Thread is interrupted by either... (thread termination, thread abort, a timer)

SJrX, Java doesn't support signaling at all.

amnesiac but is that true for the 'main' portion of my app

SJrX, you can do shutdown cleanup, take a look at ExecutorService

SJrX, instead of using plain threads use an Executor.

SJrX, that will allow you to have a lifecycle for your multithreaded application. and do cleanup at shutdown.

No I mean for the application itself

SJrX, yes, for the application itself.

Okay

java.math.BigInteger( << is that what I need?

hmmmmm

what are the actual values for getVolume() and getAvgVol()

Signal and SignalHandler....

trend, what type is getVolume and getAvgVol?

joed, is more like a shutdown hook.

it was whole numbers.. 234000 or so and 420000

I can get the exact ones.. will have to run the prog for a bit

trend, on integers, Java promotes arithmetic operations to that type...

I need to do this? System.out.println((1.0000-((double)investHist.getVolume()/(double)storageUnit.getAvgVol()))*100.00000);

trend, make sure that... whle doing arithmetic operations, you operate over values from the same type.

amnesiac: To catch clean vs unclean exits, yes, I'm saying you can get at signal handling.

:? ok

can I add to an int[]

ultravi01, eh?

i thought you could expand it

Can you resize an array - no.

i created a new int[] and I want to iterate through one array and put the values in my new int[]

ultravi01, do it by their index, considering that they're both of the same size.

Use collections.

or your new one is as big as possible to hodl all the values from the other array.

ok, thank you

/mode +b *!*arrays

double avgVolP = (((double)investHist.getVolume()/(double)storageUnit.getAvgVol())-1.00000)*100.0000; did the trick :)

thanks for the input

heh

E> arrays

~~ultravi01 javadoc System.arraycopy(*)

ultravi01, please see http://java.sun.com/javase/6/docs/api/java/lang/System.html#arraycopy(java.lang.Object,%20int,%20java.lang.Object,%20int,%20int)

thanks xeese

how can I cast 54.001230123 into 54.00 ?

that's not a cast

sorry, bad verbage

?

trend, just use a formatter\

ahh, yeah.. couldn't remember

nice :)

hey all

is there any way to loop over all the public fields in a class?

yes

lol I just spent the past two hours having a thread master, when it turns out I can't leave my thread in a blocking state anyway for timeouts since I have to sleep to interrupt, argh.

dangertools, how?

SJrX, what?

been googling but can't seem to achieve this simple task :(

look at myObj.getClass().getFields()...

I have thread A sitting in in.nextLine(), it's blocking. I needed to have it timeout after 5 minutes. So I set about having another thread interrupt it. However that won't work since an interrupt only ever happens during sleep(), so I need to just check if there is a new line first and sleep if not

seraph i use that to rewrite source code from its compiled form :)

seraph: Object.getClass().getDeclared*()

xjrn, ooer

I'll try that out

~tell seraph about reflection

seraph, reflecti