Comments for NewInstance

Comment on How to create launchers in OSX by using applescript to create a new terminal window to display log files » Zerohimselfs blog

Fri, 13 Jan 2012 02:25:18 +0000

[...] http://en.newinstance.it/2010/04/06/how-to-create-launchers-in-osx/ [...]

Comment on The following signatures were invalid: BADSIG 40976EAF437D05B5 Ubuntu Archive Automatic Signing Key by a

Tue, 10 Jan 2012 11:32:38 +0000

good!

Comment on The following signatures were invalid: BADSIG 40976EAF437D05B5 Ubuntu Archive Automatic Signing Key by Doru

Doru — Sat, 07 Jan 2012 11:11:38 +0000

For everyone still getting the same error after this procedure: please make sure that you try the manual way also:

wget http://repos.zend.com/zend.key -O- |apt-key add -

this will add the needed key to your lists. Hopefully it will help someone.

Comment on Working in Düsseldorf by Ade

Ade — Sun, 25 Dec 2011 10:44:40 +0000

Hi Great overview and info
I understand that DUS is a bike friendly city? Are there places to rent bikes for a period or would it make sense to buy one if one was to stay in DUS for let’s say 6 months- to a year?
Thanks Ade

Comment on Install Tutorial: Ubuntu 9.04, Apache with SSL, Subversion over HTTP / HTTPs, and Trac by faith

faith — Fri, 09 Dec 2011 04:51:21 +0000

hi luigi
i did chmod on the servers file n .subversion directory until it has full access permissions
but the error message still popped out..
any suggestion???

Comment on JAXB Tip: one line of code to marshall and unmarshall xml. by Luigi

Luigi — Tue, 06 Dec 2011 21:09:14 +0000

Awesome comment Tim, thanks for sharing.

Comment on JAXB Tip: one line of code to marshall and unmarshall xml. by Tim

Tim — Mon, 05 Dec 2011 23:25:53 +0000

In addition, in at least some implementations of JAXB, each call to create a JAXBContext object will result in the creation of an anonymous class that does not get garbage collected. So over time, you’ll run out of memory (PermGen, presumably, though I haven’t tried to reproduce the problem directly in order to confirm that). If you’re just running the occasional unmarshal operation, you won’t care about either of these things, but if you’re doing lots of unmarshal operations, then reusing a single JAXBContext will make a big difference in both respects.

References:
http://whileonefork.blogspot.com/2010/09/leaking-of-jaxb.html
http://stackoverflow.com/questions/3584821/jaxbcontext-newinstance-memory-leak

Comment on I/O Exception handling by blink4blog

blink4blog — Mon, 05 Dec 2011 03:28:24 +0000

I agree with the first method base on the assumptions of the following:

1. The exception is not recovrable
2. The exception is not expected to the catchable
3. There is no business value to catch the exception unless the logic can proceed without the IOStream presence.
4. The exception give technical information for system administrator or programmers.

My 2 cents..

Comment on I/O Exception handling by Luigi

Luigi — Tue, 29 Nov 2011 14:16:01 +0000

tbee, not true, I’d tipically let the IOException flow out of the method, since IOException most of the time can’t be recovered (i.e. just report the failure to the user with a proper message).

Before using try/catch[/finally], I’d think if I can actually do something to handle the exception when it is raised. I write much more often try/finally than try/catch[/finally].

Comment on I/O Exception handling by tbee

tbee — Tue, 29 Nov 2011 14:04:44 +0000

You’d have to have a second try-catch to deal with the IOException. I’ve improved the second pattern with a custom close method:

OutputStream out = null;
try {
out = new MyOutputStream(…); //may throw IOException
//use the stream
}
finally {
out = IOUtil.close(out);
}

The close only closes if the parameter null and always returns null.
BTW, the new Java 7 pattern is

try (
OutputStream out = new MyOutputStream(…); //may throw IOException
)
{
//use the stream
}
finally {
// auto closes
}