From: Robert Klemme <shortcutter@...>
Date: 2009-03-16T07:22:38+09:00
Subject: Re: Keep thread in memory

On 15.03.2009 22:35, Craig Williams wrote:
> [Note:  parts of this message were removed to make it a legal post.]
> 
> Hi,
> 
> I'm writing a simple UDP listening server. It works great using this
> 
> PORT = 1234
> server = UDPSocket.open
> server.bind("127.0.0.1", PORT)
> 
> server_thread = Thread.start(server) do |server|
>  100.times { #do something }
> end
> 
> server_thread.join

First of all, why are you using a thread here?  Do you have other work 
that needs to be done concurrently?

> The above was based off some code from the Ruby 1.9 book - I have a problem
> now trying to take it a step further.
> 
> How do I get the above to run permanently or until some other condition
> besides (100) is reached?

Depends on the condition.  If you want this to run indefinitely you can 
simply do

loop do
   ...
end

> When I run this on my remote server over SSH it works but if I ctrl-z or
> exit ssh, the process seems to be in memory but isn't actually working. I'd
> like to run it and have it go into memory until it is killed and return me
> to the ssh command.

"Go into memory"???  All programs are executed in memory.  What you 
probably rather need is a demon, i.e. a program that continues to work 
even if the terminal gets removed (ssh close).

> I tried leaving off the server_thread.join but then it just quits. Could
> someone explain exactly what Thread.start is doing here.

In your example above the thread is completely superfluous since you do 
not do anything concurrently.  A plain old infinite loop would do the 
same job.  Basically your code is equivalent to

PORT = 1234
server = UDPSocket.open
server.bind("127.0.0.1", PORT)

100.times { #do something }

Kind regards

	robert