Originally submitted by saimonmoore with a path included; please refer to http://code.google.com/p/xmpp4r-simple/issues/detail?id=6#makechanges

Simple usage:

im = Jabber::Simple.register(jid, passwd)

• will try and register user on the server
• if successfull, will reconnect and send default status presence
• if not successfull, throws a RegistrationError with a detailed message

self.register method signature:

self.register(jid, password, status = nil, status_message = "Available") #So you can still provide
custom status after registration

There are several problems with the approach to testing; admittedly, Jabber isn't an easy protocol to test, but I believe every single one of the features involved in xmpp4r-simple should be mocked to include no networking, and certainly no real Jabber accounts.

This would resolve intermittent errors due to network latency et al. It would also make writing further tests much simpler (and therefore more probable). Also, it will help with having to deal with side-effects of previous tests (if there are any) due to the external dependency.

Personally, it's painful and disheartening to sit there with the tests failing due to network problems, especially when they sit there for 20 seconds for each network-sensitive test.

This is directly related to:

http://code.google.com/p/xmpp4r-simple/issues/detail?id=7#makechanges

The culprit is start_deferred_delivery_thread with the following code:

def start_deferred_delivery_thread #:nodoc:
Thread.new {
  loop {
    messages = [queue(:pending_messages).pop].flatten
    messages.each do |message|
      if subscribed_to?(message[:to])
        deliver(message[:to], message[:message], message[:type])
      else
        queue(:pending_messages) << message
      end
    end
  }
}

end

The loop has absolutely no sleep time and will run at max speed, continually dequeueing and queueing until there is no more CPU power left, then it keeps going until nothing else runs on your computer (unless you've got god/monit monitoring it) until the subscription actually occurs.

In my production environment, this is asynchronous, and can occur even upwards of days later, and I'd rather not have my CPU raped for that long.

On http://code.google.com/p/xmpp4r-simple/ there is a pointer to the github repo. But the given url is: http://github.com/blaine/xmppr4-simple/ which is a 404.

There's a fix for 1.9 in master, but no gem. Can you release a new version? Currently you have to install this ghetto version https://rubygems.org/gems/scashin133-xmpp4r-simple

I've noticed that occasionally, in some situations, the tests that add clients to the roster, instead of adding [email protected], will add #jabber::simple:0x0xxxxxx as the JID. This is obviously wrong.

I was not able to pinpoint where exactly the problem was originating from (which tests, specifically).

I'm on xmpp4r-simple 0.8.8 (latest gem version) and xmpp4r 0.4.

What steps will reproduce the problem?

1. try to send a message to user/resource using a string or a jid object.

What is the expected output? What do you see instead?

I want it to only send to the specific resource.

What version of the product are you using? On what operating system?

Ubuntu Linux + 'ruby 1.8.7 (2008-08-11 patchlevel 72) [i486-linux]' + xmpp4r (0.5) + xmpp4r-simple (0.8.8)

The test test_add_users_to_our_roster_should_succeed_with_automatic_approval reliably fails around line 107 (assertion that client2 is subscribed to client1).

I'm not able to get this to pass.

I have xmpp4r-simple 0.8.8 and xmpp4r 0.4 installed.

ruby 2.3.1

gem install rcov

Building native extensions.  This could take a while...
ERROR:  Error installing rcov:
    ERROR: Failed to build gem native extension.

compiling 1.8/callsite.c
1.8/callsite.c:2:17: fatal error: env.h:

http://code.google.com/p/xmpp4r-simple/

"Please note that the source has moved to GitHub: http://github.com/blaine/xmppr4-simple/"

Also, you should ask http://xmpp.org/xmpp-software/libraries/ to link to the GitHub project page instead.