# starting desktopcouch
/usr/lib/desktopcouch/desktopcouch-get-port >/dev/null 2>&1 &

Unfortunately while syncing correctly the mode-bits where not set correctly so a little workaround was needed. My solution was pretty easy: I created a script in my $HOME/bin which starts the startup-script. It goes a little like this:


/bin/bash "$HOME/Ubuntu One/scripts/startup"

Making the script executable and adding the script to the startup manager finishes the job.
The only thing I’m missing is halting some programs on logout. I’ve been googling around but didn’t found a good answer. Anyone a suggestion?

I found the real solution at code.google.com. I got what I want by using jquery. So this is my current solution. On the html page we at a little javascript:

// Load jQuery
google.load("jquery", "1");
// on page load complete, fire off a jQuery json-p query
// against Google web search
google.setOnLoadCallback(function() {
$.getJSON("http://java3.tisnix.be:8888/jason.php?callback=?",
function (data) {
document.getElementById("txtResult").innerHTML = data.responseData;
document.getElementById("txtResult").style.backgroundColor = "green";
});
});

jason.php just returns a little json structure:

<?php
// Not sure of deze nodig is
header ("Content-Type: application/json; charset=utf-8");
// een heel eenvoudige response
$str = "";
$str .= $_GET['callback'];
$str .= "(";
$str .= "{responseData:'

Volgens de check van jouw browser staat mijn PC nu WEL online.

'}";
$str .= ")";
echo $str;
?>

On our page we have a div with id txtResult which get replaced by the responseData when the page is loaded.

I needed some code to make it possible to check if my computer at home was online. It’s all related to the stuff you can find in my java3 documentation. My original plan was using a little bit of AJAX which could check if my server was reachable and update a part of the page if so.

After some googling and putting stuff together I got a javascript like this


var xmlHttp;

}
function stateChanged()
{
if (xmlHttp.readyState==4) {
document.getElementById("txtResult").textContent=xmlHttp.responseText;
}
}


The code for GetXmlHttpObject has been copied from w3schools and looks like this


function GetXmlHttpObject()
{
var xmlHttp=null;
try
{
// Firefox, Opera 8.0+, Safari
xmlHttp=new XMLHttpRequest();
}
catch (e)
{
// Internet Explorer
try
{
xmlHttp=new ActiveXObject("Msxml2.XMLHTTP");
}
catch (e)
{
xmlHttp=new ActiveXObject("Microsoft.XMLHTTP");
}
}
return xmlHttp;
}


The code has been working but it got a little bug. I first tried it on my computer where it was working as it should. But porting it to the scarlet pages, my javascript debugger showed me some errors. With GetXmlHttpObject you can’t open pages on another domain. So I got stuck.

But there was a solution. The solution wasn’t vodka but putting the javascript on my local computer. So just adding this little script to my computer did the trick:

function showOnline(str) {
document.getElementById("txtOnline").textContent=str;
}


So the javascript page gets loaded from my computer. If it’s not online, you don’t get any annoying error unless you’re doing some debugging. If it gets loaded, the function replaces the element with id txtOnline with the argument you gave it. So I really had to keep it simple.

UPDATE: my test doesn’t work. I really have to do it asynchronously or the page hangs

UPDATE2: a working solution can be found in kiss part 2