Just to get started with this, I decided to show load stats and disk-space for my servers. Initially, the question was how to get this data from the servers to wherever I wanted to collect the data. The following command is my solution:

What this does, is that it sends the result of the uptime command and the df -h command as POST variables via cURL to a script that we define. The reason I want to use specifically those two commands is that uptime includes the load averages, and that df -h gives me the amount of free disk space. The command itself could then be run either manually or every minute using CRON.

The script that we access with the cURL request will contain whatever is needed to handle that data. At the same time, I am using the same script to display the data in Status Board. For example:

When I was coding this I was working from my iPad so I needed to keep it simple. For this reason, I chose to store the statistics in a file as JSON. Since I only needed to update the stats once a minute for each server, this worked just fine.

This way, I would generate an array based on the servers that contacted the script, and each of those would again hold the most recent data for that server. I identified the servers by IP in this case. The only thing missing now, is the part that generates the data for Status Board.

Now you can add this widget to Status Board by dragging in the table-widget using the url to script. It should give you a listing of the servers that called the script, showing a bar graph of how much hdd-space is used, and showing the load averages for the last minute. The table will auto-refresh every minute, so if you run the command from your servers every minute it will stay up to date.

For more info about the customisable widgets for Status Board, you can check the docs for graphs, tables and DIY-widgets. You can get the full code for this example on GitHub.

The post Server statistics in Status Board appeared first on Bensmann.

An error occurred while preparing the installation. Try running again

Now, if you haven’t used your Mac for a while, the error might be caused by the system date being incorrect. You can check this by going to utilities and opening the terminal. Once in the terminal enter the following command and hit return/enter:

date

The result of this command will be the date that the system currently has been set to. For some reason, it might have been reset to 2001, in which case we need to set it to right date. To do this, we enter a new command. This command will be entered as follows:

date {month}{day}{hour}{minute}{year}

Every bracket should be replaced with a two-digit number. For example, March 18th 2013 12:50 would become the following command:

date 0318125013

Enter the command and hit return. You can then check if it was set correctly by running the first command again.

If the date was wrong, it was likely that which caused the error, and after you quit the terminal it should able to install OS X Mountain Lion just fine.

The post Changing system date from Terminal – OS X recovery appeared first on Bensmann.

In short, this means that the company feels it won’t be able to build the browser it wants for the iOS platform according to CNET. Its not hard to understand why Mozilla feels this way. Not only would they, more or less, have to make a skinned version of Safari, but they would still have some major disadvantages. Firstly, the Safari app on the iPhone would have a performance advantage thanks to it’s faster Nitro JavaScript engine. Secondly, Apple still doesn’t allow users to select a default web browser. In other words, Firefox would have to take two steps back to begin with … but does that mean that they shouldn’t?

… there are a lot of reasons we should be on iOS even though we can’t bring our rendering engine there.

As a Chrome user I waited a long time for the browser to come to iOS, but it wasn’t because of the rendering engine or for faster load times. What I wanted was the synchronization that modern browsers offer and to be able to continue sessions across devices. This is why Firefox should invest the time in making an iOS browser. Its not about performance or rendering engines, its about accessibility and ease of use. Chrome understood this, and nailed it!

Unfortunately, I don’t think that Apple is going to allow Firefox to bring their own rendering engine. I wouldn’t expect them to do it, and frankly I don’t see this as a big problem. The UIWebView is plenty powerful, modern and fast. However, I am not a fan of Apple putting the brakes on other apps by keeping the Nitro engine to itself.

I naturally do understand the issue from Firefox’ point of view. Be that as it may, the real issue as I see it, is the fact that users can’t choose a default browser. For some users though, it would still be worth it, and they would have gotten a Firefox-experience on iOS with almost all the features that they would expect from a member of the Firefox browser family. Sadly, Mozilla doesn’t seem to see it this way.

The post Firefox won’t be on iOS without it’s own rendering engine appeared first on Bensmann.

The post Autocomplete security appeared first on Bensmann.

Saturday 26th of January WordCamp Norway 2013 will commence in Nydalen at the Radisson Hotel. Not only is this the first time I attend WordCamp, but I will also speaking at this year’s WordCamp Norway.

On this Saturday I will be holding a presentation on “Responsive design principles and techniques” based on experience, past projects and user experience considerations.

I am really looking forward to meeting the other attendees and speakers at this event. For those who are interested, tickets are available here. Will I see you there?

The post WordCamp Norway 2013 appeared first on Bensmann.

In my case I had a lot of information that I wanted to be included when searching. To keep things simple I simply decided to add all these value in a attribute, which I called “data-keywords”. In other words, it doesn’t search the rows, but it searches the attribute. The rows would look like this:

<tr data-keywords="moment.js moment.min.js date moment time">....</tr>

Now, the javascript code would need to do a couple of things. First off, it would have to update the table for each keystroke I entered in the search box. Secondly, it would have to filter the data by matching it against the “data-keywords” attribute. As a third criteria I wanted this script to constantly make sure to style odd and even rows, which meant adding and removing classes constantly. The code below shows you how I did this with only a few lines of code. (see the script in action here)

jQuery(function($){
    //index the dom elements
    var $list = $('#table > tr'),
        $search = $('#search-input');

    //filter the rows when searching
    $search.on('keyup',function(event){
        //get the search criteria
        var val = jQuery(this).val().toLowerCase();

        //show all rows if search is empty, or only matched rows
        if(val == "") $list.show();
        else $list.filter(':not([data-keywords*="'+val+'"])').hide().end()
                  .filter('[data-keywords*="'+val+'"]').show();

        //add and remove class for alternating rows
        $list.filter(':not(:visible:even)').removeClass('alternate').end()
             .filter(':visible:even').addClass('alternate');
    });
});

The post Searching a table with jQuery appeared first on Bensmann.

The video I posted on thomas.bensmann.no quickly passed 160K views on Vimeo, after being featured on several news-sites, and several people have since asked how it was made. Therefore I am posting the script that downloaded the images, which really is very simple.

<?php

//folder to store images in
$folder = "/files";

//url to latest webcam image
$url = "http://graphics8.nytimes.com/.../latest.jpg";

//get folder contents so far
$array = scandir(dirname(__FILE__) . $folder);

//i stored the images as 1.jpg, 2.jpg, etc.
//the first two elements found in the folder are "." and ".."
$num = (count($array) - 1);

//create the new filename
$filename = dirname(__FILE__) . $folder . "/" . $num . ".jpg";

//get the image
$ny_times_image = file_get_contents($url);

//save the image
file_put_contents( $filename , $ny_times_image );

Here is the cron job i added to create an image every minute:

* * * * * /usr/bin/php /var/www/labs.bensmann.no/html/ny_times_fetch.php

When the pictures were gathered I used an old version of QuickTime to put them together in a time lapse video.

Quick tip: Don’t forget to remove the cron job! My last image was called 116327.jpg, do the math.

The post Pulling webcam images with PHP appeared first on Bensmann.

The post PSB M4U 2 appeared first on Bensmann.

Meanwhile I am happy to report that my original one still works fine…

The post Jawbone Up returns appeared first on Bensmann.

Artikkelen kan du lese på Techstorm.no

The post Review av iPad mini appeared first on Bensmann.