Introduction to Windows 10

We heard it first at the Microsoft’s BUILD conference in April 2014. The new version of Windows is the seventeenth version of Microsoft’s desktop operating system, and the ninth version of the business-focused NT kernel. And, it is not named Windows 9. Instead, they choose an even number to surprise and stir the interest of a questioning clientele.

Windows 8 and even Windows 8.1 were not the best favorites of consumers who loved Windows XP and Windows 7 because of its focus on touch-friendly user interface. Some desktop users need to have some getting used to moments before they could full enjoy the system, especially those who are used to the friendly START button of the previous releases.

In the business arena, most companies skipped the route to Windows 8 and 8.1. Large businesses stayed with Windows 7 or upgraded to Windows 7 from Windows XP during the launch of Windows 8. This follows the pattern of waiting for the second release before upgrading, such as what happened to Windows Vista.

Windows XP plus Windows 7 = Windows 10?

Windows Vista and Windows 8 are both major updates to the Microsoft OS but they share the same problem, it is not just appealing to users who are loyal fans of the speed, usability, and reliability of their previous versions.

Vista was perceived be slow and resource-hungry compared to Windows XP, and Windows 8 was seen to be a struggle for those who are used to the desktop environment.

With Windows 10, it is not just the problem of Windows 8 but trying to improve Windows in general, looking at the way people love it in XP and 7. Microsoft looked closely in detail at what people liked most about Windows XP and Windows 7 and will use that to bring back or preserve the experience they have been enjoying with Windows. With any company who made a great success on their products, retaining the sellable stuff will always be a good path to take.

Why Windows 10 not 9?

Why not Windows 9? People are intrigued with the jump from 8.1 to 10. This new operating system is a major update and is anticipated to the last major version of the Windows OS. This doesn’t mean it will only stay for quite some time, but rather, like Apple did with their OS X desktop operating system, there will be no major updates after this.

For such a major release, and for the final major release of Windows, the number 9 just didn’t sound major enough, and, frankly, it would have been an odd number to finish on.

Glimpse on its Specification

Minimum hardware specification is the first thing people will look on to judge the usability and efficiency of an OS. It simply means when an OS could run with low system requirements, it is efficient enough. The usual 1GHz, 1GB RAM for 32-bit and 2GB RAM for 64-bit versions, 16GB free storage space, and graphics that supports DirectX 9 is the least hardware you can run Windows 10 with. However, you need to keep an eye on your processor. It should support PAE, NX, and SSE2 standards. This means you need a slightly modern one.

Cross-Platform

While Windows 8 is best for tablets and phones with touch-screen technology, Windows 10 is friendly with all platforms. It could run on desktops, laptops, and tablets down to phones, the Xbox One, and Internet of Things (IoT) devices, and upwards to their new Surface Hub and Holographic headset, HoloLens. We also heard that it could run on a Raspberry Pi.

How to Test it?

Consumer version will be out on July 2015 but for now you can have the Technical Preview release.

How to install Zentyal, configure, manage and set up backups for a gateway server

There’s no shortage of enterprise-grade open source applications for setting up everything from firewalls and web servers to creating virtual private networks. But setting up and configuring gateway servers requires time and effort, which is where Linux distro Zentyal comes into play. Unlike a regular desktop distro, Zentyal is designed as a one-stop server for small/home office users. You can roll it out as a unified threat manager, office server, communication server and more, thanks to the fact the distro bundles some of the best open source server tools, including a directory server, DNS server, IM Server, groupware, VoIP and content control management. That's a good package for you.

It also has a great collection of custom management tools for setting up, controlling and monitoring the various components. Although configuring these network services does require a certain level of familiarity with networking, Zentyal does its best to ease you through the process.

Install Zentyal

A good server distro must be easy to deploy, set up and manage, which is why Zentyal releases are based on the Ubuntu Server Long Term Support (LTS) releases. The current stable release, Zentyal 4.0, is based on Ubuntu Server 14.04 LTS. Thanks to its Ubuntu underpinnings, installing it is simple – it’s almost completely automated, as the distro is designed to take over the entire disk. However, you can do an Expert installation to manually partition the disk. You can also install the Zentyal server and its components on top of an existing Ubuntu server installation. Head to www.zentyal. org/server, [or the LXFDVD]download the free community edition and burn the ISO image to a disc. Boot from the disc and select the default installation, unless you have a server with RAID or want to define your partitioning layout – in which case select Expert mode (see image 1).

During installation, you’ll only be prompted for basic requirements, such as your location and keyboard layout. When prompted to select a network card, choose the one connected to the internet, not your internal network. Also enter the login details of the admin user. The installer will transfer the core modules to your hard disk. When it reboots, the system automatically launches a desktop session with a browser to configure the server.

Initial configuration

When the distro boots for the first time, it will install some core packages by downloading them from the internet, so make sure it’s connected. You’ll be logged into Zentyal’s sparse desktop, which has an open browser window to Zentyal’s web-based administration console. Log into the administration section with the admin details you created during installation.

Zentyal will then fire up a configuration wizard. From here you can install various servers and services, such as DHCP Server, Firewall, Domain Controller and Jabber. The server will automatically resolve any dependencies, inform you about the additional components it will install and then fetch them all from the internet. You can skip this step and install them later if you prefer.

One important part of this initial configuration wizard is the networking step. Typically, a Zentyal server will have multiple network interfaces – one pointing to the internet and the other pointed towards the internal intranet. In this step Zentyal shows you all the available network interfaces on your server and asks you to configure them. You’ll first have to mark all available interfaces as either External (connected to the internet) or Internal (connected to the intranet). This will impact several settings, such as the default policies for the firewall, and sets up the default listen interfaces for other modules. Next, you’ll get the option to configure these interfaces by manually specifying an IP address or asking the server to use DHCP.

The closing steps in this initial configuration wizard will vary depending on the servers you selected, e.g. if you select the Domain Controller and File Sharing server, Zentyal will prompt you to select the type of server as well as the domain name of the server. Similarly, if you’ve selected the Mail and Groupware server you’ll be asked to specify the domain name for your mail account. Again, you can skip configuring these servers for now and set them up later.

Accessing the dashboard

When you’ve finished with the initial configuration wizard, Zentyal will take you to the Dashboard. You can administer and monitor the Zentyal installation from here. In addition to Zentyal’s minimalistic LXDE-based desktop environment, you can also access this Dashboard remotely from any computer on the same network as the Zentyal server. To access it, simply enter the Zentyal server’s IP address followed by the port it’s running on, so for example https://192.168.2.5:8443. Dashboard can only be accessed through SecureHTTP and runs on port 8443 by default. Also note that since the administration portal is served via the HTTPS protocol, your browser will initially ask you to add a security exception (which you should), as the server lacks a signed certificate.

From the Dashboard you can watch various server components, such as the CPU load, plus the status of all the installed components. From here you can also install any available updates to the underlying core Zentyal distro. The navigation bar on the left of the Dashboard will list the various installed modules, as you add them.

Set up a jabber IM server

You’ll want to keep communications under your control, so to configure an instant messaging server, make sure you’ve installed the Jabber component. Head to Software Management > Zentyal Components and click the View basic mode link. Select Jabber under the Additional services page, scroll down and click the Install button. Zentyal will show you a list of additional dependency modules that need to be installed. Click Continue to install them all.

Next, head to Module Status, tick the Jabber box to enable the Jabber module and click the Save button to activate it. To configure the service, go to Jabber in the left-hand menu and set the parameters for the server. Make sure you enter a Jabber domain – you’ll then have to pass this on to your users so they can connect to this server. Feel free to enable/disable the other settings on this page. It’s advisable to use the Allow SSL option to make sure the chats are encrypted.

Head to Users and Computers > Manage, which displays a list of Groups and Users on the Zentyal server. Click the [+] button to add a new user. You can similarly add Groups and associate users with them. Click a username, scroll down to the Modules configuration section and use the pull-down menu to Enable Jabber for this user. You can also tick the box to give this user admin privileges if you want to.

That’s all there is to it – but before your users can begin chatting with each other they’ll need to configure their Jabber client. Head to https://xmpp.org for a list of supported Jabber clients. The exact configuration steps vary for each client. Make sure you select Jabber or XMPP as the protocol. Then enter your username and password in the appropriate boxes, along with the domain name you specified when you were installing Zentyal.

All users will have to follow the same process. Once they’ve authenticated with the Jabber server, they can add each other and start communicating. Users can also change their password and other details. The admin user gets some special powers and can see which users are connected to the server, set the message of the day, and broadcast a notice to all connected users.

Keeping your server ship-shape

You can monitor the health of the server from the main screen of the Dashboard. From here you can get general information about the server, such as its load average, uptime and number of connected users. Other sections give you details about the running status of the enabled modules and visualise network traffic. Head to the Logs section to configure and view detailed reports about all the enabled modules.

To share your duties with another user, head to System > General and click the Add New button inside the Administrator Accounts section. On the following page you can define the login credentials of the new admin user. Other settings include changing the hostname and domain name of the server, as well as changing the default port for the dashboard from 8443 to something else.

Zentyal can also save your server’s configuration. Head to System > Configuration Backup, which enables you to save the settings either on Zentyal’s cloud server or on your PC. For cloud backup, enter an email address and a password. Once set up, Zentyal automatically backs up settings daily. To restore them, browse the list of backups and use the Restore button adjacent to the configuration you wish to restore.

It’s also important to keep your system updated. The Dashboard lists the number of system and security updates available under the General Information section. Click the number listed there or head to Software Management > System Updates for more details. Security updates have a red shield in the component name. To update them all, scroll down the page and tick the Update all packages option–and there you have it, your gateway server is up and rolling.

Thermaltake Commander FT Review and Test

If you’re like most builders and gamers, it’s a safe bet that you spend more time picking components based on price and performance than determining how each one will impact your system’s overall cooling profile. And that’s okay, as long as you eventually take an active role in managing your temperatures. One of the best ways to do so is to install a fan controller, and Thermaltake’s Commander FT is an option with tons of extras. 

The Commander FT fits into a single 5.25-inch drive bay and features a unique 5.5-inch touchscreen surrounded by a narrow black plastic bezel. The backlit display features blue, green, and red illumination, which manages to convey a lot of information without looking busy or gaudy. It’s also very intuitive to operate. For instance, in manual mode, press one of the numbers on the right to select the desired channel, and that channel number appears in the Commander FT’s central hub, after which you can move your finger across the fan speed slider on the left to lock in your desired speed on a zero to 100% scale.

Touch the Auto icon to let the fan controller rely on temperature data gathered from the included thermal probe. There are two icons below the fan speed slider that look like a trio of dip switches in the off and on positions. These correspond to silent (fans running at their lowest speed) and performance (fans running full tilt) operating modes. Each of the channels supports up to 10 watts per fan. Although the five fan connectors support 4-pin fans, the PWM function isn’t supported on this controller; the Commander FT relies instead on temperature control or your manually input settings.

When you select any given channel using the numbers on the right, you’ll also see that fan’s current RPM as well as the temperature being registered by the thermal probe. The Commander FT also has a temperature warning, which you can adjust, to help you balance temperatures and noise output. Other features include built-in short circuit protection and support for temperature display in either Celsius or Fahrenheit.

If you’re looking for an affordable, attractive 5-channel fan speed controller, then the Commander FT is a solid choice. The touchscreen also adds a coolness factor that few other fan controllers can touch (pun not intended).

Aerocool DS200 Window Edition Review and Test

Aerocool certainly knows its way around PC tower cases, but even for a company that has as many hits on its hands as this one does (see also the GT-S, the Strike-X Air, the XPredator X3 and XPredator Cube, et al.), the DS200 is a fantastic enclosure. 

For starters, the DS200 Window Edition (we received the black variant for this review) is swathed in the soft-touch material that feels like some kind of space-age rubber/leather hybrid. The stuff coats the top, front, and bottom. It frames the sides, as well; only the removable side panels are solely made of steel and acrylic. The one drawback to this material is that it tends to be pretty eager to show fingerprints, but the counterpoint is that it’s easy to wipe the case down and quickly make them disappear.

The exterior also looks pretty smooth, thanks to clean, fluid lines and a front panel that eschews the usual mesh ventilation in favor of a recessed, slotted vent strip that runs along the sides and around the bottom. This is a unique design element, and the effect on the DS200’s overall look is profound. To top it all off, Aerocool built a round LED temperature display and fan control unit into the center of the top panel near the front. The display shows the temp of whatever component you’ve attached the included sensor to (usually the CPU) in either Celsius or Fahrenheit, and the three-channel fan controller’s buttons let you select one of four settings for all connected fans. (The first setting is Off.) Builders who are serious about getting all lights to match will appreciate that the readout can display your info in one of seven colors: red, green, yellow, blue, purple, light blue, or white. You can also turn it off if you prefer a PC that’s not seen and not heard.

Another nice feature related to the DS200’s top panel is that removing it reveals enough space above the inner metal frame to mount a 240mm or 280mm radiator. (The case also comes with an alternate steel mesh top panel for those who like to maximize airflow.) This is cool because you can mount the push/exhaust fans just beneath the frame, but the radiator itself doesn’t cramp the case’s interior, which by the way comes with one 140mm intake fan behind the front panel and a single 120mm unit to push hot air out the back. You can add a second 140mm or 120mm fan behind the front panel to increase overall airflow.

A look around the rest of the interior reveals all of the amenities you’d expect, including multiple tool-less internal drive bays (some of which are removable), space for graphics cards up to 11.4 inches long (or 16.1 inches with the middle drive cage removed), a motherboard tray that’s ready for cable management and convenient CPU cooler installation, a PSU mount with rubber pads to reduce vibration, and so on.

If you like the look of the DS200 but think it needs a little more spice, you’re in luck: As you can guess from the pictures above, Aerocool also offers the case in red. The red version has the same soft-touch coating, the same side panel window, and the same price tag as the black one that we reviewed. Regardless of which color you choose, the Aerocool DS200 will provide you with an attractive, builder-friendly home for your next build.

How to Build an arcade cabinet with Raspberry Pi 2

In the 1980s and 90s there was an explosion of gaming consoles and home computers. Computer users of a certain age will have fond memories of machines such as the ZX80, BBC Micro and Megadrive/Genesis. These machines are now quite rare and this is where emulation plays a big part in helping re-live those glory days. 

The Raspberry Pi is a popular choice for emulation because it’s powerful enough to run emulators from Atari to upright arcade cabinets using MAME. The release of the Raspberry Pi 2 has also meant we now have an even more powerful processor that will enable emulators for later consoles, such as the PlayStation One, to run more smoothly. We shall be using the RetroPie emulator (http://bit.ly/ RetroPieProject). You’ll need to copy the downloaded image to a blank high capacity SD card, the larger the better, and you’ll need to use the dd terminal command to copy the image to an SD card.

With Retropie on your SD card connect all of your peripherals, including any USB gamepads that you wish to use, to your Raspberry Pi and power up. On first boot RetroPie will ask if you would like to configure your joypad to work with the user interface. Note: There’s a separate configuration for each of the emulators due to their differing joypad layouts. If you have a wired Xbox 360 controller there’s a great guide on the RetroPie GitHub (http://bit.ly/ RetroPieXbox360SetUp) for helping set it up.

RetroPie uses the popular Emulation Station as its user interface and by default there are a number of systems that can be emulated out of the box. These systems are identifiable by being active on the user interface. For each system there’s a games library beneath it and this is created by inserting a USB drive into your Pi. A script creates the necessary file structure on the drive. Insert this drive into your PC and copy the ROMs from your computer into their corresponding folder on the USB drive. Now return the stick to your Raspberry Pi and RetroPie will automatically copy the contents to your install. By copying ROMs in this way their corresponding emulator is enabled in the Emulation Station user interface.

At this point, we need to advise you that ROMs are copyright material and remain the property of their owners. Their use in emulation is a grey area and many games are now so old that they are no longer commercially available; however, this doesn’t mean that they are out of copyright.

With your ROMs installed go back to the user interface and select which system you would like to play. RetroPie will now ask if you would like to catalogue your games, before accepting, ensure that your Pi is connected via Ethernet to your router. RetroPie will search the internet for your games and download any box art and information about each title. Where there’s a conflict it will ask you to choose which is correct, or if nothing is found it will ask you to manually enter the names of the games. RetroPie can emulate a number of systems, such as the Commodore Amiga, Sega Genesis and Super Nintendo, and on the Pi 2 there’s no need to overclock the system as it runs six times faster than its predecessor. If you are using an original Raspberry Pi then you can streamline your setup using the Advanced Configuration steps on the RetroPie wiki.

NOTE:

There were many different games consoles and computers from the 1970s to early 2000s. Companies such as Commodore, Sinclair, Acorn, Dragon, Sega, Atari all competed for home computing dominance. This drove the rise of bedroom programmers; people who learnt to code from magazines of the time.

In each magazine there would be pages of BASIC code to type into your computer and learn coding via experience. This lead to a boom in the number of games being produced in the UK, and fuelled a generation of coders who are now hoping to reignite that spark with the Raspberry Pi. David Braben, for example, co-wrote the iconic Elite game in the 1980s. Elite was an entire galaxy full of planets, space stations and pirates in only 32Kb of memory. David Braben also happens to be one of the co-founders of the Raspberry Pi Foundation and is very keen to see children creating and learning with computers, rather than simply consuming content.

Build a SmartTV using your Raspberry Pi 2

We have many ways to consume media these days but our main method is still through our TV (even it might be via streaming technology). The Raspberry Pi is well suited to displaying media and since the day of its release users around the world have been creating media centers based on the Pi. In this project, we’ll show you how to do it for yourself. 

We start this project by copying the operating system to the SD card. For this we’ll need to extract the contents of the file downloaded from the OpenELEC website. This will extract an image file, which we need to copy to the SD card. To copy the image to our SD card follow the guidance at their website. Note: For whatever version of Raspberry Pi that you use, ensure that you have the correct image file as Pi 1 and 2 images are currently not compatible.

With the image copied to the SD card, insert it into your Raspberry Pi along with your wireless keyboard and mouse and other connections, such as Ethernet and HDMI to your TV. The last connection to make is the power which will boot your Raspberry Pi.

On your first boot up OpenELEC will lead you through a series of screens that will set up your media center. Once complete you’ll be returned to the main menu which is split into sections to catalogue your pictures, videos and music. With any of these categories you can easily import content from a variety of sources and the following steps are repeatable for all types of media that you wish to import. If you haven’t already inserted a USB hard drive, with movies copied on it, into your Raspberry Pi and then navigate to the Videos menu and then to Files. From here a new dialog box will appear and you’ll be able to browse to the USB drive and the movies contained therein. Your movies will appear in the Videos library and you can select and play them from the list. Content can also be stored on network drives, such as a NAS, and OpenELEC can connect via many standards such as SSH, NFS and Samba, enabling content to be centralized in your home and available to multiple media devices.

OpenELEC also comes with a series of add-ons that can enhance the experience and you can find a selection of these in the Add-on menu for each category, eg a great video addon is the Revision3 channel that broadcasts lots of maker and hacker-based shows. You can easily install the add-on by finding it in the list and clicking Install, and OpenELEC will handle everything from there on. For those of us in the UK, you can also install BBC iPlayer functionality thanks to Kodi.

Controlling OpenELEC using a wireless keyboard and mouse is the default method, but there are other stylish and consumer-friendly ways of doing things. First, there’s FLIRC, (available from the PiHut store) which is an infrared receiver that can be programmed to use your existing remote to control OpenELEC.

Alternatively, if you have a spare Android tablet knocking about there’s a free app called Yatse that turns your tablet into a multimedia controller. This app can show your media catalogue on the tablet screen, and enable you to browse and select media which is then played on your television.

NOTE:

Media centers were one of the first projects to emerge for the Raspberry Pi on launch and they have gone on to become increasingly popular. In fact, they are so popular that the Raspberry Pi Foundation has invested in their development. Speaking to Eben Upton recently, he said that he’s a keen user of OpenELEC and is extremely happy with the performance of the latest version on the Raspberry Pi 2 with it. While the Raspberry Pi 1 and 2 both share the same Videocore IV GPU, which means that they both can easily work with large 1080p video files. The OpenELEC user interface is also quite a CPU-intensive process and it was common for users of the original Raspberry Pi to overclock their Pis to coax out every last ounce of performance. For Raspberry Pi 2 this isn’t the case, because of the immense improvements made to the CPU and RAM, and means that Raspberry Pi 2 is clearly the stable base that you need to build your media center upon.

How to Remote control your Raspberry Pi 2

Using your Raspberry Pi is normally accomplished via a keyboard, mouse and monitor but at times this might be awkward, e.g. when controlling a robot or time-lapse photography rig. Remotely controlling a computer is nothing new but it‘s a handy solution to accessing your hard to reach Raspberry Pi project. In this project we’ll install an SSH server, which will enable us to remotely control the Raspberry Pi using the terminal from another machine. We’ll also install a VNC server which will enable us to use the Raspberry Pi desktop over a network. 

Connect up your Raspberry Pi and boot Raspbian to the desktop; you may need to log in using your username and password. If the Raspberry Pi 2 doesn’t boot straight to the desktop type startx in LXTerminal and press Enter. If you’re using a Wi-Fi dongle, make sure that it’s configured and has an IP address. If you’re using an Ethernet connection, open LXTerminal and type ifconfig to find your IP address.

Now we must configure the software that will run on the Pi and give us the access that we need. Open LXTerminal and type the following to setup an SSH server:

sudo raspi-config

Navigate to the Advanced Options Menu and look for SSH server, enable it and exit the application. If you’re asked to reboot, do so and return to the desktop. With the SSH server installed we can now test that it works using another computer. We’re using a laptop running Linux Mint and SSH’d into the Raspberry Pi using the terminal:

ssh pi@IP OF YOUR PI

The first time that you connect to the Raspberry Pi, SSH will ask you to confirm that the Pi is what you think it is via a unique fingerprint, and for this project we can accept that it’s correct. After a few moments you will be asked for your Raspberry Pi password and need to successfully log into your Pi. Any command issued from this point onwards will be run on the Raspberry Pi.

With the SSH server successfully installed, our attention shifts to installing the VNC server. In the SSH session run the following command:

sudo apt-get install tightvncserver

This installs the VNC server and to run the server type:

tightvncserver

You’ll be asked for a password – use a strong one with a maximum of eight characters. You will also be asked for a view-only password that isn’t needed at this time.

In the terminal you’ll see the output of the tightvncserver command, and it will advise you that a VNC session has been started and that you can connect using the IP address or hostname followed by: 1.

Now we need a VNC client on our computer to connect to the server running on the Pi 2. We’ve used Vinagre which comes installed on Linux Mint and Ubuntu. Open Vinagre and click on Connect, in the host box enter the IP address of your Raspberry Pi, followed by: 1 and click Connect. You will be asked for your VNC password. With the correct password entered you can now use your computer’s mouse and keyboard to control your Raspberry Pi and the Raspbian desktop is visible on your desktop.

For those of you looking to play Minecraft in this manner, we’re sorry to say that, at this time, it’s not possible due to the way Minecraft renders itself on the Raspberry Pi. Generally though, you’re now able to use your Pi remotely from another computer in your home. In fact, you can have more than one VNC connection running, which will enable two people to share a Raspberry Pi.

NOTE:

SSH is used by system administrators to remotely work with servers across the world, for instance, a VPS or server running a website or web application. Because there’s no physical access to the device a user has to remote in using SSH either via a terminal, or using a web interface provided by a host.

SSH connections are encrypted and it’s highly unlikely that anyone can see what’s being done, unlike VNC which should not be used over the internet. VNC sends its information via an unencrypted connection and any VNC ports that are open can be scanned by malicious parties. In fact, there’s a website that lists websites with open VNC ports and enables you to connect just like Chat Roulette. However, tunneling a VNC connection via SSH, grants you a level of security.

There are SSH applications for all OSes and on Linux it comes by default, and VNC applications are installed with Ubuntu and Linux Mint in the form of Vinagre.

Which is the best Desktop Virtualization software this 2015?

Virtualization technology has fundamentally changed the computing landscape. We’d get in trouble if we claimed that it was a recent invention (early mainframes used it as a means of provisioning resources), but CPU innovations introduced about ten years ago mean that near-native performance can be achieved when virtualizing x86 code. Now that we can squeeze tens of (easily backed-up and restorable) machines into a single box, data center workloads can be made drastically more efficient. We can also simulate foreign architectures too, e.g. ARM, which is useful for developers working on embedded systems. 

It’s a boon for casual desktop users too – trying out a new OS need no longer be a time-consuming exercise in frustration that jeopardizes your current setup. Even if you just want to test a new piece of software, it’s generally safer to do so in a disposable virtual machine rather than risking your current configuration. Support for in-kernel virtualization (through KVM) and the Qemu emulator mean that Linux users no longer need to rely on proprietary tools. Back in the old days, Sun’s (formerly Innotek’s, now Oracle’s) VirtualBox used to be the only real option. But times have changed so let’s see how the other virtualization apps measure up.

Ease of Install

VMware and Virtualbox both rely on kernel-specific modules being loaded in order to work their magic. VMware will need to compile these, which requires you to install a kernel headers package, and all the necessary compiler jazz. Further, it provides an init script to load said modules, but this will be of no use to Systemd users. If this is you, then you may wish to make your own unit file, rather than having to run this script as root (or see the same error messages) all the time. At the time of writing, 3.19 series kernels needed the VMware sources to be patched, but hopefully this is has been remedied by the time you read this. Most distros ship with Virtualbox packages, so if you’re using a stock kernel you won’t need to worry about anything here.

Virt-Manager needs the libvirtd service to be started on your system before it will work, though a user-friendly message informs you of this, and if you’re using a fully featured desktop environment then it will do this for you – all you’ll need to do is type in the root password.

Both VirtualBox and VMware Workstation are reasonably easy to get to grips with, so long as you don’t get distracted looking at every single option. In comparison, VMware Player doesn’t have too many of these, so you will be able to get your machine up and running in no time. However, you’ll want to install the guest add-ons if you want to make the most of any of them.

The Linux guest additions for VirtualBox are much easier to install (the CD will autorun) than the extra ‘tools’ for VMware, which seemed to require copying stuff off the imaginary CD, changing permissions and then running a script. What is this, 1999? Once you’ve done that though, you’ll be rewarded with better graphics and a number of extra features, which we’ll discuss later on.

‘Boxes’ is by far the most easy to use of our selection, if only because it exposes the bare minimum of Qemu / libvirt’s functionality. VMware Player is a joint second with VirtualBox and then their paid-for counterpart is hot on its heels (these are nominally harder by virtue of having more options). VirtManager isn’t really hard to use, but it does expose enough of the Qemu labyrinth that it may overwhelm novices. It also uses a rather convoluted mechanism for managing storage, especially if you want to store your virtual drive somewhere non-standard. In this case you first must add the required directory as a ‘storage pool’. Any weird modules it needs will be supplied by your distribution though, so it’s got that going for it.

Results are: Gnome Boxes first, VirtualBox and VMware Player second, Virt-Manager and VMware Workstation last.

Performance

Thanks to hardware extensions and paravirtualization techniques, it’s theoretically possible to run some tasks at very close to bare-metal speed.

Things being the way they are though, there is usually some kind of performance hit. In fact, pretty much every configuration where we tested a guest desktop resulted in some measure of sluggishness. But you get used to it, and you can deactivate any eye-candy, or use the appropriate fallback mode, if it really bothers you. For the record, we’re not taking 3D performance into account in this category – that would be unfair and it has its own category over the page. For everyday Terminal-based tasks though, you’d be hard pressed to notice any difference between the candidates. Our kernel compilation experiments suggested that VirtualBox lags measurably behind the competition. The Javascript SunSpider benchmark also agreed with this analysis, with both tasks being about 20% slower than the competition. Disk I/O (especially when your virtual disk is stored on an SSD) and network traffic were swift across the board. VMware clinches it in the end thanks to its support for newer Intel processors.

Results are: VMware’s first, Gnome Boxes and Virt-Manager second, VirtualBox last.

Features

All of our candidates are aimed at different usage scenarios, and so all have their unique selling points. Of course, there are some standard features that we have come to expect, such snapshotting, support for Intel VT-x and AMD-V processor extensions and flexible configuration of the virtual hardware. All of these are catered for by all of them, but some do so much more. It’s important to note that Gnome Boxes and Virt-manager, are just frontends for Qemu (via the libvirt abstraction layer). Qemu is at its heart a processor emulator that can provide virtualization through KVM, but is a whole world unto itself.

Gnome Boxes – 2 stars

When called from the command line, Qemu supports a huge number of options, most which of don’t feature in Gnome Boxes , which aims for (and achieves) a simple and clean look and feel. Through its friendly wizard interface it’s possible to set up a virtual machine in roughly three clicks – you just need to point it towards a suitable ISO. Boxes abstracts the difference between virtual and remote machines and you can connect to either through the VNC, SPICE (which allows audio to work over a network) or OVirt protocols. ‘Boxes’ doesn’t offer much control over networking your virtual machine, but at least provides all the NAT voodoo required so your virtual machine can talk to the outside world. The other candidates will happily set up NAT, network bridges or host-only networks, all of which may be useful given appropriate circumstances.

Virt-Manager – 4 stars

(aka Virtual Machine Manager) exposes much more (but not the entirety) of Qemu ’s functionality. In fact, it seems to have gone overboard with its list of supported x86 operating systems, particularly the Linux genus. That aside Virt-Manager makes it relatively straightforward to setup as complicated a machine as you would like - all manner of hardware can be added, including multiple network interfaces. Besides KVM VMs, Virt-Manager can leverage Qemu /libvirt’s support for Xen guests and LXC containers. It can also optionally poll guest resources and in so doing provide very nice graphs (such as those in the Performance section, which follows about 30 seconds of a Windows 10 startup). Virt-Manager also makes use of Qemu ’s recently added support for USB 3.0 pass-through.

Compatibility and snapshots

Sometimes you’ll want to move VMs between hypervisors. All the software here can import machines stored in the Open Virtual Applicance (OVA) format, which is just a tarball containing a VMDK (VMware) disk image and data about the virtual hardware. VirtualBox allows you to export in this format, but has its own Virtual Disk Image (VDI) format and can work with all kinds of others too.

The qemu-img command can be used to convert between all manners of formats. Of particular note is its preferred QCOW2 format, which allows for multiple snapshots to be stored internally using efficient Copy On Write (COW) trickery.

Boxes, VirtManager, VirtualBox and VMware Workstation all have snapshot support, allowing you to efficiently save various states of your VM as you mess with it. VMware Player only enables you to have one snapshot per machine, in addition to its current state. If you’re doing quite involved regression testing, this will be a deal-breaker. VirtualBox and VMware Workstation also allow you to ‘clone’ VMs, and this is an efficient way of snapshotting, since data is only written to a linked clone when its state differs from that of its parent.

VMware enables you to mount a guest’s VMDK image on the host, which can be useful as well, although the same thing can be achieved by converting to a raw disk image and using standard Linux tools and arithmetic to calculate partition offsets.

Results are: VMware Workstation, VirtualBox, and Virt-Manager first, Gnome Boxes second, VMware Player last.

VirtualBox – 4stars

The original desktop virtualization client, VirtualBox is probably still the go-to tool for many. The program has a clean layout which makes setting up a virtual machine straightforward, but it also manages to offer a host of useful options. As well as limiting the number of CPU cores a guest OS can access, VirtualBox enables you to specify a percentage execution cap for host CPU utilization.

VirtualBox also has support for video capture, so by night you can record Windows tutorials for your YouTube channel. It can import all manner of virtual disks. It only offers USB 2.0 host controller support though, and even then only if you install Oracle’s proprietary extension pack. The shared clipboard and drag and drop options (uni or bi-directional, as befits your whim) are pretty useful though. VirtualBox has handy indicators for disk and network I/O, as well as CPU usage.

VMwarePlayer – 3 stars

The free offering from VMware has come a long way since its inception back in 2008. Most notably, it’s no longer just a player and is more than capable of setting up a quality virtual machine for your amusement. As well as supporting a number of networking configurations (NAT, bridged, host only etc.), it offers some nice traffic shaping options, which is useful if you’re testing the latest version of, say, a DDoS client or seeing how much malware you can get onto a Windows XP virtual machine before it implodes. VMware also supports USB3.0 devices, and installing the guest tools will also enable shiny graphics, shared clipboards and shared directories. Player falls down slightly when it comes to snapshots (it only allows you to make a single one).

VMwareWorkstation – 5 stars

There is a non-free version of VMware Player (VMware Player Pro) available, but we reasoned that for this Roundup we may as well explore the more upmarket Workstation, instead. There are a plethora of extra developer options here, including the ability to group virtual machines into a sort of virtual phalanx, so that they can be simultaneously spurred into action with a single click. VMware Workstation also has support for the new instructions that are available on modern Intel processors, as well as enabling you to set up machines with up to 16 vCPUs and 64GB of RAM. But the Workstation program is largely geared towards integration with the rest of the (considerable) VMware suite, and hence will find most of its relevant use in enterprise environments.

Graphical acceleration

Any of our candidates would serve you admirably, if you just wanted to install a simple virtual machine and weren’t going to be doing anything graphically intensive. However, if you require 3D acceleration then you will have to go with VMware or VirtualBox. Once you’ve jumped through the hoops of getting the guest additions installed (including sidestepping VirtualBox’s double negative question that tries to make you avoid its experimental WDDM support, which is required if you want to use Windows 7’s Aero interface, for instance) then you can enjoy 3D acceleration in VirtualBox and the two VMware clients.

VirtualBox allows you to allocate up to 256MB of system RAM to the virtual graphics card, whereas VMware allows up to 2GB. Note that this doesn’t come from your real graphics card, so often you won’t see any benefit past 256MB. VirtualBox also provides 2D acceleration for Windows guests, which ought to speed up video rendering, display stretching and color correction, though it very much depends on the host configuration – on a fast machine this setting didn’t seem to make much difference. A Windows virtual machine is probably not going to run Battlefield 4 or Middle-earth: Shadow of Mordor (fortunate then that it’s being ported to Linux) with everything turned up to 11, but older or less demanding games will be fine, we wasted a good hour playing popular indie title Don’t Starve before we realized this Roundup was due. It was a much smoother experience on VMware than VirtualBox, but may be that was configuration related – we were using the latest proprietary NVidia driver on Arch Linux which may have favored one over the other.

Results are: VMware’s leads again, VirtualBox second, Boxes and Virt-Manager last.

Licensing and development

VirtualBox used to be subject to a Personal Use and Evaluation License. However, the guest additions and most of the codebase were released under the GPL v2 license at the end of 2010, following Sun’s acquisition by Oracle. However, the exception to this general open source release is the Oracle extension pack for providing a USB2.0 host controller, RDP support and PXE network booting, which remains proprietary. Technically, VirtualBox is subject to a dual-license, so that enterprise customers can pay for a license if the GPL scares them. In fact, some claim that Oracle is neglecting this one time de rigeur tool as all the releases over the past two years have been maintenance-only updates. If this were the case at least Oracle is doing its neglecting of products with the right kind of license.

VMware Player is only licensed for non-commercial use so you may need to part with some dinero to use it legally. A VMware Workstation licence will set you back about £180, which is only really worth it if you want to interoperate with the rest of the formidable VMware suite. VMware has been regularly updating its products, but (apart from 3D support) the open source offerings are hot on its heels. Gnome Boxes , VirtManager , Qemu and libvirtd all come under GPL or LGPL licenses and enjoy frequent updates. Qemu in particular has come on in leaps and bounds since development shifted to GitHub. Also, while we’re on the topic of licensing, we note that VMware has recently been slapped with a GPL violation lawsuit in Hamburg. Naughty.

Results are: Boxes and Virt-Manager first, VirtualBox and VMware Player second, VMware Workstation last.

The verdict

Virtualization is a tough ol’ Round up to judge. If you want to run 3D games, then you won’t even be looking at Gnome Boxes or Virt-Manager, and, unless you have better results with VirtualBox than us, you’ll probably find VMware is the hypervisor for you. But then again, this technology isn’t anywhere near mature enough to compare with DirectX11 running natively. You might even have better luck running games in Wine [Ed –or not] with some of the experimental performance patches. We don’t really like the VMware licenses, especially the one that makes you pay money after a 30-day trial, but the enterprise features of Workstation will be a boon for some. In particular, if you grab VMware’s vCenter Converter, then you can virtualize a machine with the click of a button, ideal if you want to test something new on your current setup.

We haven’t covered the wealth of command-line tools that are provided with all of our candidates, but they are there, along with more than ample documentation. You can use them in your scripts when you go on an all-out virtualization binge, summoning to action an army of virtual machines from the safety of the command line. Hacker types may also prefer to use Qemu directly, and people wanting a simple free and open source solution will be happy with Gnome Boxes.

But there can (except in the event of some sort of tie) be only one winner, and this time it goes to Virt-Manager – we couldn’t really let VMware win after all. Virt-Manager enables you to harness a huge amount of Qemu’s power without having to mess around with lengthy terminal incantations. Machines can be paused, reconfigured, moved around and cloned – all without fuss.

The only thing it lacks is support for graphics acceleration, and who knows, maybe one day this will come. VMware Player and Gnome Boxes tie for second place, as they both score highly for simplicity, and we like the Boxes icon which features a tesseract (or hypercube, or box in a box if you like). VirtualBox and VMware workstation take joint third place, the latter costs money and the former really should perform a little better than it does.

1st – Virt-Manager.

2nd – VMware Player.

3rd – Boxes.

4th – VirtualBox.

5th – VMware Workstation.

KaOS 2015.02 Review

Unlike most other Linux distros that burden users with ample choice, KaOS has consciously decided to keep the options limited. The installable live distro is available only for 64-bit machines and offers only KDE as the desktop environment (DE). Its repos only feature x86_64 packages and the install yields a GTK-free base system. 

The package and desktop restriction isn't to ensure a bloat-free release, but because the developers believe in working only with the best available DE. By its own admission, the distro is designed for users who’ve tried others only to conclude that KDE is the best option for them.

Being a rolling release, users don't have to worry about fresh installs and complicated upgrades. With the underlying Pacman package management system, a single command is all you need to keep an installed system up to date.

Building blocks

The distro features the distro independent Calamares installer. Like most popular distros, KaOS doesn't let users choose what packages to install. However, the installer does provide several partitioning options, and the chance to carve out partitions to your liking. While there aren't any helpful tool tips to guide first time users, the installation process is straight forward and the install intuitive.

While the distro ships with the usual complement of standard apps, such as browser, office suite, music player and IM clients, it only carries the QT alternatives for each as far as possible. This isn't to say that GTK apps are entirely absent. But you’ll only find these where the QT counter parts aren't up to scratch or the software’s popularity necessitates their inclusion.

The distro doesn't use any upstream repos, choosing instead to build each package from scratch specifically for KaOS in its three repositories. The Core repository contains all the essentials like the Linux kernel, Systemd, etc. The Main repository comprises the drivers, libraries and firmware needed to get the apps to function properly. The apps themselves are relegated into the Apps repository, which is fully rolling. This means that KaOS provides rolling updates for all packages and that no package is ever older than one year, which ensures tighter integration.

The distro ships with Plasma 5 as the default desktop and features KDE 14.12 with all language packs preinstalled, proprietary NVidia drivers and Linux Kernel 3.18.7. It can play YouTube videos out of the box and supports a variety of multimedia formats. Due to its focus on QT apps, the distro offers Calligra 3.8.92 as the default office suite, although LibreOffice is available in the repos. The distro also includes various QT/KDE-specific internet apps such as Qupzilla web browser and KDE-telepathy.

You can install popular GTK apps such as Firefox 35.0.1, Thunderbird 31.4.0, etc. from the repos using the powerful Octopi package management tool. The latest release also features CacheCleaner, a tool for Octopi to remove packages from Pacman cache.

KaOS is designed for a very specific user base: KDE aficionados and advanced users who've finally tired of hopping distros and are skilled enough to work with a rolling release.

This is a well-rounded distro that delivers on its mission statement – a fully functional, rolling KDE release. We gave this a Cloudy Circuitry score of 85/100.

Bodhi Linux 3.0 Review

Bodhi Linux is essentially a one man project, so Jeff Hoogland’s retirement seemed to sound the death knell for the beloved build-your-own distro. Thankfully, he had a change of heart and returned to continue work on the release that was under development when he quit.

This third major release of Bodhi is based on Ubuntu 14.04 LTS and uses Linux kernel v3.16. This is also the first release that supports the UEFI BIOS found in nearly every new machine as well as the SeaBIOS found on some newer Acer and HP Chromebooks.

On the other end of the spectrum, the latest release also works with processors that don’t support PAE, such as the Pentium M Centrino.

In addition to borrowing the underlying infrastructure and core libraries, the project also uses a customized version of Ubuntu’s user friendly Ubiquity installer.

An enlightened desktop Bodhi is one of the few distro that uses the light and attractive Enlightenment desktop, and the latest release features a customized desktop based around Enlightenment E19.3. The one major change that long term users will notice is that unlike previous releases users aren’t asked to choose a layout style when they boot into the desktop. The developers argue that the consistent look will help make Bodhi more approachable to new users, although it wouldn’t have hurt to include other predefined styles in the Profile Selector tool like before.

In line with Bodhi’s philosophy of installing a minimum base for users to build their system, the distro ships with the Midori web browser and a bunch of Enlightenment apps written using the Enlightenment Foundation Libraries (EFL). These include the file manager, the Terminology terminal emulator which includes bells and whistles like a visual tab bar and Gravatar support, and the Ephoto picture viewer among others.

There’s also the ePad plain text editor which is written by Hoogland himself in EFL and replaces the Leafpad editor featured in previous releases. ePad in its current form supports most of the features you would expect from a text editor. Another in-house app of note is eepDater which is a simple but fully functional update manager.

Any distro that ships with a minimal set of apps and calls upon its users to flesh out their installation must have an exceptionally good package management system. Bodhi Linux, piggy-backs on the excellent apt-get packaging tool and the AptURL protocol, has an app installation platform that’s flexible and easy to use. New users can just use its online AppCenter, optimized for the included Midori web browser, while experienced users can use the Synaptic . Bodhi’s online Appcenter includes a limited number of frequently used apps – such as LibreOffice, Firefox, Okular, VLC, Handbrake, Thunderbird, Pidgin etc –has also been overhauled in this release. The search functionality has been improved and now looks up queries in the software titles as well as their descriptions for more meaningful results. Furthermore, users now have the ability to rate the apps.

Bodhi Linux is more than a stripped down version of Ubuntu. The distro adds features that’d be of use to users irrespective of their skill set: You can e.g. use the distro in Presentation mode which temporarily disables the screensaver, lock and power-saving features, and if you forget to enable this mode and quickly disable the screensaver, the distro offers to enable the Presentation mode for you. Now that’s what you call user-friendly.

This Linux distro is a wonderful minimal version for DIYers that provides an ideal base and all the suitable tools for customization. We gave this a Cloudy Circuitry score of 95/100.

Imagination Creator C120 Review and Test

Imagination Technologies is most well-known for its PowerVR division, In February2013, Imagination acquired MIPS Technologies, masters of the most popular instruction set you've never heard of: MIPS. Being thus possessed of GPU and CPU silicon, it makes strategic sense to release a product with both.

The fruit of its labor is the Creator CI20. As we know this class is dominated by the Raspberry Pi, there are plenty of alternatives though. Many of these, e.g. the Beagle Bone Black and Intel's x86 Galileo, offered a little more computational horse power. But now the Pi has had a reboot and is capable of running a desktop smoothly, these rival boards have lost their raison d'être.

The Creator is powered by a dual core, 32-bit, 1.2GHz CPU, it features 1GB of DDR3 memory and 8GB of NAND storage. Besides that there is an SD card slot, 100Mb Ethernet, 802.11n wireless, Bluetooth, an IR receiver and three USB2.0 ports, two full-sized and one mini. USB On-The-Go is enabled through a jumper, and the mini port is paralleled with one of its larger siblings, so only two ports can be used at once.

The board itself is powered by the same barrel connector as the Sony PSP. As you'd expect there's an HDMI port and a 3.5mm audio jack. There are also all manner of I/O pins for the hard core.

The Creator officially supports Debian, or vice versa, since Debian has supported the MIPS architecture for some time. There are also images available for Gentoo and Arch Linux, as well as the embedded-centric Angstrom and Yocto Sano.

Interestingly there is also an Android 4.4 image available, although being based around the Android Open Source Project you'll have to side load any apps you want. The MIPS architecture is fully supported by Android, but many apps are still compiled solely for ARM, limiting the current potential here. However, those wishing to develop for Android may be tempted by this low cost device, though they'll have to put up with a 5minute (or 1cup of tea) boot time.

Highest on our list of gripes is the board's fussiness over displays: the developers warn that passive HDMI-DVI adapters may not work, but we found at least one monitor that it failed to talk to over HDMI, leading us to suspect that someone broke it. In fact, the 3.0 series kernel shipped with the Debian image can't query EDID data, it is possible to compile your own 3.16 kernel, but at the time of writing this breaks acceleration.

Numerically the Ingenic JZ4780 SoC seems impressive, but comparing the Creator's MIPS architecture to ARM or x86 is akin to comparing apples and oranges. The Raspberry Pi 2's four cores will trump the Creator for multithreaded applications, but floating point-heavy math operations will favor the Creator's FPU. The PowerVR graphics can do 1080p at 60Hz, or 2k at 30Hz, outdoing the more basic Pi output. Booting Debian takes a couple of minutes and the XFCE desktop was sluggish with frequent tearing when moving windows. Games like Supertux and Chromium BSU run fine though, and it does come with some quite impressive 3D demos. One such demo achieved about 9 million triangles a second as so many blocks arrange themselves into various shapes above a valley. But pretty demos alone do not a Pi-conquering board make, and, sad to say, the CI20 Creator's extra features aren't quite justified by the £20price differential.

If not for the Pi 2, this could tempt many hobbyists. It still has a few extra features, but only for a small niche. We gave this a 6.5/10 score.

Intel Edison Review and Test

Originally, the Edison was intended to use the SD form factor and be powered by Intel’s shiny new Pentium-class Quark processor. Feedback from users of the Quark-based Galileo development board, however, suggested this idea wasn’t going to fly thanks to extremely limited performance. Intel’s answer was a swift redesign into a module format, and to combine a 100MHz Quark with a 500MHz dual-core Atom.

The Edison is still impressively tiny, especially when you look at the feature list: as well as the dual-core CPU and low-power co-processor, the 35.5 x 25mm package includes 1GB of LPDDR3 RAM, 4GB of eMMC storage, and 802.11a/b/g/n dual-band Wi-Fi, plus Bluetooth 4 and 2.1 EDR. Its 40-pin Hirose high-density connector at the rear also carries 14 GPIO channels (four of which support pulse-width modulation), two UARTs to an SD card channel, I²C, I²S, SPI and even USB.

You might wonder how a hobbyist gains access to these wonders, and you’d be right to do so. The official answer from Intel is by using a motherboard into which the Edison bolts; it boasts an Arduino pin-out for GPIO and power, along with USB and micro-SD ports. Using this board, you can turn the Edison into a ridiculously overengineered Arduino Uno, complete with embedded Linux OS and wireless network connection – it isn’t 100 per cent compatible with the Uno, but it’s good enough for most cases. Doing so, though, negates the size and weight benefits of the Edison: the Arduino motherboard is nearly as large as three Arduino Uno boards.

A better option, although I haven’t been able to test it at the time of writing, comes from US-based hobbyist electronics specialist SparkFun. The SparkFun Blocks for Intel Edison are significantly smaller than the official motherboard, and each one is designed to break out a particular feature: there are Blocks to add analogue to digital conversion, USB ports, motor control, pulse-width modulation, GPIO and even one Block featuring a built-in rechargeable lithium-polymer battery. Each Block has a Hirose connector on each side, enabling them to form a stack that combined whatever features you desire with the Edison as the cherry on top.

The fact the Spark Fun Blocks had to exist at all, however, shows that Intel has perhaps missed the mark. The Galileo may have been a poor performer, but its form factor was at least well suited to hobbyists; the Edison is only immediately usable with either the oversized breakout board, making it a more powerful Galileo, or with third-party add-ons.

For industry, the Edison offers more possibilities. People who can spend the time designing circuits with tiny high-density connectors will find the Edison more tempting than the relatively feature-light Raspberry Pi Compute Module, and with the added bonus of using the familiar x86 instruction set architecture.

For educational use, combined with the SparkFun Blocks, there’s certainly an argument for the Edison. For hobbyist use, though, a Galileo is a better option, as is just forgetting about x86 compatibility and using a proper Arduino board at a fraction of the cost.

The Intel Edison is available for £47.94 including VAT as a bare module (useless to hobbyists) or £81.54 including VAT bundled with the Arduino motherboard as reviewed; a cheaper breakout board bundle costs ____.

How to create your own Raspberry Pi 2 chassis

We already used 3D printers for PC modding, taking you through the basics of using TinkerCad to design your objects. There’s far more to TinkerCad than making 2D fan grilles and modding components, though, and in this post we’re going to use it to create a case. 

Most reasonably priced 3D printers can only cater for sub mini-ITX motherboards, but Intel NUC and Raspberry Pi cases are definitely options for printers with 6in³ printing capacity or more. As Raspberry Pi 2 is the hot hobbyist product of the moment, we’re going to create and print a custom-designed case for it using the 3D Systems Cube 3 printer, while also showing you how to go about creating your own. As the Raspberry Pi 2 is fanless and doesn’t have any active cooling, this job is really easy and you can have some fun with the case design too.

Tools You’ll need:

1. TinkerCard

2. 3D printer or 3D printing service

3. Digital vernier caliper

4. Finger files

How to…

1. DOWNLOAD RASPBERRY PI STL

To create your Pi 2 case, you first need a 3D model of the PCB. Thankfully the Pi B+ PCB is physically identical to the Pi 2 in terms of case layout, and is available to download for free from [www.thingiverse.com]. Search for the Raspberry Pi B+ board by jayftee and grab the 9MB STL file.

2. IMPORT MODEL INTO TINKERCAD

Importing the STL file into TinkerCad [www.tinkercad.com] is easy, and you can then start to build your case around it. In the right-hand menu, under Import, simply point TinkerCad at your downloaded STL file and upload it.

3. START WITH A CUBE

Now you can start building the case, starting by dropping a cube shape into the workplane from the same menu. You can, of course, choose any shape but we’ll be making a rectangular case so we’re starting with a cube.

4. ENLARGE CUBE TO FORM BASE

Modify the cube so it’s the right size for the base of your case. You can either use the measurements provided by TinkerCad to guide you, or simply make the base a little larger than the PCB. The total PCB model size is 58.5x 86.89 mm, so your base needs to accommodate that area measurement.

5. CREATE SCREW HOLE INSERTS

With the PCB lined up on your base, create some cylinders for inserting holes into the case to mount the PCB. You can use standard PC screws – just make a cylinder that’s slightly smaller than the screw thread (a digital vernier caliper will come in handy), so the screw can bite into the plastic.

6. PLACE INSERTS INTO BASE

Move the inserts on to the case and position them in the PCB mounting holes, in order to ensure they line up once the case is printed – another good reason to have a 3D model of the PCB to hand.

7. CHECK BASE FOR CLEARANCE

You now need to raise the inserts, so they don’t go all the way through the base when you convert them in to holes. Lift them by 3-4mm so the screws can pass a little way into the base – you can check they’re clear by looking underneath the case.

8. CREATE CUT-OUT INSERT

As we saw in last month’s 3D printing feature, it’s best to create holes in as many parts of your project as possible to cut printing time and costs. We started by laying down a large object in the base to create an opening – we’re also moving the PCB model out of the way for now.

9. SWITCH TO HOLE MODE

Now convert the object into a hole by selecting it, then clicking on Hole in the Inspector box, which converts it from a colored object into a hole. It should now appear translucent, which indicates that it’s acting as a hole and not a solid object.

10. GROUP OBJECTS

To finish creating the hole, select the base and hole only (leave the mounting holes separate for now), and select Group from the top tool bar. This process will delete the area of the base with which the hole is merged.

11. CREATE CUT-OUT DESIGN

You can then create a vent using individual pieces or, as we did, drop some letters and numbers into place. You can simply enlarge or reduce them to size as needed, depending on your chosen vent design, or you could lay them on top.

12. INSERT SUPPORTS

To hold the characters in place, we created small supports that connect them to the case at the edges. They’ll still look like they’re suspended, but this reinforcement will help them to survive any knocks. We also reversed the lettering, so it will read the right way when you pick up the box.

13. GROUPOBJECTS

Merge your vent design objects into one group by selecting them and clicking the Group icon again, being careful to avoid any objects that you don’t want to group yet.

14. CLONE BASE

We want to apply the same design to the lid of the case, and the easiest way to do this job is to clone the entire base, then remove the PCB support holes as shown. You can move the lid to one side for now.

15. ADD PCB STANDOFFS

You don’t want to screw the PCB directly into the base, so create some small standoffs into which you can insert screws. These bits are just like the standoffs in a PC case, except you only need them to be 5mm or so high. Place these standoffs over the mounting holes.

16. CREATE HOLES

You can now convert the mounting holes from solid objects into holes. Check if they’re aligned with the standoffs, then click on each one and convert it into a hole using the same method employed in step 9.

17. GROUPOBJECTS

In the same way you grouped the vent objects together, you can now group all the base objects together so you can move on. Hold down the left mouse button down, as you would to select an area in Photoshop, or characters in Word, then click the Group button in the top menu.

18. CREATE SIDES

Now put the PCB model back in place and create some side sections. Only two sides of the Raspberry Pi 2 have ports, so the other two can be solid objects. We dropped in a couple of cube shapes and manipulated them to form 3mm-thick walls that run from end to end as shown.

19. CREATE LID MOUNTS

You’ll need away to mount the lid too. You could have a sliding lid, but we preferred the idea of securely screwing it into place. We created two mounts that could sit inside the case, level with the side sections.

20. INSTALL LID MOUNTS

These mounts sit diagonally from each other, and only two are needed so save space – most of the lid will rest on the base’s side sections anyway. We’ve placed the mounts so that there’s still enough clearance between the PCB and the side sections – at least 2-3mm.

21. ADD LID MOUNT SUPPORTS

We went through to the end and printed a test base at this point, but found the supports were too weak, breaking when screwing the lid into them. We went back and added these corner braces to strengthen them. You can also add vents in the sides – we opted for a hexagonal pattern.

22. PLACE LID

Now you can move the lid into position, lining it up with the base and side sections. You can then move onto finishing the side sections and creating holes for the mounting screws.

23. SURROUND PORTS

You don’t really need to create a sealed case with the Raspberry Pi 2, but a sealed case definitely looks neater, so we went some of the way, and surrounded the end ports at the side and the bottom by extending the lid.

24. CREATE PORT SIDE SECTION

We then added a further surround in the lid section, as well as a long side section as the final piece –we’ll cut holes into this part through which the ports can protrude. The idea will be to delete the PCB model, leaving a 3D-printable and two-part case.

25. CONVERT TO HOLE

By converting the side section to a hole, it becomes translucent so you can see the ports you need to cut out to provide access. There are just three – the HDMI port, micro-USB connector and audio output.

26. ADD PORT AND LID HOLES

You then need to recreate the ports from appropriately shaped objects, and then resize them to provide plenty of clearance around them. For the HDMI port, we opted for a rectangular hole, rather than a larger-sized hole shaped like an HDMI port, to ensure there was enough clearance.

27. PRINT YOUR DESIGN

Click on Design, then Download for 3D printing to get the STL file for your design, and either print it on your own 3D printer or use a third-party service, then enlarge any trouble some holes and file the edges. Here’s our design, printed on 3D Systems Cube 3, with a Pi 2 installed.

Raspberry Pi 2 Alternatives

The Raspberry Pi may not have been the first single-board computer on the market, but its success – selling five million units and counting – has led to an explosion in the market never seen before. Here’s a look at the most notable competition to the Pi.

Lemaker Banana Pro

The follow-up to Lemaker’s Banana Pi, as reviewed in Issue 131, the Banana Pro improves on the original design by adopting the new B+/Raspberry Pi 2 GPIO header and adding integrated 802.11/b/g/n Wi-Fi. Additional features over the Raspberry Pi 2 include on-board SATA, infrared support, Gigabit Ethernet and wider compatibility with various operating systems. Its dual-core AllWinner A20 processor, however, lags behind the Pi 2’s BCM2836 in the performance stakes, and there are only two full-sized USB 2 ports – although, unlike the semi-compatible ports of the Pi, they’re fully standards-compliant.

Intel MinnowBoard Max

One of several open-hardware projects Intel is throwing at the hobbyist market, the MinnowBoard Max is significantly more expensive than the Raspberry Pi but includes a full-fat dual-core 64-bit x86 Atom processor and 2GB of RAM in its top configuration. As detailed in Issue 136’s review, it’s a significant improvement over the 32-bit original and boasts some of the best software compatibility of any hobbyist single-board computer. It lacks the community built up around the Raspberry Pi, however, and the add-on Lure boards for it are currently thin on the ground.

SolidRun HummingBoard

The HummingBoard began life as an internal development board. Available in several editions, ranging from a single-core model to the reviewed HummingBoard-i2ex, it uses an uncommon computer-on-module (COM) design. As such, it’s theoretically possible for SolidRun to release upgraded modules you can buy at a lower cost to replacing the entire device, although no such modules have appeared yet. A promised quad-core variant is also missing, leaving the top-end model lagging behind the Raspberry Pi 2, despite using the more efficient Freescale i.MX6 SoC.

CubieTech Cubieboard 4

The Cubieboard 4 might be bulky and pricey, but it’s a beast. Based around the AllWinner A80 SoC, it packs four Cortex-A15 1.8GHz and four Cortex-A7 1.2GHz cores in ARM’s big.LITTLE layout, but with the ability to run all eight cores simultaneously, giving highly threaded workloads a great speed boost. The 2GB of RAM is also generous, but the PowerVR G6230 GPU means hardware graphics acceleration is unavailable outside Android – a major blow for anyone hoping to run a more flexible operating system.

Hardkernel Odroid C1

As soon as the Raspberry Pi 2was announced, Hardkernel was quick to position the C1 as a more powerful alternative board. Based on an Amlogic SoC, the Odroid C1 includes four 1.5GHz Cortex-A5cores, a Mali-450 MP2 GPU and 1GB of DDR3 memory, along with Gigabit Ethernet and a Raspberry Pi+/2 compatible 40-pin GPIO header. Support for Android and Ubuntu 14.04 is included as standard, but the board again lacks the excellent community enjoyed by the Raspberry Pi.

Imagination Technology Creator CI20

The Creator CI20 uses the MIPS instruction set rather than the more common x86 or ARM ISAs. The CI20 has enjoyed an upgrade since its first release; with double the flash storage at 8GB and an eye catching purple PCB. Its dual-core 1.2GHz CPU is also more efficient than its ARM rivals, but can’t beat the quad-core BCM2836 for multi-threaded speed. Its PowerVR SGX540 also lacks acceleration support outside Android, although Imagination has promised to address this issue in the future.

How to overclock, update, and multi-boot your Raspberry Pi 2

The similarities between the original BCM2835 and the new quad-core BCM2 extend to their support for overclocking, providing power users with a means to squeeze a little more power out of their Raspberry Pi the risk of a shortened lifespan. The safest way to get some extra speed out of your new device, assuming you’re running Raspbian, is to use the configuration wizard. At the terminal, type: 

sudo raspi-config

Scroll down to 7 Overclock, hit Enter, read and acknowledge the warning, then choose the bottom option – labeled Pi2 – from the list that appears. Make sure you don’t use any of the other options, including None; these settings are designed for the original Raspberry Pi, and could leave your shiny new Pi 2 running slower than stock speed. Exit the tool by tabbing to Finish and confirm you’d like to reboot when asked. For even higher speeds, you can adjust settings directly within the boot configuration file. Run the following command at the terminal:

nano /boot/config.txt

Head to the bottom of the file, and use the following to adjust the performance of the processor:

arm_freq - Core clock frequency in megahertz

gpu_freq - GPU clock frequency in megahertz

sdram_freq - Memory clock frequency in megahertz

over_voltage - Increase the SoC’s voltage, in steps of 0.025V

Be aware, however, that manually adjusting these settings – in particular the voltage – can render your warranty void.

GPIO expansion

Although the latest Raspberry Pi models have a longer general-purpose-input-output (GPIO) header than their predecessors, it still lacks many functions of rival devices – in particular support for hardware pulse-width modulation (PWM) or analogue inputs. The solution is a GPIO expansion board.

There are plenty of add-ons from which to choose, many of which are tailored to a particular task such as robotics or home automation. A comprehensive list is available at [elinux.org/RPi_ Expansion_Boards].

Updating the firmware and kernel

You’re almost certainly aware that Raspbian needs regular updates to keep it performing in tip-top condition, but you may be running an older kernel and firmware release if you’re relying on the traditional method of using this command:

sudo apt-get update && sudo apt-get upgrade

To ensure you have the very latest kernel and modules – providing bug fixes, stability and even performance improvements – you need to use a separate tool bundled with Raspbian. At the terminal, type the following command:

sudo rpi-update

This command refreshes the firmware, kernel and all the modules from the official GitHub repository, ensuring that your system is bang up to date.

Multi-boot via NOOBS

While its name may suggest that it’s designed for newcomers, NOOBS (New Out-Of-Box Software) includes one feature that power users may find very useful indeed: multi-boot support. Download NOOBS from [www.raspberrypi.org/downloads], and extract the contents of the Zip to a FAT32-formatted micro-SD card – there’s no need to mess around flashing raw images. Insert the card into the Pi and boot it, then tick the boxes next to as many operating systems as you’d like to install.

When installation is complete, the next boot will load a menu asking which of the installed operating systems you want to boot. Future boots will display this menu for ten seconds, after which the most recently booted OS will be automatically selected. When installing your OS, however, be aware that not every choice available in NOOBS is yet compatible with the Raspberry Pi 2.