I owe a lot to this little, £130, underpowered (even at the time) machine. Sat in my car refining and sending off my CV for numerous jobs whilst waiting for my hateful call centre job to begin helped me break into a technical role.

The end of Windows 10 support in October 2025 could be considered a full stop on this laptop’s usefulness, however Windows claimed this as a victim long ago. The 32GB eMMC storage was entirely consumed by just the operating system and its pending updates, which relegated this system to version 18.09.

One point of the X205TA that really impressed was the battery life, with an OS estimated 10 hours at full charge it was something 2016 me had never seen before.

The thought occurred that I’d like a lounge terminal, something with a proper keyboard and screen that I could use to SSH into other machines, with a web browser for reference.

Could I breathe new life into the X205TA with the introduction of a Linux Desktop?

Specifications

All that is required is a terminal and a web browser.

This could be achieved with a simple window manager but due to the laptop’s svelte specs (Intel Atom Z3735F, 2GB RAM, 32GB eMMC storage), I don’t think a barebones setup is going to give much of a performance boost.

Debian is my preference, naturally the system will be built on this. Luckily, and something that needs consideration these days, is this is a 32-bit processor.

Secure boot, enabled by default, will need to be disabled in BIOS in order to install a Linux system.

Desktop Environments

Debian’s wiki has a page for the X205TA , but the referenced kernel versions suggest this may not be the most up to date.

Attempting a fully fledged environment in the first instance…

KDE Plasma

For this I opted for a minimal install of Debian 12, followed by a minimal install of KDE Plasma on first login:

apt install kde-plasma-desktop plasma-nm

This give the KDE desktop, with a couple of issues. Firstly and immediately obvious, despite WiFi working for the net-install ISO and on first boot, post KDE install the WiFi would not connect to my network.

This was resolved by commenting out ALL of the /etc/network/interfaces file, then, through the GUI, forcing the wireless to connect over 2.4Ghz to my SSID that is active of both 2.4 and 5Ghz.

Following this, audio was not detected, not a dealbreaker but something worth noting.

The touchpad controls (tap to click) though not enabled, was easily configured. However, its was quite sluggish, with RAM at 1.4GB at idle, and a browser (Firefox or Chromium) exacerbated this.

Next was a renowned lightweight desktop…

LXDE

Similar to KDE, install via a net-install ISO went seamlessly, but on first boot to the desktop no signs of WiFi or networking was to be seen.

This could of prompted a troubleshooting session, but the lack of touchpad customisation was to be a dealbreaker.

On to the next…

XFCE

Like LXDE, this has the same goal of being a lightweight desktop, but seems to be under more active and continuous development.

Installation was via the same net-install ISO, but on first boot, everything just worked!

With all installation, Debian advised that propriety firmware was recommended prior to installation, yet I’m not seeing any detriment to skipping this step.

RAM usage at idle sits at ~800MB, yet I feel offers the same customisation offered in KDE at 1.4GB, especially with tap to click on the trackpad.

With this level of simplicity of install, XFCE is the one for me.

Summary

Also offered with XFCE are the same comforts of a modern UI, such battery life status and estimates. It’s amazing that this cheap is still able to offer 9 hours of battery life at 10 years old, this was after a charge cycle and booting the system off-grid.

It’s over optimistic to think that this could be a serious workhorse, now or even back then. But to have a simple terminal machine to hand when lounging on the sofa as a more convenient option to a frustrating mobile experience still gives this little cheap laptop some value.

Switching to a Mac for daily use takes some getting used to, so my hope at least was to preserve my 3x 1080p monitor setup with the laptop in clamshell mode (lid closed). I’d been accustomed to for many years so was hoping this would aid the transition to a whole new Mac ecosystem.

The specifications and online hear-say suggest that the maximum supported outputs are two monitors plus the Mac’s display. But there is a way to get 3 screens running on an M1 Macbook Pro…

For MacBook’s of this age, at least, the secret to getting more monitors is a dock with DisplayLink, a method of driving displays over USB.

This is a great solution to overcome limitations of the hardware, however it comes with some considerations:

• As the extra displays are delivered over USB, it is limited by the USB bus speed. Luckily for me a 1080p monitor is not (these days) not that much of a bandwidth hog. But for those using 4K or higher it may hit the limits of the USB bus, or its use will impact transfer speed of other devices.
• The extra displays are managed by software, so the experience may not be as seamless as a dedicated display output. Many times a reboot was the solution to solve the extra displays not … displaying.
• A constant “your display is monitored” icon is in view on the taskbar, granted this is a minor caveat, but can be distracting as it seems to appear and disappear at random intervals.

Dock

Given the port density of a modern MacBook the natural choice to add three screens was to use a dock, which allows all the extra screens plus a few more connectivity options.

I settled on the Dell UD22, this is a more universal version of the Windows centric Dell dock’s I’ve previously been used to. In addition to being Mac compatible it is also capable of being able to give 96 watts of power delivery to the host machine, allowing a one cable serves all connection to the laptop for a cleaner look.

Connection

Despite the dock having 3 monitor connections (2x DP, 1 HDMI), knowing the limitations I chose to connect the 2 of the monitors via a DP and HDMI port. The third connects via the Thunderbolt port on the rear of the dock.

However it will not work yet…

Software

To enable the third (or More) screens, download DisplayLink software, and allow to run at startup.

A restart later and all 3 screens are running off the same dock.

Caveats

From experience, exploiting the hardware limits are not as seamless as hoped:

• The laptop if left in clamshell mode 24/7 mostly works fine, but any restarts for me stops all screens from outputting video. For this the laptop needs to be unplugged, opened, closed, then plugged in again.
• Sometimes, the DL screens will not wake. This requires a reboot and will need to refer to the above point.
• The DisplayLink screens will not behave the same as the normal outputs. Oddly I notice the DL screens will wake before the others connected to the dock. Resulting in the “LIVE DESKTOP” being offset to the others.

Conclusion

For adding an additional 1080p monitor at least (in my case), the DisplayLink monitor for general web browsing is perfectly useable and I cannot see any degradation in quality or framerate to a point where “this is the USB monitor”, I can’t vouch for multiscreen gaming however as; a) it’s a work laptop, and b) it’s a Mac. 

For background, I was previously in a FTTC environment getting average speeds due to my distance from the cab, however good enough to assumably be overlooked for the next phase of the Openreach Ultrafast rollout. Not that I’m complaining, 2020 is the year my speed gets a much-needed boost.

2020 also turns out to be the inaugural international work from home year, so had the opportunity to have a front seat view from my home office on the activity and timeline that brought FTTP home.

I’d like to share my observations and timeline as an example of what you can expect should you get the inkling of fibre coming to your street soon.

Before we get started, as my expectations rose, I found this post by Andy’s World invaluable for identifying activity and helping me confirm that FTTP was on its way.

Background

As mentioned, I was already in a FTTC enabled area. From the cab the copper “last mile” travelled underground from the cab to my nearest telegraph pole, then overhead to my property.

I was getting speeds on average of 40Mbps down and 5Mbps up due to the distance from the cab, I also suffered loss of sync approx. every other day. I notice that my overhead line was a lot older than others on the pole, identified by the greying colour and how you could identify both cores on the cable rather than the single looking shielded core of adjacent cables. A 48h MTBF was not ideal but well within Openreach service levels, meaning I couldn’t get this drop cable replaced without a significant cost.

Being in an FTTC area, no matter how bad the connection is usually meaning that it is not up for consideration for upcoming phases of an Openreach FTTP rollout. Couple this to a Virgin media rollout along my street the previous year and this would reason that the demand for FTTP would be low.

Timeline

Early days – May 2020

Entering my second month of working from home, notice a lot of Openreach vans passing my door. I may be more observant of these given my profession, but it was a lot of activity over the normal fault-based callouts.

Later and as I started venturing out of the house again, it looks as though they were commissioning a new cabinet a couple of streets up, therefore assumed the activity was for this.

June 2020

I happen to live near a train station, and was intrigued by the letters through the door.

Since moving in a year earlier, I’ve received a couple of letters of this type. But a new ticketing system combined with increased Openreach activity made me think if maybe this would include fibre. Hopes were still low at this point, but even if the rail infrastructure went for leased line circuits, this would increase the likelihood of FTTP.

Hopes were increased later in the month, when I noticed active involvement in the pole supplying my house.

To coincide, another letter relating to communications works on the rail system piqued my interest further.

August 2020

Another letter, this time for roadworks to clear blocked ducts for Openreach, with my and adjacent poles labelled for work.

This was very telling, as FTTC was already available with a on pole distribution point (DP), either this was to increase the number of copper circuits, or for deployment of a new technology.

25^th August 2020

The conformation I needed; thanks to the Andy’s World post I knew what to look for.

This day a cherry picker arrived soon after 9am, and then commenced the installation of a FTTP termination at my local pole. It was a day later before the “Fibre lines overhead” label appeared but it was clear FTTP was on its way.

Through my work I’m able to check and order connectivity for premises, so now began my daily check to see if/when FTTP would be available.

10^th Sepember  2020

Daily ISP provision check 16, and FTTP appeared as an order option, at 9am it appeared as an option but unable to order to lack of capacity, but an hour later it was fully available:

Before getting ahead and ordering the full 1000/115Mbps, time for a reality check to see what my current router can actually process, resulting in an order of 220/30Mbps later that day.

My ISP like many others allow me to upgrade speed within contract but now downgrade, so this will be a good starting point with the option to increase should I get a router upgrade.

Activation date was set for Wednesday 23/09/2020.

Monday 21^st September 2020

Activation week, and more activity around the Openreach pole. Involving a cherry picker, a team of 3 descended to check over the FTTP distribution point.

Also noticed that the tensioners (Or the metal wire that attaches to the ring on top of the pole and is wrapped around the drop cable hold it to the pole) of my current copper drop cable were unwound slightly.

Tuesday 22^nd September 2020

The day before I had an unexpected reconnaissance visit from the Openreach engineer assigned to my activation the next day.

My previous drop cable for reasons I cannot fathom was run into the house via a corner of the uPVC window frame, so was eager to have the fibre enter the house via a new route. And the five-minute visit was worth it as it was determined that a cherry picker was needed at pole side due to the slope of the road.

Wednesday 23^rd September 2020 – Install Day

ONT- Optical Network Termination – The modem that converts light to electrical data.

Openreach engineer turned up at 9am as promised, and a recap of the installation path.

The assumed installation work would disconnect the copper to the building, and replace the analogue master socket like for like with the new ONT. I needed the FTTC to remain in use while I transition IP addresses so at request the old copper would remain active.

Sub-plan was to leave the copper untouched, and run a new route of fibre only cable into the house. However, when the cherry picker turned up so did a supply of combined copper/fibre cable, allowing the supply of all services via the new route, and to replace the aging copper.

Installation

Fibre installation comes in two parts, the distribution point (either overhead or underground) to the outside of the property, and a fibre from the ONT inside the property out to meet the other.

First the hole was drilled close to the ONT location, the ONT was mounted on the wall and the internal fibre ran to the outside. The internal fibre is much thinner than from the pole due to the less armoured sheathing and benefits with more flexibility, although this comes at a risk of snapping or kinking the fibre, resulting in light loss and ultimately network loss.

Next stage with the help of the cherry picker was to plug in the fibre to the DP on the pole, and run it overhead towards the house. The fibre was secured to the house and run down the side of the building. During this the old copper was disconnected and new was connected at the pole.

Thirdly, the internal and external fibres are joined together. A new box is attached to the external wall that has inside a spool that will hold a few windings of each fibre, as the joining of the cables is considered a failure point, as few windings are included during the initial installation to give extra fibre should the cables need to be re-joined in future.

A portable splicing machine is used to join both fibres by lining them up, heating to fuse the two together, and shrink wrapping the join for protection. The unit also checks the light levels at this location, mine being 17 at this join, 14 at pole.

The fibre is coiled in the external box and closed, completing the installation of the fibre.

Commissioning is a simple as Openreach completing the order, then waiting for the PON light on the modem to illuminate. Such is the technology, there’s no fine tuning or testing of the signal, as long as the light levels at the splicing phase is good, the connection should just would. Although a sanity check wouldn’t go a miss and a quick speed test confirmed I was running on the new connection.

Fibre installed, the analogue side was reconnected with the use of a new mk5c socket, the copper in the new combo cable is extremely thin to the point where you’d struggle to believe it’s a pair. That so, VDSL sync speeds are comparable to the old drop cable, although I don’t think longevity was in mind with this design.

Equipment

As FTTP enters a maturity, the efficiency and cost saving measures improve (depending on your side of the fence).

Early installations had modems which included a PSTN port for analogue phones, and were accompanied by a battery backup unit to keep the phone line active during power failures. Later units dropped the PSTN and battery options, in favour of four RJ45 ports, with the potential to provide 4 separate internet connections via the single fibre and modem.

The new ONT supplied for my residential installation was something new to me, a Nokia G-010G-Q. This unit is far smaller than previously seen, and is sized perfectly to replace in situ a BT analogue master socket. The compromise is from the bare minimum of I/O, with the only connections being power, fibre termination and a single RJ45 port for modulated data.

Results

As expected, my internet connection has been perfectly stable and just what I need when working from home, in particular the upload speed and lower latency is a much-needed improvement for this website, which I host from a home server.

Ironically the install of FTTP has greatly improved the FTTC connection its replacing, now it syncs for days at a time instead of a hopeful 48hour maximum.

If you are like me and insist on using your own router, double check the WAN throughput before deciding on a speed profile. Even though Draytek advertises a 400Mbps firewall throughput, I’m reliably getting a maximum of only 180Mbps. This I would attribute to three active WAN and a couple of VPN all with load balancing options applied.

While there I came to the realisation that I needed to be happy in my employment, something that’s been missing for the past decade, and since I no longer had any loyalty to a company I refocused and started looking for IT jobs again. The late summer was not a good time to be made redundant, due to technology roles being scarce and in high demand.

Roll on to last January and the job market is much more buoyant, I managed to attend a few interviews and very lucky to have been offered a job at a technology company. And it has been the best opportunity I’ve ever had. The role mainly revolves around VOIP services and IT consultation and support, I absolutely love it.

To work alongside likeminded people who are driven by technology and choosing the best solutions available for customers is something I’ve never experienced, but to chat about issues in full geek speak terms before translating into easier to understand language is amazing and I hope it long continues.

The lack of posting comes down to time, where I had an understanding on IT, VOIP sounds like it requires the same operational understanding, but the level at which this company operates is a whole new experience, so most of my spare time has been dedicated to bringing myself up to speed.

So what does the future hold for this blog? It was created as a portfolio to showcase my computing skills in hope to gain an IT job, and now I have one it’s brought up a dilemma I never thought I had to deal with. With the knowledge I’m gaining every day in the new role, it’s a conflict of interest to publish my learnings on here which would be a detriment to the company.

The blog however was set up for another reason, to document and be a reference for those obscure issues that I’ve overcome and seldom crop up. And so this is what the blog will continue to be about, and if this helps others who face that same annoying issue, the better.

So rest assured the blog is here to stay, it’s done one of its jobs, and will remain to satisfy the other.

I’m talking about the Asus X205TA, essentially a low cost, low weight, basic Windows laptop that does the basics. At a retail price of £179 (less with the offers it seems to attract) its price point puts it among lower end Android tablets and Chromebooks. But here the big attraction is its ability to run full versions of Windows, making it a contender as an ultra-portable daily driver.

Leading up to the Asus my mobile computing solutions were an iPhone (screen too small), iPad (Too expensive to leave on display) and a Samsung Galaxy Tab 4 10.1. Focussing on the latter, the Galaxy Tab paired with Android is a versatile environment, doing nearly everything a PC can, but in a more disjointed way that can get frustrating and impact productivity. None more so the when using Excel, where creating even basic spreadsheets takes an age when only having a touchscreen for input.

The X205TA is a small machine, coming with a 12-inch screen that fills up the upper clamshell, and with a weight of 1kg it is definitely a portable machine, swapping a Galaxy Tab 10.1 out for this in my bag made no difference in carrying around. With the size, compromises have to be made to the keyboard, it’s not full size but begin typing and your fingers adjust to the size within minutes, however the difference becomes apparent when using a full size keyboard at the same time, with missed keys becoming common. What I like however is the organisation of the keys, even though it’s been squeezed into a small space, the spacebar and special keys have retained the proportions of a full size counterpart, which makes the transition all the easier. The space gained from the smaller keyboard is filled with the trackpad. I’m not a fan of trackpads in general but the larger size makes it easier to navigate the screen precisely, without multiple swipes to reach each end of the screen. However, with more use frustration arises with the touchpad missing tap to click actions more frequently than is acceptable, and I find myself reaching for a Bluetooth mouse when more mouse work is needed. Multi-touch is supported but as of publication only scrolling seems to be available on Windows 10, it does this well with a smooth action, but that’s all it is.

With a laptop of this price range the specs do not make great reading, actively shopping in this price bracket is a case of finding the best trade-offs and making the best of a bad situation. The Asus has an Intel Atom Baytrail quad core processor clocked at a max of 1.3GHz. Paired with that 2GB DDR3e RAM and 32GB eMMC flash storage, both of which are soldered on the board so the upgrade option is a nonstarter. Combine that with a non-user replaceable battery and there really no point in unscrewing and opening the panel underneath. With a system that simply won’t upgrade, it’s a gamble to pick a machine that will remain useable the longest in its current form.The 12-inch screen has a 1366×768 display that is driven by the Intel integrated graphics, and thankfully the system delivers smooth HD playback from VLC player and YouTube on Chrome. The panel itself is bright even on default battery power settings, but colour fade begins as soon as you move you view from dead on, combined with a glossy screen that will reflect anything in light and it can become hard to see the display without interruption. These issues suggest that the biggest cost savings were made with the screen, but nevertheless a functional display.

The combination of Windows 10 on this machine really helps to create a useable system with such low specs, RAM usage is much lower at boot compared to Windows 7 or 8 which is essential with a hard limit of 2GB. That said the eMMC drive, essentially a SSD allows faster virtual memory when the real stuff runs low, however I don’t understand with Asus put BitLocker encryption on the storage drive out of the box, as it chokes data transfer speeds making the laptop artificially slow.

I thought the RAM on a full Windows system would be the biggest limitation, but with a couple of months use its inevitable that the 32GB eMMC drive will be the first to cause critical issues. From a factory install, Chrome, Office 365 and Visual Studio Express Web was added and now free space on the drive is down to 6.5GB, which with the constant bombardment of Windows updates is sure to reduce even further. There is a Micro SD slot available to increase space, but looking a year or two into the future I’m sceptical that the full 32GB space would be enough for Windows alone.

Despite the low specs, it performs very well at the basics, it can handle a couple of Google Chrome tabs while running Word or Excel at the same time, can even manage Visual Studio Web and a Chrome tab simultaneously without too much screen refresh lag when switching between. What makes it a great system for me is that I can carry a full Windows system around ready to go wherever I am, ready to solve simple issues quickly that would have been tedious on an Android or IOS device. Plus, with a battery life that is truly on par with a tablets and an iPad it can be on standby for days and still be of use when needed, or as a workstation all day away from the mains.

A Windows laptop that can run all day on batteries was once the realm of expensive Ultrabook’s, now its achievable at a fraction of the price. Of course the lower price will mean compromises, but if a willingness to work within the limited specifications is achievable or the high performance of an Ultrabook will go to waste, this is a great machine.

With such a small form factor it is going to be compared to Ultrabook’s, and that’s where it’s at a disadvantage as the X205TA is simply not designed to compete with highly engineered machines that have high prices to suit. I would like to say it’s more of a netbook, but that also doesn’t do the Asus justice as the word netbook has been sullied by the mid 2000 machines that were too small and woefully underpowered to do anything on an operating system not suitable for them, looking at Vista here. I would like to think of the Asus X205TA as a new generation netbook, or an example of what a netbook should have been.

I am using the 64bit build 10074 available publically via the windows preview site, and with the release date of 29th July just being announced, it shouldn’t be far from complete.

First up, starting it in VMware and logging in saw a spike in my host’s machines memory, true to form the new Windows is looking to be a memory hog with 1.1GB taken on boot, not bad but this is a clean copy without any 3rd party software and the inevitable bloatware that PC manufacturers will inevitably add.

A few minutes of looking around the desktop and i quickly get the feeling that its an OS I could live with daily. Without showstoppers that I had a few years back when trying the Windows 8 preview (a false sense of security and then stumped at trying to find the shutdown menu).

Luckily the basic layouts inside the settings are more of the same, with networking options staying the same since XP, however getting to them was a new experience as again the settings menu has had a shake up to try and make it more user friendly, but only alienates seasoned visitors to the control panel. Also to note for some reason the monitor and personalisation settings have been separated in this release, and Computer Management has been banished to only the Administrative Tools folder in the Start menu.

Moving on to the next new feature, Edge browser or as its still called in this build Project Spartan is the replacement to Internet Explorer. I’m not sure if Microsoft will ever get rid of the stigma that surrounded it’s attempts at browsers but on first look it is a much cleaner, almost sparse interface, borrowing very heavily from Chrome, even the drop down menu for past visited sites has been removed, the only thing I missed when converting from IE to Chrome.

Testing involved visiting YouTube and for a visualized machine it did rather well, apart from a few seconds of choppy video and audio while Spartan took CPU load to 100% of the two borrowed threads from a Gen 1 Intel Core i7 I gave Windows 10 it played flawlessly. That goes for the whole system too, animations were smooth and a neat trick i noticed that opening a program (sorry, app now) it would show a window instantly, filled with an icon while it loads, this stops the unsure wondering if a click to open a program was successful, and I’m sure will pay off on low powered devices the Microsoft appears to be targeting here.

Overall this build is impressive but this close to release its by no mean finished with bugs cropping up. In my case drop down menus rendered text incorrectly, the Photos app plain didn’t work and I hope the system icons were not final as they look like they were created in a late 90s copy of Macromedia Flash.

I have stuck with Windows 7 all this time and would I use my free upgrade option that Microsoft pushed to all Win 7 and 8 machines? Yes, but not immediately. Thats not based on the preview (although the bugs are not putting me at ease) but that a new OS always has problems at launch that need resolving. I recall XP at launch could have folders deleted via a URL, and Vista’s abysmal file transfer speeds.Its best for others to find the catastrophes before you do!

The HP Microserver I am running has four HDD bays all populated with 1TB drives in RAID 1 configuration, so to increase capacity I had to replace two of the disks. I went for two Western Digital Green 2TB disks as the 1TB variants I currently had proved reliable.

Moving the data over to a new disk would be tricky, the simplest solution would be to put the old drives into a USB caddy and plug into the server, but I had previously found that USB drives don’t like RAID. But since RAID1 means redundancy, I could tackle this another way…

Step1:

Know your hard drive configuration both in Windows and physically, the disks may have been installed for years and the order you installed them may be long forgotten. The HP Microserver is a compact unit so it’s impossible to follow the SATA cables from the drive to motherboard, so an online search did the trick.

Step2:

Knowing the hard drive layout, it’s time to break the RAID mirror by removing a disk and replacing with a larger one. Open the Computer Management window (Right click My Computer and choose Manage) then click Disk Management.

On this system I will be upgrading the volume labelled Data, therefore either Disk 2 or Disk 3 can be pulled out, after shutting the system down of course.

Step3:

After a restart to swap the drives Windows will detect the new blank drive, so a visit to the Disk Management console will greet you with this:

Follow the wizard to create a new BASIC partition, perform a quick format in NTFS, assign a drive letter, in this case I kept the label as New Volume.

Step 4:

Time to copy the data to a new drive, I prefer to use the xcopy command line as once run it does not interrupt and stop for user input such as copying system or read only files.

xcopy A: B: /s /c

Where A: represents the old data full drive, and B: the new empty one. Here the /s switch copies all sub-directories, the whole drive in the case of this command, and /c ignores errors that would interrupt the transfer.

The benefit is you can set it to run in the evening and it will copy unattended all night, great as it will take a while for a full 1TB drive.

Once completed its worth looking at the drive space to make sure it’s all transferred. At this size it may be a Gigabyte out but this is good enough.

Step 5:

Shut down and remove the final remaining old drive, and replace with the other new drive. Upon starting up again the Disk Management now looks a mess:

Not to fear, this requires a little housekeeping. Firstly remove the Missing drives that represent the drive you removed as the wont need to be represented on here anymore, right click on the missing drive’s volume and choose Delete Volume.

Then right click on the Missing drive icon and choose Delete Drive.

Repeat for the second missing volume and the old drives are no more.

Step 6:

Finally it’s time to mirror your new drives, this can be done the same as when your first mirror was set up. Right click the icon for the New Volume, choose Convert To Dynamic Disk

Then on the Simple Volume right click then Add Mirror, in the pop up dialog select the other new disk, identified as the one with all the Unallocated Space. Allow many hours for the mirror to rebuild and everything is back to normal, with added free space.

When studying GCSE IT at school, the first chance for me to use a computer at school, the first programming centric software used was Winlogo. A simple drawing application that used code to manipulate a turtle (yes, I kid you not) to create vector shapes on a screen. Basically an overly complicated version of paint.

I used Winlogo on the at then school standard Acorn A3020, a RISC based 12mhz processor machine, with my final project code taking a full 3 minutes to run on. Luckily at the time I just got the internet at home and managed to obtain a copy of Winlogo for my first home PC. Running on an AMD K5 133mhz the same project code ran in mere seconds.

And this comparison is where I think my love for the Raspberry Pi came about, even though I had the ability to create something that ran quickly and get the same results. I preferred running it on a slower machine as it felt that my work had more prominence, taking up many more processor cycles on the older machine than new. Plus watching my code run at a pace I could read and understand helped me relate to it and establish what I was doing and, in the case of Winlogo where I went wrong, seeing which vector took the wrong path as it happened on the A3020 rather than trying to fathom out from the final scribbled mess that my PC displayed.

I relate this to the current methods of computing I have, if I want a new system to learn like Debian, I could easily create a Virtual Machine on my laptop or launch an instance on Amazon EC2. But I have been drawn to the Raspberry Pi firstly due to it being a physical machine, like the A3020 used in the past, even though it was not the fastest or superior machine available I could manipulate it to use its entire capacity on what I choose. And the slowly scrolling status messages help me to understand what’s going on when I type in commands, essentially helping my learn better.

Not sure if this post places the RPi in a positive light, or myself for that matter,  but it is a great learning tool.

The post before that however had more gravitas, with my server being offline for close to a month all due to a one second power cut made me feel vulnerable to another downtime incident over something I couldn’t control. It was time to look into a Uninterruptable Power Supply to protect my server from power cuts that could knock my RAID out of sync.

My purchasing requirements were as cheap as possible,  but had to be an online system, meaning that the UPS would notify the server when it was running on batteries and shut down the system down before the juice ran out. An offline system would be less inexpensive, but a long term power outage would just postpone the problem, plus since I don’t intend to supply backup to the network hub, I would be powerless (on both levels) to do anything about it.

I settled on the APC CS350, a lead-acid unit with serial and USB connectivity. Making thing simple I connected the UPS to the server via USB and Windows 2003 found it without extra drivers, making it as familiar to use as a laptop with its battery management.

After allowing the battery to charge, testing is easy, switch off the mains and see how it does. Even from this modest unit my HP Microserver in idle gets 50 minutes of extra life before the battery runs out. I set Windows to shut down at 25% charge, more than enough to shutdown, given that during a power outage the router would be down so the server would be unable to communicate with anything, data preservation is the main game.

The UPS cost £46, my preferred WD Caviar Green 1TB drive costs £48, if another power cut occurs then my it would of paid for itself, another cut and its paid for itself. Yes?

Amongst other web applications, WordPress is the most popular and it makes the install on Windows as simple as a few clicks.

So far I’ve tried it on Server 2003 R2 and Server 2012 R2, both with instant success.

It’s available at: http://www.microsoft.com/web/downloads/platform.aspx