2017 Spring Budget - fibre optic broadband & 5G

Budget day!

So, the two key areas for telecoms in this budget are:

  • "Full-fibre" broadband, aka fibre to the home (FTTH) or premises (FTTP)
  • 5G, the next big thing in cell phone technology

The Government says it wants to "support the market roll out of fast and reliable full-fibre communications for consumers and businesses", which is on one hand excellent news as it's now on the national agenda, but on the other hand.. what does "support the market rollout" actually mean?

The budget statement refers to making "£200m available to fund a programme of local projects to test ways to accelerate market delivery of new full-fibre broadband networks." Full-fibre broadband is where the optical fibres terminate at the customer's premises - which means much faster speeds can be delivered on a more consistent basis to everyone, regardless of where they live.

So who should get this £200m? Should BT? Well, they've just spent £1.2Bn on football instead of investing in their network, so maybe they shouldn't qualify. In any case, their fibre to the home technology, where they provide it, is quite different and slower (300Mbps/30Mbps) than other companies' services but I believe they have plans to change this.

The remarkable community broadband project B4RN could do an awful lot with just £1m, as they quietly go about delivering 1000Mbps or 10,000Mbps (count the zeros - there's no typo) symmetrical services in the rural North West of England. Similar groups such as B4YS and B4RDS using a cookie-cutter style model could also benefit.

Of the more traditional service providers there's Gigaclear, that focusses on rural fibre optic broadband, Gigler based in Portsmouth, and Hyperoptic that focusses on upmarket MDUs, and Cityfibre that's not so much of a service provider but more of an "open access" infrastructure provider for other service providers.

All of these have skin in the game and would be deserving of assistance.

Instead, the budget is suggesting building on public sector assets to reduce risk and costs - shared infrastructure. It's not immediately clear how this will help national rollout of fibre to the home, and we've been in the shared infrastructure space before with PIA (Physical Infrastructure Access) which didn't work out too well.

5G is often touted as the panacea for all telecoms ills. It promises much, but so did 4G and we haven't really got 4G yet either. 5G uses higher frequencies to deliver faster bandwidth but those pesky laws of physics mean that the signals go less distance and have less building penetration capability. So more towers have to be built with more base stations so the signalling system has a chance to deliver its asymmetric 100Mbps/50Mbps service.

This will require a lot more fibre optic "backhaul" to cope with the increase in bandwidth, and all of this won't come cheap as there will need to be tens of thousands of more base stations around the country. Given that our "4G" services already deliver 30Mbps to 50Mbps speeds - is the cost really worth it?

The budget refers to investing up to £16m (why is everything to do with broadband always described as "up to"?) in "a cutting edge 5G facility with the technology to run the trials, delivered through cooperation between leading 5G research institutions." We already have one of these at a research institution, at Surrey University, and it's well worth visiting.

Personally I'd have liked to see something much more adventurous - assistance to deliver Universal Service Obligation (USO) of say 100Mbps symmetrical and inducement to base broadband tariffs not on a hypothetical speed at the exchange or cabinet, but on real world speeds as they are delivered and experienced by the customer. Oh - and something about fibre tax too - there's a playing playing field that definitely needs levelling.

Anyway, the relevant bits (see what I did there?) from the budget are below - draw your own conclusions.

National Productivity Investment Fund (NPIF)

4.15 At Autumn Statement 2016, the government established the NPIF to provide over £23 billion of high-value investment between 2017-18 and 2021-22, with a focus on priority areas that are critical for improving productivity: economic infrastructure, housing and R&D. This built on existing plans for major investment over this Parliament, including resurfacing 80% of the strategic road network and the largest investment in the railways since Victorian times.

4.16 The NPIF provides the financial backbone to the government’s Industrial Strategy, and will:

  • support market roll-out of the fast and reliable full-fibre connections that will help businesses to grow
  • tackle congestion and ensure the UK’s transport networks are fit for the future
  • enhance the UK’s position at the forefront of technological progress globally
  • accelerate new housing supply

4.17 This will provide a signifcant boost to the UK’s productivity in the long term. The Budget sets out further detail of how NPIF funds will be invested in priority transport, digital communications and R&D programmes.

4.18 Digital infrastructure – The NPIF will invest £740 million in digital infrastructure by 2020-21, to support the next generation of fast and reliable mobile and broadband communications for consumers and businesses. The Budget announces the first steps towards this ambition.

4.19 5G – The government’s 5G Strategy, published today, sets out steps for the UK to become a world leader in the next wave of mobile technology and services. This includes:

  • a new National 5G Innovation Network to trial and demonstrate 5G applications. The first phase will invest up to £16 million in a cutting edge 5G facility with the technology to run the trials, delivered through cooperation between leading 5G research institutions. A new centre of 5G expertise within government will oversee this programme, working with public and private sector partners. Funding for future trials will be awarded on a competitive basis
  • the government’s response to the National Infrastructure Commission’s Connected Future report and recommendations on 5G. This will include developing commercial options for improving coverage on roads and rail, and working with Ofcom to ensure the UK has a regulatory environment fit for 5G

4.20 Full-fibre broadband – Starting in 2017, the government will invest £200 million to fund a programme of local projects to test ways to accelerate market delivery of new full-fibre broadband networks. These will combine the following approaches:

  • bringing together local public sector customers, to create enough broadband demand to reduce the financial risk of building new full-fibre networks
  • offering full-fibre broadband connection vouchers for businesses, to increase take-up of services where new networks are built through the programme
  • directly connecting public sector buildings, such as schools and hospitals. This will bring fibre closer to more homes and businesses, allowing them to be connected
  • opening up public sector assets, such as existing ducts, to allow fibre to be laid more cheaply

4.21 Complementing the NPIF programmes, the new Digital Infrastructure Investment Fund will be launched in spring 2017. Government investment of £400 million will be at least matched by private sector investors, and will accelerate the deployment of full- bre networks by providing developers with greater access to commercial finance.

The Unpicking of the London Underground

The London Underground is one of the most comprehensive public transport systems in the world. Opened in 1863, the network now has 270 stations and 250 miles of track, on a daily basis ridership totals 4.8 million, or 1.34 BILLION annually.

The London Underground is one of the most comprehensive public transport systems in the world. Opened in 1863, the network now has 270 stations and 250 miles of track, on a daily basis ridership totals 4.8 million, or 1.34 BILLION annually.

Quite simply staggering.

However, there’s a problem. And I think this problem is badly affecting the user experience and is also weakening the Underground brand.

As a frequent user of Eurostar I’ve tried every combination of getting to StPancras from my home town of Maidenhead, and I’m really looking forward to CrossRail, or the Elizabeth Line, going live as I’ll be able to get there directly without changing trains.

Outside of rush hour, the mainline operator GWR have deemed that no one’s interested in fast rail services into London, resulting in a sufferingly long journey from Maidenhead. Having eventually arrived at Paddington, there’s the Tube journey to StPancras itself.

The most obvious route is via the Circle line, but the Circle line is no longer a circle, it’s broken. It is in fact a spiral, and this is where I first noticed something was amiss.

The "Spiral Line" 

The "Spiral Line" 

Previously, the Circle line really was circular in nature, no matter which station you started from, you could be certain you’d end up back there again in an uninterrupted journey. I could board a Circle line train at Paddington and be certain it would trundle around its track and arrive at StPancras a few minutes later.

However in recent years, the Circle line has been extended out westwards to Hammersmith running on the same line as the pink Hammersmith & City, and now terminates at Edgware Road rather than running straight through, as shown in my rendition of the Tube map above.

Having arrived at Paddington with heavily loaded wheelie-suitcases suitable for a visit to Paris, I’d either end up on platform 2 or platform 3 at Edgware road Underground station where the Circle line service was more often than not terminated. This was either a minor inconvenience or a major challenge, depending on which platform I arrived at. If I was lucky, it would be platform 2, in which case I could easily saunter across to platform 1 and wait for the connecting train (either Circle line or Hammersmith & City) to St Pancras. 

If I was unlucky, I’d arrive at platform 3 and have to schlep suitcases up the Victorian-era stairs as there are no escalators let alone a lift, cross the bridge and then schlep the suitcases back down another flight of stairs to platform 1 to continue the journey.

I couldn’t help but be frustrated by this nonsensical break in the journey which caused massive inconvenience for passengers, utter confusion for lost tourists whose English wasn’t their prime language, and which resulted in such a poor user experience. 

Then it tumbled, the penny dropped, and I realised that not all Circle line trains are the same. In fact not all Circle line trains are Circle line trains at all.

You can see from the older map inset in the diagram above that the green District line terminates at Edgware Road, and the old “true” Circle line carried on straight on through. This perfectly logical difference has been swept away not just on the map, but also on the trains themselves and it’s this that’s the root cause of the problem.

To explain this we first need to go back in time and look at two of the most influential characters in the development not only of the London Underground as we know it, but also of design, and service design thinking, Frank Pick and Harry Beck.

Frank Pick

Frank Pick Hon. RIBA was CEO and vice-chairman of the London Passenger Transport Board from its creation in 1933 until 1940 and is on my list of top 10 people I’d like to spend time with, were it possible. His influence went far beyond that of the underground electric railway, and his interest in design and branding went deep within the underground.

In 1908, Pick was responsible for the marketing of the Underground Electric Railways Company of London (UERL), where he’d end up as Managing Director in 1918.

Pick had a strong interest in design and its use in public life. He steered the development of the London Underground's corporate identity by commissioning eye-catching commercial artgraphic design and modern architecture, establishing a highly recognisable brand, including the first versions of the roundel and “Jackson” typeface still used today.

Under his direction, the UERL's Underground network and associated bus services expanded considerably reaching out into new areas and stimulating the growth of London's suburbs. His impact on the growth of London between the world wars led to him being likened to Baron Haussmann and Robert Moses.

He was a founding member and later served as President of the Design and Industries Association. He was also the first chairman of the Council for Art and Industry and regularly wrote and lectured on design and urban planning subjects.

Pick's design philosophy was that:

 "The test of the goodness of a thing is its fitness for use. If it fails on this first test, no amount of ornamentation or finish will make it any better; it will only make it more expensive, more foolish.”

It’s this expression “the test of the goodness of a thing is its fitness for use”, combined with his working within a service provider (the London Underground provides a transportation service) that for me, places him among the greats of service design thinking.

Harry Beck

Harry Beck was an English technical or engineering draughtsman at the London Underground Signals Office. Prior to Beck’s now iconic map, the various underground lines had been laid out geographically, often superimposed over the roadway of a regular city map. This meant the centrally located stations were shown very close together and the out-of-town stations spaced far apart, leading to maps that were difficult to read and which adversely affected the user experience of the Tube network.

Beck’s three strokes of brilliance were firstly an understanding that passengers on the Underground were not concerned with geographical accuracy, they were more interested in how to get from one station to another and where to change trains  i.e. the most efficient user journey.

The second stroke of genius was to base his design on the then established design principles of an electric wiring diagram. Don’t forget that at the time he was a technical draughtsman at the Signal’s Office, so this would have been a very natural thing for him to do. Electricity wasn't universally available as it is now, it and its motifs were considered very modern, so Beck's design tapped into the Modernist zeitgeist of the time.

The third stroke of genius, which may seem trivial to some, was to use a different colour for each of the then operating lines. This added clarity to a complex map which in turn aided the user journey especially when linked to the colours of the trains themselves.

Beck first submitted his idea to Frank Pick in 1931 but (in)famously it was considered too radical because it didn't show distances relative from any one station to the others. The design was therefore rejected by the Publicity department at first, but the designer persisted. So, after a successful trial of 500 copies in 1932, distributed via a select few stations, the map was given its first full publication in 1933 (700,000 copies). The positive reaction from customers proved it was a sound design, and a large reprint was required after only one month.

Coming full circle

It’s now 2017 and all of the above is being unpicked, and maybe you can now see the brilliance of my witty blog entry title. How so? How is this being undone?

I mentioned that Beck used colours to differentiate each Tube line on his map, the current corporate colours are defined on the Pantone colour range as follows:

And in practice this is what we see on the trains – the photos below for example show the Central Line on the left, red (Pantone 485) line on the map, red fittings throughout the carriage for further clarification. On the right the Bakerloo line, brown (Pantone 470) line on map, brown fittings and fixtures throughout the train:

But what about this train? (Hint: the "Semi-fast" dot-matrix display is the giveaway for those in the know).

It is in fact a Metropolitan line train:

The Metropolitan line’s colour scheme is defined as being Pantone 235, so a purple colour on the map, but what’s inside the train? Yellow. The same colour as the Circle line (Pantone 116)! 

Psychologists tell us that cognitive dissonance is the mental stress or discomfort experienced by a person who simultaneously holds two or more contradictory beliefs, ideas or values when performing an action that contradicts those beliefs, ideas, and values; or when confronted with new information that contradicts existing beliefs, ideas, and values.

It is of course the job of the service design thinker, or product manager, to remove as many instances of cognitive dissonance as possible to deliver the best possible user experience or smoothest possible user journey - a literal user journey in the example of the London Underground.

So this train is being represented simultaneously as a Circle line train AND a Metropolitan line train - but it cant be both, it has to be one or the other.... no wonder confusion abounds! But it gets worse.

Which Tube line is this?

You can’t tell from this picture – there’s lots of Circle line yellow on the fixtures and fittings, but then there are two other possibilities as indicated by the maps – the District line (Green Pantone 356) as well as the pink (Pantone 197) of the Hammersmith & City line.

And this is why my suitcase-laded journey across London was so poor. I was looking for the most readily accessible identifier of the tube trains as they pulled into Paddington – the interior colour scheme. But as all Circle line and District line trains that use Paddington are now yellow, it’s less easy to tell them apart and easier to get onto the wrong or less convenient train, resulting in a disrupted, less convenient journey. The different lines arrive at different platforms at Edgware road.

Why is this?

Why is the strong link between the colour scheme of the map being separated from the colour scheme of the trains? I can think of only one answer, and that’s the accountant, the finance team. It presumably is cheaper to buy 1000 identical carriages than it is to buy 10 lots of 100 different carriages, and it’s probably easier to manage too, as rolling stock can be readily mixed and matched.

But often it seems to me that accountants know the cost of everything and the value of nothing. How do they account for user experience? Where on the balance sheet does “fantastic user experience” go? Or brand value, for that matter? Pick’s insightfulness and deep thinking about the Underground brand as he created it is being weakened, undermined, and we the users are all the poorer for it.

Remember Pick’s design philosophy:

“The test of the goodness of a thing is its fitness for use…”

This unnecessary disruption to journeys caused by this colour confusion is rendering the Tube’s fitness for use less so. There’s such a strong link between the colourways used on Beck’s map, distribution throughout the trains and the user experience that it seems to me  foolhardy to disrupt it. These physical elements of design are key to the success of the design of the service.

However as it seems inevitable that train interiors will become a standardised, corporatised, aneathetised, homogenised Pantone 116, I humbly present to you, dear reader, my own version of the legendary London Underground map, with full apologies to Beck for my amateurish skills:

Groundhog Day

At the end of Y2K I became an illegal alien.

We’d successfully sold the Motorola business unit, but the new owners didn’t want to sponsor my L1 work visa, so there I was, stranded in Rhode Island, with the clock ticking… 

The serendipitous solution presented itself with an opportunity for a telecoms service provider called IDN, or Inter-Digital Networks. As Marketing Director for IDN, the future looked bright… although by the time I’d landed back with all my goods and chattels, IDN had rebranded to the perhaps more familiar name of Neos Networks.

So, a brand new job and a brand new brand. What were we going to do with this?

The dilemma Neos had was that their business model didn’t provide the investors with the return that they were looking for, and neither did it offer them a way to exit, either by floating or selling. Neos had a London metro network, and sold some SDH. That was pretty much it.

We needed a different business model, and I needed a new car. After much research I ordered through a leasing company an Audi A4 Quattro.

We needed to scale, but to scale up, we needed a national network. The trouble with this was that we’d be very late to market with a data services network – or more precisely with the necessary four networks we’d have to build:

  • An optical underlay network
  • A layer 2 frame relay network
  • A layer 2 ATM network
  • A layer 3 IP network

And all of these would need their own OSS and BSS ecosystems. And all of this came with a pretty huge price tag. Which was all very well as long as there was significant differentiation for us as we were at least 10 years behind all of our potential future competitors, all of which had a proven track record of delivering these services to a demanding and sceptical market.

The order for my Audio Quattro got knocked back as Audi decided to do a model refresh and there would be a bit of a delay, and the car was finally delivered a few months later.

The Solution

Our late entry to the data network services market was in fact fortuitous, precisely because we didn’t have to struggle with the financial consequences of having sunk an awful lot of money into:

  • An optical underlay network
  • A layer 2 frame relay network
  • A layer 2 ATM network
  • A layer 3 IP network

Technology came to our rescue. We realised that the innovation of Metro Ethernet networks could be scaled up to provide a national network, and, when combined with the then recently launched optical LAN Extension Services (LES) tail circuits from BT, would provide a complete end-to-end “optical Ethernet” solution.

The result would be for us a “fully delayed” network, with no legacy technology to complicate either the technical solution or the business case. We could very cost effectively build a national network that offered MPLS and VPLS Ethernet services with bandwidth price point we knew would make our competitors’ eyes water.

And because this new network was entirely Ethernet-based, and so cost-effective, we could provide bandwidth at any speed on 1Mbps increments, from 1Mbps to 1000Mbps. To facilitate this, we provided a “bandwidth dial” on our website that our customers could use to easily turn their bandwidth up and back down again. Not only did we offer fantastic flexibility, the network performance stats were off the chart. We called this Liquidbandwidth™, created a cool logo and thereby a sub-brand, and watched the market react.

This was radical stuff in telecoms back in 2001. 18 months later we’d signed up 59 LiquidBandwidth™ Ethernet customers, with on-net users averaging 92Mbps per order and off-net customers 54Mbps.

We were so successful, we delivered the exit our investors wanted by way of a trade sale to SSE, the energy company.

Back to the future

In Autumn 2015 I became unemployed.

After sucessfully selling Neos to SSE, for a number of years I went freelancing and freewheeling. I raised a small fortune for a webTV venture in a bid to become an internet millionaire, but we didn’t make it. I spent a lot of time in Paris and developed a deep appreciation of the art of the patisserie.

In 2015, I successfully completed a brand refresh for a small IT services company, driving their page rank from the deepest of dark depths to the sunny uplands of sunlit Google visbility. The project ended and there I was with freshly printed P45 in hand, with the clock ticking...

The serendipitous solution presented itself with an opportunity for a telecoms service provider called SSE Enterprise Telecoms. Nigel Pitcher, my old sparing partner from Neos’ No1 competitor Fibrenet, and now SSE Enterprise Telecoms’ Marketing Director, suggested we meet for coffee, cakes and a chat.

And here I am, back at Neos, I mean SSE Enterprise Telecoms, surely the very definition of karma. Some old Neos hands are still here, which is lovely, and it’s interesting to see what SSE Enterprise Telecoms have made of things since I was last here. Our network is larger, much larger, as is the size of our customer base, our revenues and market share. What was sexy and new has become passé and established. Ethernet services are de rigour, everyone offers them.

But here’s the thing. Our investor, our parent company, SSE is a £14 billion market capitalised, A+ credit rated and FTSE top 35 listed energy company, that sees us as being a high growth engine for SSE Group and is prepared to invest to enable that growth.

But to really grow we need to enter a new market, and that means creating new services. And for that we need a different business model, and I need a new car. After much research I ordered through a leasing company an Audi A3 Quattro.

Today most of our revenue comes from the wholesale market. We offer dark fibre, lit fibre and fantastic Layer 2 Ethernet services, using MPLS and VPLS, building on what we started all those years ago at Neos. Even so, that’s pretty much it.

How are we going to create new services and what will our new business model be? We believe technology will come to our rescue.

The buzz in telecoms this time around is all about Network Function Virtualisation (NFV) and Software Defined Networks (SDN). We realise that virtualised networks with workflow automation will enable us to create unimaginably exciting new flexible services, that are easily turn-upable and downable.

Unlike many service providers, our network is still really clean. It doesn’t have an eclectic legacy of conflicting technologies, a patchwork quilt of various vendors, a jumbled bundle of BSS and OSS infrastructure. We believe we can cost-effectively build a whole new style of telecoms service provider, with the emphasis very much on service and customer experience.

This is radical stuff in telecoms, as telecoms completes its transition to commodity IT-based services.

Oh -  the way - I got an email the other day from the car leasing company. My order got knocked back by Audi as they decided to do a model refresh

 

Hello

For those that don’t know, “Hello” is hallowed by the MacFaithful, as it was the first word written on a Mac screen in public back in 1984.

So for Apple to publicise an event “Hello again”, made our MacGlands swell to bursting with anticipation for yesterday’s launch. 

However, instead of an insanely great product all we got were MacBookPros that were, as a very dear friend said, “a bit faster, a bit thinner, with a bit of an iPad glued on”.

Yes. It’s true. A desperate need to be seen to innovate, while desperately not turning the Mac into an iPad, has led to the TouchBar, a programmable bit of touchscreenery where the practical function keys used to be.

This, combined with a supersized haptic touchpad produces an "amazing experience", which Uncle Phil says we’re going to love.

Well, we might do but when the entry 15inch model comes in at over 2,000 of her Majesty’s finest, it’s an experience a fair few of us won’t be loving for quite some time.

Microsoft, having copied everything Apple have done for the last 35 years, have finally conceded Apple’s business model is the superior one and are starting to make the product and the OS and the apps. This resulted in them, the day before Apple’s event, launching the Microsoft iMac - basically a very large tablet that swivels up to be iMac-like and swivels down to be a large table-like tablet. To help use this, they’ve come up with a natty puck-like device they call the Dial.

I tell you, it’s a pucking sad day when Apple get out-Appled by Microsoft.

This Joy of Tech cartoon just about sums it up:

I'm told there will be new iMacs and MacPros with us by the end of Nov. If this is the case, Apple have a fight on their hands.

The Philosopher's Macs

Some years ago I came across a philosophical conundrum known as “The Philosopher’s Axe”. The Philosopher’s Axe poses the question “If over the course of use, the head and handle of an axe is replaced, is it still the same axe?”.

You may have come across this in the brilliant comedy series “Only Fools and Horses", in a sketch called “Trigger’s Broom”, click on the following link for your delectation and delight:

https://www.youtube.com/watch?v=BUl6PooveJE

While googling this, I discovered this idea goes back even further than Delboy and Rodders. Here’s an extract from the Wikipedia entry:

“The ship of Theseus, also known as Theseus' paradox, is a thought experiment that raises the question of whether an object that has had all of its components replaced remains fundamentally the same object. The paradox is most notably recorded by Plutarch in Life of Theseus from the late first century. Plutarch asked whether a ship that had been restored by replacing every single wooden part remained the same ship.

Centuries later, the philosopher Thomas Hobbes introduced a further puzzle, wondering what would happen if the original planks were gathered up after they were replaced, and used to build a second ship. Hobbes asked which ship, if either, would be the original Ship of Theseus.”

Now that second paragraph really bakes my noodle.

I’ve been using Macs since 1985 and in that time across all the systems I've used, I’ve had two internal power supplies fail, one motherboard fail and a disk drive fail. My current MacBook Pro celebrated its 5th birthday recently – but is it in fact the same Mac that I bought?

Last year, during the summer, the graphics card failed. Research at the time showed that Apple had built a batch of Macbook Pros with faulty cards and were running a recall program, but my Mac fell outside of this program, by virtue of it not being old enough. 

Rather than paying the £475 ex VAT repair bill, I played the EU consumer law card, which basically says that if a product unreasonably breaks or fails within 6 years of purchase, then it should be repaired or replaced by the manufacturer (something to possibly consider in the in/out/shake-it-all-about debate). Apple played ball, and the work was done by an Apple Dealer called Stormfront in Windsor. When I collected it, the store manager made what I thought was a strange comment “We had to replace the whole lid assembly as well as the screen”…

… an odd thing to do I thought, but that’s half a new laptop, thanks very much!

Three weeks ago the screen froze on my Mac, even the seconds on the clock were stuck in time. I rebooted, but half way through the boot sequence, the Mac booted into a completely grey screen. It did the same thing when booting into “safe mode”, and even from “internet recovery” and even more worryingly, from an external bootable hard drive.

Beth, the MacGenius at the Genius bar in the Apple Reading store diagnosed the problem, checked the notes from Stormfront, and concluded that while they had indeed replaced the screen and top lid assembly, they hadn’t actually replaced the faulty graphics card, which is what I took it to them to do! (quite how it's been working since then I have no idea).

No matter, Beth reassured me, Apple were now running a quality program for my vintage of MacBook Pro, and she duly booked it in for repairs – which consisted of swapping out the motherboard, the graphic card being an un-swappable subcomponent.

A week later, I’m happy to report, it was returned in full working order with a delightful £0 fee.

During September of last year, I’d cracked open the casing and fettled my Mac. I increased the RAM from 8GB to 16GB, and swapped out the original 128GB SSD for a 500GB SSD. As I no longer need to burn DVDs, I replaced the internal optical drive with a second internal disk drive that acts as a Timemachine backup drive, the optical disk drive going into an external USB enclosure for the occasional times I need one. 

All this activity means the number of original parts is few: the cooling fan, the keyboard and trackpad, the battery and the aluminium “unibody” enclosure are all that remain of the Mac as it rolled off the production line. It is in essence a completely new Mac, ready to go for another five years.

But it’s still my much loved Mac.