Deb Bassett is a freelance Rails developer based in Otley in West Yorkshire. She is a mainstay of the Leeds tech scene and the organiser of the Leeds GeekUp. I interviewed her the day after the inaugural Geek-Up in Leeds, and a couple of weeks before her marriage to Rob Lee, who is also an Otley-based developer.
CS: What’s your background?
DB: I’m originally from Stone in Staffordshire. I did a University degree at Manchester in Computer Science – I had applied to study Physics but I changed my mind at the last minute after realising my forte was computers! After finishing my degree, I moved to Leeds for my first job, as a software engineer for PA New Media, which became Ananova (remember – “the world’s first virtual newscaster”?) and was 
eventually bought by Orange. I was with them for three years, mainly developing in Perl, with some Java and PHP in between. I then went to work for Energis, which is now Cable & Wireless, as a full-time Perl and Java developer. After 4 years there, I left and went travelling for 10 months with Rob, and we have both freelanced thereafter.
Things moved on while I was travelling, and I picked up Ruby/ Ruby on Rails when I got back – this time I had the choice of where to take my career and Rails seemed to have everything I wanted in a framework.
The Rails community has grown substantially since I first started, and are such an enthusiastic bunch of people. The amount of Rails applications in production has increased substantially too since I started and is steadily creeping into the mainstream.
CS: How do you find working from home, when there’s two of you doing it, and your personal
life is there as well?
DB: It’s really good, I enjoy it! Sometimes it can be difficult to separate home and working life when there’s no clear boundaries for when you should start and stop working. You sometimes find yourself working the whole day and into the evening until quite late, and sometimes your cycles move round so, for example, you might work from 12 noon until 4 in the morning: it doesn’t really matter when you’re working from home what times you work. But there’s two of us, and it’s actually very nice to have the support – I don’t think I’d like it if I was on my own.
Otley’s a great place to work from – it’s a small town in the countryside just north of Leeds and we’re lucky enough to have some good friends there. There’s quite a tech community which mainly stems from my ex-employers Energis (Cable & Wireless) and Orange – most of us have worked at one or the other at some point and we are techie to varying degrees.
I occasionally head over to York to see Edd who I work with on his exciting start-up, Expectnation.
CS: How long has GeekUp being going on in Manchester?
DB: I believe it started about a year and a half, maybe two years ago. It is the brain child of Andrew Disley, a web developer from Orrell, near Wigan. I first found out about it at Reboot in May last year where I met Andrew, and it had already been going a good while then.
CS: And at what point did you think you could do one of those in Leeds?
It’s an hour and a half from Otley to Manchester and we’ve done the journey quite a few times now. It seemed a natural progression for Leeds to have a GeekUp, a bit closer to home – we had talked about it a few months before with Andrew and others but we were a bit tied up because Rob and I were getting married (they got married in July – ed) so we said we’d leave it until after then. We had
underestimated how vibrant the Leeds tech scene was, and by May lots of people were asking about Leeds GeekUp, so Andrew asked us if we could start it a bit earlier, so we kicked things off in June.
CS: Andrew seems to be an Agent Provocateur …
DB: I like to call him the father of GeekUp, although I’m not sure he’d like me saying that! He’s a nice guy I don’t think it would have the GeekUp edge without him.
CS: While I think it’s great there are User Groups for Linux, Perl etc, I think it’s important that GeekUp is not just about one technology. If someone is setting up a Start-Up, they need to be able to program, to need to be able to sell themselves, so there’s marketing skills involved, they need to be able to set up a website so there’s web design, there’s the money side of things …
DB: GeekUp is great because you get to meet a mix of people from the industry who you wouldn’t necessarily normally meet. For example, I get to meet programmers fairly often, but, say, designers – I’m not really a designer so getting the opportunity to meet others from that community is invaluable. Also, if you are a freelancer, working from home, GeekUp provides that important connection to the community that you might otherwise get from a work environment. I end up picking up lots of sound-bites of information, and reading about them in depth later… it’s an efficient, effective way of catching up with the community.
CS: So last night was just conversation?
DB: Pure conversation.
We’re hoping to have talks next time, we haven’t quite decided on the format, maybe 20:20 Lightning Talks. The Manchester GeekUp last month did 5 quick sessions, and people could talk about anything techie. It’s nice to have that mixture of socialising and formal talks.
CS: I think it’s important to leave spaces at these events so people can have a gab.
DB: Yes, It’s a well known fact that the best conversations happen in the hallways at conferences, we think that GeekUp can learn from this.
CS: It’s a way of people getting excited about what’s going on. There’s a real buzz these
days.
DB: There’s a complete buzz at the moment – it’s fascinating. Although it worries me a bit, I’m hoping it doesn’t match in behaviour to the original dot-com boom.
CS: So how many people turned up last night?
DB: More than 40! Initially we were expecting around 10 people as we didn’t realise the Leeds tech community was so big so we were really pleased!
CS: And beside the Rails guys you were with, who else came along last night?
DB: There were lots of web developers and designers plus a few from PR and marketing and some bloggers and entrepreneurs.
CS: How many people knew each other?
DB: I knew quite a few people, as I’d got in touch with them and invited them along, but there were lots who I didn’t know – groups of people from certain companies that came together and then there were individuals who just turned up and mingled, so it was a complete mix. It was great to see lots of people from the Round Foundry Media Centre too – I’d never met any of them before.
CS: Did everyone volunteer to come again?
DB: We had lots of positive feedback, there was even talk of a barbecue next time.
CS: There seemed to be lots of people writing on Upcoming how they were pleased this was happening in Leeds. Has there been anything happening in Leeds like this up until now?
DB: WYLUG has been around for some time now and there’s been quite a few flickr and blogger meetups. More recently OpenCoffee has started in Leeds – the first one was last week. It’s during the day so hard for me to get to, but they had a really positive turnout with 30+ people turning up in Starbucks (I think next month they are changing the venue). There’s going to be a BarCamp Leeds very soon, they haven’t set a date yet, but it could be late September or October, it’s a really positive step for Leeds. Just last month there was a Barcamp Sheffield. It’s really great to see so much happening up north.
There’s also going to be an Open Street Map mapping party in Leeds on 15th & 16th September. The organiser, a guy called Tim Waters was at GeekUp last night. Leeds isn’t very well mapped, although it’s getting better: I went to the Sheffield Mapping Party and I’m really pleased we’re having one in Leeds as it’d be great to see more of Leeds on the map.
CS: They were a courier company at first, weren’t they, using GPS to keep track of where their couriers were?
DB: They used a courier company to get lots of the initial GPS traces. It’s really fascinating: have you seen the animations of the OSM coverage growing in London and the UK? It really excites me, over the last year I’ve been following it, I’ve watched the maps grow rapidly in coverage, the mapping completed by regular, open source-minded people.
CS: How far a-field did people come from last night?
DB: Some people came from over 30 miles away including some of the Manchester GeekUp lot. Andrew didn’t make it – he really wanted to but he’ll be there next time. I’d love to see Leeds, Liverpool and Manchester have a combined GeekUp gathering one time, it would be great to get everyone together.
CS: So when’s the next GeekUp Leeds?
DB: GeekUp Leeds is every third Wednesday of the month. Manchester is the second Tuesday of the month and Liverpool in on the last Thursday!
CS: That’s quite an algorithm to work out!
DB: There’s method there somewhere. One thing I noticed was that there seemed to be less Mac users than I expected. Every time I go to a conference there’s whole mass of Mac users. There seemed to be a mix of Windows, Linux and Mac users.
CS: Are you a Mac User?
DB: Linux. Ubuntu.
CS: And is it going to be a geeky marriage, if you don’t mind me asking?
DB: Quite! We are using moo cards for the place names, Rob’s generated everyone’s name using images of the letters from flickr using the same principle as flickr spelling. I wrote a guest management system for the wedding in Rails, too. We also are planning on having a wii room at the evening do.
CS: Weddings are well served by the internet.
DB: There’s certainly lots of wedding-related content out there, it’s a vast market! People used to put disposable cameras on the tables, but now there’s no need, you can tell everyone to upload their photos to Flickr and tag them with a specified tag. That way, everyone gets to see everyone else’s photos. Our gift list is online, too – we are using kaboodle (didn’t get chance to write our own!) – so that we could list gifts from different places and not be tied to a specific store.
CS: Do you mind if I ask about women geeks? Someone was saying that Hack Day was predominantly male, and I wondered if you see things changing and what would make things change.
DB: I don’t see things changing that much, which is a shame as it’s a really interesting and creative industry to work in. I do find it a difficult question to answer as it seemed a natural progression for me to head into computing and I have found the computing industry really welcoming as a whole.
Of course, you are aware that you are a minority, and there are both advantages and disadvantages to this. Jeni Tennison has written a fabulous article which is the first article I’ve read that really hits the nail on the head for me: Getting Women into Computing. Edd wrote a follow-up to Jeni’s article – of particular interest to me is the point made about self-efficacy – as Edd says, a complete revelation, and something I’ve never quite been able to put into words.
CS: Are there any other interesting events in the North that you go to, and further a-field?
DB: Rob and I head down to Manchester as often as possible for North West Ruby User Group, run by Dave Verwer. There’s also a group in York called We Are The Monkeys – they’ve been going for some time now and meet on a monthly-ish basis. In July, Manchester will host “The State of the Map” – the first Open Street Map international conference (http://upcoming.yahoo.com/event/153405/). Unfortunately it’s on the same day as our wedding otherwise I’d be there! Also in Manchester is rumour of a Barcamp – I’m looking forward to hearing more about that.
Further a-field, we headed down to Interesting 2007 in London where we manned a stall for folksy, the start-up that Rob has been working on with a guy called James Boardwell. Folksy is an online market place to enable crafters to sell their stuff but with a social, Instructables slant. I helped them with their stall where we showed off our hardware hacking and crafting skills, which went down really well. It was cool to see people making stuff in person at an event, it doesn’t normally happen. There seems to be a movement towards making and crafting at the moment, it’s great to see people getting out their soldering irons and sewing machines and innovating again.
In May, we were lucky enough to head over to Paris for XTech, chaired by Edd Dumbill, the theme this year being “The Ubiquitous Web”.
In May, we also travelled to Copenhagen for Reboot, ‘a community event for the practical visionaries who are at the intersection of digital technology and change all around us’, and last September we were at RailsConf Europe, in London.
CS: What are the major employment opportunities in Leeds for a techie? What kind of work is it? How do you find freelancing? Is it a good place to be a start-up?
DB: There are fairly good employment opportunities in Leeds covering a wide range of skills. Larger companies such as Cable and Wireless, Orange, William Hill, HSBC and the Halifax have a large IT presence here and there are also many web design and development agencies too.
In terms of freelance/contract opportunities, I’ve not seen many advertised Rails positions, however the more mainstream skills are well represented. I’ve normally found that freelance work comes via word-of-mouth and it matters less where it is based as you generally work from home.
In terms of being a good place to be a start-up – I don’t think Leeds is any different to any other major UK city – nowadays I think it matters less where you are based, as long as you are close to major transport links, and have good connectivity!
CS: Are there major tech events in Leeds or and how do they impact on the daily techosphere?
I’m sorry to say I can’t think of any major tech events in Leeds other than the impending BarCamp. Maybe that’s something that GeekUp can help resolve!
CS: Who are the major bloggers and characters?
DB: The major characters and bloggers that I know of in the Leeds area are: Tim Waters, geo-specialist and developer, Edd Dumbill, writer, programmer, entrepreneur and free software advocate, Imran Ali, emerging technologies specialist and organiser of Leeds Open Coffee and not forgetting Rob Lee, fiancé, developer and entrepreneur!
I was at Manchester Polytechnic between 1986 and 1989, playing about with online chemistry abstracts and yet not quite realising the potential that lay at my fingers. Manchester is a great city to live in, with the facilities of a major conurbation but the feel of a small town, and it has a fantastic tech pedigree. The city is a vibrant metropolis at the heart of the North-West and though it’s much gentrified since I was there, it has a busy grass roots that goes out of its way to make things happen.
Andrew Disley and Paul Robinson are two players on the Manchester geek scene who are making real efforts to bring people together. Andrew started the GeekUp franchise, which has spread virally to the mouth of the Mersey and over the Pennines to Leeds. Like Andrew, Paul has quickly evolved from being purely a developer into an event organiser of note, co-founding User Groups and Community Groups because nothing at the time provided the forum that he needed. Both are strong advocates for the city, the North-West and the North as a whole, championing the cause of and encouraging the participants in the multitude of activities that take place in the towns and cities either side of the M62. These are guys with a world outlook but a local base and they move seamlessly from picking through the minutiae of a block of code to taming the maelstrom of a Northern tech meet.
They were good enough to answer a few questions about Manchester and the North-West.
Who are you? What do you do? What are you aiming to achieve?
Paul Robinson (PR): I’m Paul Robinson, the owner/founder of Vagueware Ltd which is a software development company and soon-to-be online publisher dedicated to innovation in the software industry. I’m loosely attached to GeekUp but more involved in the Manchester BSD User Group and the NWDC. I worked on public sector stuff a few years back, and before that was working in the local ISP scene. I graduated from UMIST (now part of Manchester University) in Software Engineering and have lived in the city centre for the last decade.
I’m a kind of a developer with strong sys admin skills who wants to get more into writing. ;-)
In more broad terms, my main business is Vagueware Ltd. It started as a software company building Rails apps freelance and doing bespoke work, but I’ve evolved my business plan in public. I’m now a few days
away from releasing the latest version of that plan, aka “the simplest thing that will work”, which is best described here: http://blog.vagueware.com/2007/8/5/almost-there
My driving force is that there isn’t enough innovation in the open source space, and developers aren’t plugged in to listening to their user base. For now, I want to get the 400+ ideas for software products on my desk/in my head out there for somebody to be able to play with. If in the process we build ad-hoc open source projects or even commercial businesses off them, great.
Some people ask how I intend to monetise this, to which I shrug my shoulders and say “I’ll work it out eventually”. Like I said in my last post: doing interesting things appeals more than doing something boring with a guaranteed return. Or maybe I’m just dumb.
Manchester seems a great place to be a techie at the moment. Has that always been the case? Does Manchester have a history of encouraging technology in general and computing specifically?
Andrew Disley (AD): Manchester is said to be the birthplace of modern computing, Alan Turing was based at Manchester University and his stored-program idea led to Tom Kilburn’s “Baby” (Manchester Mark I) – the world’s first
stored-program electronic digital computer.
PR: Wow, where to start…
OK, first off Turing came here after the War and worked at Manchester University on “The Baby”, the World’s first stored program digital computer. The University is justifiably proud of this, and built a replica of that machine to mark the 50th anniversary: http://www.computer50.org/
Between 1948 and until at least the mid/late 1950s the University was one of the few places you could buy a computer if you were a government type in need of one – it was really a pioneering centre of innovation in the early days. If you have access to the BBC Archive Trial there’s an interesting docu in there called “The Brain in the Box” in which Profs Kilburn and Williams are interviewed.
As you probably know, Manchester went through a period of economic decline in the late 50’s through to the 1990s with much of the North in decline. During that period there was a focus on industrialisation and Manchester kind of lost its way and ended up in a bit of trouble.
In the early 1990s there was some interest in digital media in the city, but it was mostly coming from the angle of “new media” – i.e. web design, interactive content, etc. There is an argument this is related to the rise of the Northern Quarter (where even now most web companies are based), which in turn was helped by the rise of the music scene in Manchester.
What helps this argument is that whilst most centres of digital innovation (Silicon Valley, Taiwan, China, Estonia, Israel, et al) are driven by the economics of the sector, Manchester is pretty unique in that it’s driven by the creative possibilities. We need to make money, sure, but doing something interesting always seems to have appealed to the Manchester companies more than doing something outrageously profitable but boring.
After the ’96 IRA bomb which required a major rebuilding of the city centre, there was a renewed sense of purpose in the region and the North West Development Agency (the NW’s RDA) combined with ERDF giving the region Objective 2 status meant the council and several Universities were able to throw money at a purpose. This was described to me as “To make the North West of England the Silicon Valley of Europe.” I was involved in one project at MMU to help businesses engage in digital media and to try and promote business development within the sector: help little web companies find their way to breaking big and dominating in their niche. That experiment broadly failed in immediate terms, but the knock-on effects have been considerable. A lot of companies in Manchester now know of each other and look out for each other as a direct consequence of that project.
In addition, the council set up MDDA which is still running. They’re more interested in helping disadvantaged parts of the community gain access to technology but have a broad interest in business development as well.
If you add on top of that Manchester Digital – the city’s own trade association for new media/tech companies – and the effort that goes into Big Chip Awards, you can see there’s a lot of effort to try and make the sector really viable here.
Has the original goal been achieved? Are we Silicon Valley? No.
But then, I think the emphasis has changed: we’re no longer interested as a community in being Silicon Valley – we’re interested in being Manchester, and being the very best Manchester we can be. There’s a lot of grass roots tech work going on here, with the largest University campus in Europe, with a massive R&D spend, and it’s affordable and enjoyable living here, that’s where the vive is headed…
What is the scene like with regards people getting together to talk geek?
PR: Strong. There are to my knowledge at least half a dozen regular geek get-togethers every month covering everything from high-level topics of interest to business managers in the sector down to piss-ups with Unix sys admins.
Do you have a vibrant User Group infrastructure?
PR: There is a regular meeting of organisers of the various user groups (known as the North West Digital Communities, or NWDC for short) where we sit down face-to-face for a couple of hours, talk about what we’re all doing and what we can do next.
I get the feeling we’re all determined to make people talk to each other and make interesting things happen. We’re supported in part by MDDA and Manchester Digital which gives us focus and means we get cross-pollination. Suddenly a guy who runs an event for VCs is talking to a Unix sys admin and trying to work out if those two groups can overlap somehow.
It’s early days, but it’s getting interesting…
Is there a cross-pollination between different tech areas, eg do the wireless guys party with the Unix crowd, do the developers and designers hang-out together?
AD: As far as I can tell it’s only been recently (the last 2 years) that techies in Manchester have started to cross-pollinate. At the end of 2005 I formed GeekUp, after I failed to find an exisiting User Group that would connect me with other designers and developers in the area. GeekUp, while predominately web related, isn’t based on a specific technology so all techie’s are encouraged to come along. GeekUp’s take place each month and proven to be successful in Manchester there are now GeekUp’s in Leeds and Liverpool. There is also a number of technology specific User Groups:
* BSD User Group
* Linux User Group
* .Net User Group
* North West Ruby User Group
In February of this year North West Digital Communities formed. NWDC brings together the leaders of local communities to improve the local digital community by sharing resources and pooling ideas.
PR: Sometimes. The web guys tend to stay away from the Unix/Windows sys admin guys, but there is an overlap where developers mix between the two. VC events tend to draw SME developers looking to network and so it seems the Venn diagram puts development across the major groups.
I might be biased though: I’m a SME developer and I go to nearly all the events I can, and the only people who tend to show up to all of them with me are developers. We’re like some kind of weird glue. :-)
What events are specific to your part of the world? And what is the benefit of live events to the attendee? How does it affect their output?
PR: GeekUp is exclusively Northern at the moment. To my knowledge me and Sam Smith run the only regular BSD User Group in the UK here. The OpenCoffee stuff was started down in London, but Manoj Ranaweera is doing a lot to promote them across the region.
There are plans for a much bigger event next year, but we’re keeping that close to our chests for the time being. It should be quite interesting though if we pull it off. That will be very Mancunian in the modern/cosmopolitan/forward-looking sense of the word “Mancunian”.
In terms of benefit, part of it is just knowing that there are other geeks around here. Knowing you don’t need to move to London or CA to make something happen is something I think helps people. And then there is the conversation, the beer, oh the beer… :-)
What are employment opportunities like for a techie in Manchester?
PR: Variable. If you’ve got the skills, the work is here. A lot of kids end up going elsewhere in the UK but work hard to get back here if this is where they studied – it might have a reputation for rain, but there’s something addictive about living in and around Manchester.
If you want to work for a small company (and take everything that involves) there is plenty of work around for somebody who knows their stuff. Bullshitters and coasters don’t tend to last long in my experience: you either have the kung fu, or you need to go to London where you won’t get noticed in the swarm of what I can only kindly refer to as “mediocrity”. :-)
The BBC move is expected to produce a lot of jobs, but I’m cynical of how many will be recruited from local population…
In geek-related areas like biotech and engineering there’s quite a lot of opportunity around here as well. Astra Zeneca invest in the region, and the University is trying to bring in a dozen or so Nobel laureates.
Where things really fly around here though is if you’ve got the balls to set up on your own…
Are there big companies that dominate?
PR: Not really. It’s a SME city in many ways. That’s expected to change to some extent with the BBC move to Salford, but big companies tend to move out to the suburbs or down South: a large web tech company in
Manchester would be one having more than a dozen employees.
At the last count I saw in later 2005, there were reckoned to be around 4000 such companies in the North West, but that had a pretty broad definition of “digital” attached to it.
Do they innovate?
PR: Not as much as they should. I’m trying to do something about that. :-)
Most innovative company I think that geeks should know about, based a 10-minute train journey away from the city centre:
Starchaser
Yes, the rockets are real and they work. :-)
The Big Chip Awards is the place where a lot of local firms show off their digital media stuff, but I think in the next few years there is going to be a lot more in general tech locally.
As you probably know, innovation tends to come from smaller, risk-taking companies than it does much bigger firms. We’re primed, we’re just not moving as hard as we could yet, IMHO.
AD: GeekUp runs it’s own Job Board. The up and coming Media City UK will no doubt open up massive employment in the area, with the BBC set to move 1500 London based jobs up North.
Is the work on offer interesting or routine?
PR: It tends to be bespoke development work for larger organisations. High street names often bring their digital portfolio to be developed here. What’s just starting to break out now is the start-up with their own technology funded out of VC or other investment channels.
I expect in the next few years Manchester will become more of a mixture of companies that develop for themselves as part of their own business as well as handling bespoke development as part of a media campaign.
Does the work focus on any particular technology or admin/programming skills? Are there Open Source opportunites?
PR: I’d say it’s dominated by web and ISP work (we have the largest concentration of telecoms and data centre work outside of London) but there are some interesting light engineering firms doing interesting stuff with embedded systems. They’re not prominent though, and don’t talk themselves up.
What about the big national/international tech companies – do they have big set-ups that co-ordinate with the grass roots developers? Any developments in the pipeline to look forward to?
PR: Sun Microsystems is working hard to reach us, and they’re just starting to succeed. They’re based down the road, and doing much better at engaging than the regional IBM office (which is mostly sales). Google has a Manchester office but it’s sales only.
If they took notice of us, we’d take notice of them. As it is, we’re quite happy co-existing if they’re not interested in what’s under their noses.
Does the culture support start-ups?
PR: Yes. The Northern Quarter is very bohemian, and the city itself is full of relatively cheap office space. The cost of living is comparatively low for a cosmopolitan city, and VCs are starting to take interest. The Universities run R&D outfits and incubators and because the city itself is so small everybody kind of has a way of knowing everybody else – useful when you need ideas and contacts.
Are the government helpful in this regard?
PR: The Objective 2 funding has ended, but the NWDA still are trying to help in ways they can. The council are keen to help where they can through MDDA, but it’s becoming more “self-hosting” through Manchester Digital and GeekUp and NWDC. We’re trying to break the culture of saying “why isn’t somebody doing something about this?” and actually doing something about it ourselves. It’ll take a while, but if the enthusiasm is there, it’ll work.
Is there a ready supply of venture capitalists eager to invest in the talent of a promising set-up?
PR: I know of one VC firm on the look out for local companies, and there are rumours of more expecting knocks on the door. To my knowledge, it’s only these guys who go to the effort of trying to knock on geeks first: http://www.enterprise-ventures.co.uk/
Are other techies supportive?
PR: Very. Sometimes the mailing lists for local techs can be a bit aggressive, but that’s what happens when you get a bunch of intelligent, opinionated people who know what they’re talking about together. In general though, great bunch of people.
Do the best ideas come from the best techies or do they come from outside the pool of Manchester geeks?
PR: All over the place. The geeks tend to be the waterwheel of the local industry, but they’re not always the people who decided there needs to be a waterwheel in the first place. :-)
Is there a particular business model preferred round your way, eg do the start-ups build to sell, use advertising as a model, give the app away and hope that somehow money will follow, or do people develop purely for fun?
PR: It’s mostly bespoke development, so you pay, we play, you get a working set of code to make you rich.
There’s a mild increase in VC and investment funding, and they’re keen to follow open models where possible. A mixture of build-to-flip and build-to-dominate in my experience.
There is a lot of hobbyist work around here though – lots of passion out there for doing interesting things. “Hacking” is starting to mean something pure around here these days, which is nice given that for years the only geek group that met were 2600.
What part do key bloggers play? Is there a feedback loop that helps everyone keep in touch with what others are doing? What sites do you all read?
AD: GeekUp has an active mailing which keeps everyone up to date with whats going on.
PR: The GeekUp afeeda is closely tracked and has several blogs on it from local geeks doing different things, there are plenty of mailing lists knocking about, etc.
What major conferences go on in your neck of the woods? How do they affect the day-to-day life of a techie? Are they a source of inspiration or something that goes on in the background for a while but doesn’t really touch the grass roots geek?
AD: Up to now the only major grass roots conference I can think of that Manchester has hosted was the first international OpenStreetMap conference. Next year Manchester will be home to a four-day long Futurology: 2008 Manchester UK conference. On a smaller level GeekUp hosted a one day mini conference on Ruby and Rails, and because of its success, Dave Verwer formed the NWRUG.
PR: Have laptop, will travel: it’s not unusual for whole groups of us to suddenly appear in Brighton, Berlin, London or Paris for a conference. Tends to be focused on EU and UK confs, but sometimes people wander off to the US.
I wouldn’t say they’re a major driving force, but make for interesting conversation. Like I say before, there are plans for us to do something here in Manchester in the next year which I hope will be notable, and then we’ll see where we take that as a city.
Antonio Bonanno is a resident of Milan, Italy. He is one half, along with Giorgio Montersino, of Digital Natives, a start-up created to help Italian companies put together a Social Networking business strategy. I met him through Franceso D’Orazio after the first Social Media Lab at IULM, when, after he and Giorgio had driven us around the city, he kindly agreed to write a piece summing up the Milanese and Italian tech scenes.
Hi there! First of all, I’ve got something to confess: this piece is the result of pizza.
Yes, I mean it. I met Craig Smith of O’ReillyGMT at the first of a series of events called Social Media Lab, and we had a great time together eating all sorts of pizza after the conference. Over pizza, it happened. He proposed I should write a piece about Italy and Milan, what’s happening here in the Web 2.0/Live Web revolution, so here I am. My name is Antonio Bonanno and I’m one of the two partners of Digital Natives, a start-up based in Italy. What we do is try to help out companies and folks who want to create social networks.
When we talk of social networks, we actually mean a whole lot of things. Social networks have different technological platforms: they can be web-based, on your mobile, in Second Life, in your company’s intranet. Each of them is about starting a conversation: with your friends, with your collegues, with some appealing stranger with the same musical tastes. My business partner Giorgio Montersino and I have been active on the Internet for about 8 years, during which time we have seen many developments happen, and we started to play a part in the late 90’s when we began building websites. 
Although we’ve been working together for about 6 years now, we only recently started our own company. During these years, we met a lot of interesting people, and with a mix of their experiences and ours, I will try to outline what’s happening on the Italian scene in the Internet field. On a personal note, I would like to thank Craig for this opportunity. So, here we go!
CS – Are the tech scenes in Italy and Milan buzzing? Does Italy have a history of encouraging technology in general and computing specifically? What is the scene like with regards geeks getting together?
There is a lot going on and the community is very active, but I wouldn’t say that Italy has a history of encouraging technology, in general. We’ve had our great times (Olivetti in the 70s), but we recently lost a lot of that competitiveness in the field. I asked Mafe De Baggis, from Daimon, about this at LeWeb3: in her opinion, Italians don’t like to take risks: they prefer to work for an assured gain, rather than risk high capital. Also, in Italy we have an “innovation caste”, that makes it very difficult for people with good ideas to emerge from the underground. Mafe also played a part in the formation of the community of internet innovators in Milan: she told me about the very beginning, when people started to meet at a bar called Movida: at first they were seen as Martians by members of other business communities, but with time the community started to grow and regain the status of earthliness as the Internet itself grew. The variety of members of the community started to vary, too, from techie-only to a wide range of professionals, from journalist to the emerging jobs of the latest years, community managers, project managers and so on. The community is today still quite strong even off-line.
CS – Do you have a vibrant User Group infrastructure? Is there a cross-pollination between people in different areas of technology?
Working to organize a network of freelancers, I personally think intercommunication between different tech areas is vital both for the growth of new ideas and for the correct development of projects. Working without talking to each other doesn’t lead anywhere: that’s why, at 
Digital Natives, we don’t forget that the person as a whole, and not only their specific abilities, are key to every project. Federico Fasce, community designer and blogger currently working with Daimon, has a different idea. He thinks that there are some interesting communities of people willing to help each other, some even with a solid knowledge base, and gives HTML as an example. It has active groups in HTML, PHP and general webapps and the forum managed by Giorgio Taverniti is a great source for SEO and web marketing information. These kinds of communities tend to form around technical topics. Cross-pollination is more of an issue: there is some within specific communities (web app developers talk to web app designers) but not among the communities themselves. One example is the web and the game industry in Italy: so far there haven’t been contacts between the two that lead to a reciprocal cross-pollination of businesses.
CS – What events are specific to your part of the world? And what are the benefits of live events? How do they affect the attendee’s creativity and output?
Especially lately, events have multiplied at all levels: governmental entities started organizing events with the aim of evangelizing people who are not familiar with the technology and the revolution that’s happening; private organizations focused on getting people together to talk business; members of the tech communities self-organized to promote events with more specific targets, with a strong preference for the “barcamp” model.
Nicola Mattina spoke and helped organize a lot of these events. He recently talked business at 
Web2Oltre in Milan: in his opinion, the event was way too expensive, although it was targeted at corporate executives. BarCamps instead have no cost for attendees, but unlike the ones that happen in other parts of Europe (UK) and where they were born (California), they tend to derail too much towards fun and put innovation and real exchange of ideas aside. It is really hard, Nicola says, to hear interesting things at BarCamps in Italy. Both creative and business output are not very well developed, although they are a good place to meet new friends and recruit people to work with on a variety of projects. Nicola suggests that for future events it would be best to adopt the OpenConference model, rather than continue with barcamps.
On the same note, Federico Fasce thinks the most important activities at events is networking. Barcamps have so far been successful in that, because they have become “the place” for networking among bloggers and members of the different parts of the community. What Federico also sees is a great deal of attention coming from people who are not part of the community, trying to understand what is happening. Events like the recent one organized by Top-IX show an effort in this direction.
CS – What major conferences take place in Italy and Milan? How do they affect the day-to-day life of a techie? Are they a source of inspiration or something that goes on in the background for a while but doesn’t really touch the grass roots geek?
From the perspective of Luca Conti, among the most popular bloggers in Italy with his blog Pandemia, most of the events that take place in Italy are nothing but an attempt to involve the big players that have something to do with the web but are not web-based businesses, thus come from the marketing/advertising side, to stimulate new opportunities and evangelize them. For this reason, the average content level of these events is quite low: it is often the case these events are little more than a showcase, where no real innovation takes place. The targets are traditional and new media and corporate managers who represent either big advertising agencies or companies that are interested in adopting web tools. The geek world is seldom participating in these events, because they have high entrance fees, and if they’re free, they’re very often simply showcases of products of these or that company. 
That is why Paolo Valdemarin and Beniamino Pagliaro organized a 2-day conference called “State of the Net” (February 8th-9th, in Udine), that has the aim to change this trend and speak of what innovation is really about: as Paolo told me, the event will be focused on the web and its impact on society from a professional and cultural point of view, with topics ranging from information management to the web’s social impact. They recently launched the event, which is going to have a very interesting line-up that will be revealed day by day on the website of the event. The event is structured into 3-4 keynotes, 6-7 sessions in which a certain numbers of topics will be faced, in panels with a mixture of web professionals, both Italian and foreign. The event aims to gather both professionals and people who are just interested in those aspects of communication which involve the web.
CS – Does the culture support start-ups? Are the government helpful in this regard? Is there a ready supply of venture capitalists eager to invest in the talent of a promising set-up?
Speaking of start-ups in Italy at this time is a hard task. On one side, we see a lot of new stuff coming up, with very interesting products to be delivered to the market soon. On the other, it becomes more and more evident the need for a big change of culture towards enterpreneurs. Most of the investments and money are in the hands of the big players, who are just now starting to think of investing in smaller businesses (with some exceptions, of course).
Alberto D’Ottavi, consultant and blogger, member of 1Generation and NewTwo, a network of entrepreneurs encouraging the intiatives of young entrepreneurs, thinks that despite the vast number of initiatives, mostly linked to incubators in universities, the level of innovation is quite low. However, the VC system in Italy is opening up to small initiatives. Today there are VCs in Italy – the real problem they have is that they don’t find worthwhile initiatives: most of the initiatives that are now coming up are B2B, and they don’t offer enough perspective growth. Italy had a great deal of innovative products in the past: Yoox, VolaGratis, Dada, Tiscali, Buongiorno.it, all Italian intiatives that had success both inside and outside Italy. The same enterpreneurs now are starting to welcome the new wave, which is actually already here.
Luca Mascaro, owner of Sketchin, a company based in the Italian canton of Switzerland and operating a lot in Italy, has observed that a few Italian VCs started investing also in Ticino, to expand their range of action and possibilities. The start-up culture we have here in Italy (and Western Europe in general) is much different from the American or Northern European one: our entrepreneurs tend to risk much more, to reduce start-up costs. Our competitivity is mostly on risk, than on quality; as Eric Reiss noted in his recent keynote at the Italian Information Architecture Summit, in Italy we do more invention than actual innovation.
Mafe De Baggis also notes that most of the initiatives of innovation that governmental entities launch are really hard to access: the deal of documentation that is needed to even access them is so high that creative and brilliant people are discouraged to even try.
Lele Dainesi, one of the most popular bloggers and now working for Cisco Italy CEO Stefano Venturi, has been overviewing some really interesting initiatives Cisco is taking in Italy. From its position, Cisco is trying to replicate in Italy the Silicon Valley culture that made it the company it is. The Cisco facility in Monza is the biggest R&D center in Europe, with top level engineers, where a lot of the products Cisco does in the whole world are sent for testing and developement. In this facility, Cisco recently started a series of events trying to connect Italy and Silicon Valley, to encourage investments and innovation in our country. The most recent one is an event that took place with the presence of the US Ambassador in Italy, Mr. Spogli, and a panel of experts in venture capital and related subjects. As a follow-up, a group of Italian enterpreneurs and innovators recently went to Silicon Valley to present their products and look at opportunities there. Cisco itself is promoting a VC culture: one of the most recent acquisitions, WebEx, was done in that spirit. What Dainesi hopes is that Cisco will manage to replicate the same culture in Italy.
CS – Are other techies supportive? Do the best ideas come from the best techies or do they come from outside the pool of Italian geeks? Is there a particular business model preferred round your way, eg do the start-ups build to sell, use advertising as a model, give the app away and hope that somehow money will follow, or do people develop purely for fun?
Alberto D’Ottavi, as member of 1Generation and NewTwo, has a very good perspective from which to look at these matters: he thinks that we are now learning to be supportive of each other, because the working culture in Italy for many years meant looking for a stable job rather than taking risks as entrepreneurs. What happened recently is that on top of the few big companies, there are now a lot of smaller initiatives, led by a professionals or small companies. In the first New Economy, around 2000, we saw the explosion and the creation of a lot of initiatives in the tech field, even of a considerable size, mostly concentrated in Milan, which is the advertising pole, and Rome, the telecommunications pole. Now the situation has changed, we have a huge number of small professionals, who work and cooperate on a national scale, communicating and exchanging experiences through the Net, and events like BarCamps and such. What we’re still missing is a structured ecosystem. The business model that is most used in Italy is the one of development and integration of new technologies, particularly B2B.
Paolo Valdemarin agrees on the present situation, where there are a lot of protagonists who work in more than one project of development. He also notes that the level of competitiveness is low, because of the lack of structure of the company ecosystem, which makes it really hard to establish which is the most used business model, in his view.
CS – What part do key bloggers play? Is there a feedback loop that helps everyone keep in touch with what others are doing? What sites do you all read?
Since I was in Paris with many of the most prominent bloggers in Italy (according to stats), I took the occasion to ask them what their views about blogging are.
Luca Conti writes on many blogs, the most important being “Pandemia“. He has been a protagonist of the Italian blogosphere for a long time. The part played by bloggers, he thinks, if you compare it to the blogospheres in other European countries, is less prominent, even if the potential is high. The situation is slowly changing, even if the dimension of the blogosphere is small, and we almost all know each other. The fact that we have these connections, however, doesn’t mean the blogosphere is fully cooperative inside. There is a good deal of competition, mostly based on views and visits more than content: in Luca’s perspective, there should be more cooperation to help raising attention to some very important issues we all feel as fundamental (i.e. net neutrality, technology tools diffusion and so on). His five blogs: Stefano Quintarelli, Andrea Beggi, Giovy, Tommaso Tessarolo, Robin Good.
I asked Nicola Mattina what he thinks of the Italian blogosphere. Some of the relations he had as a blogger then became work opportunies (like the one with Sole24Ore, the leading economic daily newspaper in Italy, which has a Thursday special on Technology called “Nova24”). Blogs, in his view, are used to start conversations: speaking through a blog is almost easier than speaking in the “traditional” way. One of the problems of the blogosphere in Italy, he thinks, is the fact that it’s self-referenced: if you write in English, or speak of a project you made in any other language than Italian, only a few will pay attention. I asked Nicola (and all the others, too) to give me 5 addresses of blogs he usually reads; his list is: Zoro; Luca De Biase; Maestrini per Caso; Luca Conti; Gianni Cuperlo.
Lele Dainesi thinks we’re in a critical moment of the Italian blogosphere and of Web 2.0 in general, in Italy. The attention that is now given to bloggers is not based on content, but on what they do and say or who they know, in a sort of cult of presentialism. It’s not healthy to keep the blogosphere closed as it is now: new entries are less and less, and this creates the akward effect of encouraging personalism. Lele thinks we should get back to content, and speak less of charts, self-exposition, or self-promotion.
The future of blogging, in Lele’s view, is in social networks. He sees a lot of creativity coming from sites like FaceBook or Xing: bloggers will form microcommunities of specialists in certain matters. What is emerging now in the blogosphere is the need for identity management: people want to be different personas in different situations: being a “public persona” is radically changing.
Mafe De Baggis, in a very interesting remark, said that the real crisis of blogging happens when bloggers sell their capacity for suggesting and criticism. One of the most important roles that bloggers should keep is the role of the “maven” (a quote from Tipping Point, Gladwell, 2001), people who are genuinely interested in providing you with the information to do the best choice. Five blogs for Mafe: Mamma per sbaglio, Zoro, Kurai, Squonk, Mantellini.
Before my conclusions, I would like to give you a sneak view of a couple of really good initiatives that are originated in Italy or by Italian entepreneurs and developers I got in touch with in Paris, and before in London at the Virtual Worlds Forum.
One is Dixero. It was launched January 15th, and it’s a service that was developed by Sketchin (owned by Luca Mascaro, who told me about it) and Phiware, with a mixture of Italian and Swedish capital. What it does is it gives the possibility to aggregate and vocalize your favourite RSS feeds with your preferred voice, transforming any entry in a podcast. It works with any standard RSS, from your calendar to your e-mail. In Luca’s vision, it will help people “prolong” their 24-hours day, with the possibility of putting things to listen in background while doing other things.
The other one is Myrl. Myrl is a social network for the Metaverse, a cross-world platform aiming to bring together users from multiple virtual worlds. The initiative was started by Francesco D’Orazio, who now lives in London, with Italian funds. The first upcoming release of Myrl will have a specific focus on social virtual worlds, bringing together users from worlds like Second Life and There.com. Among the most interesting aims and uses of Myrl, which are many, there are these two: Myrl helps bridge the gap between virtual worlds and the web and to integrate web2.0 tools in the Metaverse user experience, while gathering users from multiple social virtual worlds in one place to build a cross-world community of users, in order to stimulate cross-world interaction (world-hopping, not world interoperability), to stimulate a cross-fertilization amongst different virtual world user bases and to promote an organic vision of the Metaverse, as opposed to a walled garden style constellation of worlds. We are waiting for a public beta release of Myrl, which should happen in the next days. Stay tuned!
I think that’s all. Speaking to all these people with different views was extremely interesting and I hope it gives a nice portrait of the situation in Italy. For any other quesion or information, you can leave a comment here or contact me by e-mail.
Thanks for your attention!
Gordon Guthrie is CEO/CTO of hypernumbers.com, an early stage semi-stealth start-up.
He has previously had a range of senior technical and business positions including Chief Technical Architect at if.com and more recently as Solutions Architect at BT/City of Edinburgh Council.
I bumped into Gordon at FOWA in London a few months ago, and asked him to write an article about Erlang, which he was extolling at the O’Reilly stand, and he came back with this piece about why Erlang should (and shouldn’t) be used in the workplace.
Erlang – The CEO’s View by Gordon Guthrie
Disclaimer
This is a CEO opinion on Erlang –it is not a technical article but it does require technical knowledge about programming and programming languages. The poet Robert Lovell once said: “there is no money in poetry, and no poetry in money”. Readers seeking The Joy Of Erlang (and joy there is) will go away from this screed unsatisfied.
Introduction
CEO’s don’t have opinions on programming languages. Or they shouldn’t. Programming languages are merely an epiphenomenon of the business requirements – if a programming language can do ‘it’, whatsoever ‘it’ might be, then it’s in.
So why might a CEO, like myself, venture such an opinion? – and such an opinion attached to the new sexy language de jour. Oops. Old hands will immediately know where to pin the blame, a techie ‘playing up’ – for ‘up’ read ‘out of their league’.
Indeed, what opinion could a CEO offer on a programming language. It would seem to me it would have to be one of two stark options either you should use it – or you shouldn’t.
So before I tell you that you should use Erlang, as tell you I will, it is probably worth making the case
against first.
(Still) No Silver Bullet
The traditional response to this sort of article is to pull out a battered copy of Fred Brooks’ article No Silver Bullet – Essence And Accidents Of Software Engineering [1] – and with good reason. No Silver Bullet asserts:
There is no single development, in either technology or management technique, which by itself, promises even one order of magnitude improvement within a decade in productivity, in reliability, in simplicity.
In business speak this translates into a firm injunction to the CEO that there is not make or break commercial decisions on programming language alone for it to be within their core remit. Certainly there are critical decisions about technology that will impact the performance of the company, but the job of the CEO is to get the best technologist in, build the best corporate culture and enable the troops to get on with it with the tools they think are best fit for the job. Decisions on technology are best left to the CTO and their team. In my other job as CTO I have written [2] about Erlang from a technical perspective.
Brooks separates two different aspects of software – the essential and the accidental. The essential aspects are those that pertain to the business problem that the software is addressing, and the accidental aspects are those that are constraints placed on the outcomes by actually using real software on real machines with real limitations.
Brooks’ argument is that step-changes in software productivity can only come by addressing the essential aspects – there aren’t enough accidental constraints left that their being addressed can deliver the goods.
There are two ways to promote Erlang in these circumstances. The first is as one of the group of functional programming languages (a ‘family’ that includes Lisp, Haskell and F# amongst others) which focuses on addressing the central question of state. Functional languages allow the decomposition of programmes into functions with side-effects (ie that depend on state) and those (the majority) that don’t. Side effect-free code is deterministic – the same inputs give the same outputs every time.
This addresses the core essential complexity of software programming. To quote NSB:
From the complexity comes the difficulty of enumerating, much less understanding, all the possible states of the program, and from that comes the unreliability. From complexity of function comes the difficulty of invoking function, which makes programs hard to use. From complexity of structure comes the difficulty of extending programs to new functions without creating side effects. From complexity of structure come the unvisualized states that constitute security trapdoors.
This line of argument focuses on using the differing semantics of different individual programming languages (and families thereof) to partition the possible state space of the programme and thus reduce complexity dramatically. Functional languages are designed around side-effect-free functions that are their essential (in the Brooks sense) nature. Functionality with side-effects (like writing to disk for instance [3]) are accidental in nature. The FP argument goes that if you don’t design a language such that side-effects are considered constraints on your language then you are mixing accidental and essential problems willy-nilly in all code written in that language – you should develop your systems in a functional language. As a techie I agree with this argument, you should have a bias to FP – but as a CEO I don’t think it is important enough to warrant my time.
The second line of argument is a sort of super-set of the first one which focuses on the Erlang concurrency model. Erlang uses ‘processes’ as basic units of concurrency – processes which are analogous to operating system processes. In Erlang even functional components that have side-effects have no shared state (at the conceptual programming language level). This can be made a convincing argument at a technical level.
But Brooks is right. There is no silver bullet and the Erlang programming language isn’t it, which would appear to leave this article somewhat holed below the waterline. Maybe it is time to make the positive case for Erlang.
Erlang – The Long View
NSB lists four Promising Attacks On The Conceptual Essence:
· buy not build
· requirements refining and rapid prototyping
· incremental development, grow, not build, software
· great designers
On this side of the house, Erlang appears to be a clear loser. There is no evidence that Erlang is superior to other languages in terms of rapid prototyping or incremental development. By definition great software designers are not tied to languages except in terms of the availability of experienced ones – and ‘small’ languages like Erlang naturally lose to ‘large’ ones like almost all others on that line anyway.
‘Buy not build’ really should be ‘procure not build’ with the rise of open source and free software. In language choice this is normally expressed in terms of library support – and on these grounds Erlang also loses. There are certainly ‘missing libraries’ in Erlang – particularly with respect to string management – and these are partly a consequence of limited uptake. The cussedness of Erlang and its obdurate refusal to ‘play nicely’ with other languages doesn’t help – no handy piggybacking.
But it is on this unpromising territory that I will stand my ground. The essence of this argument is that with Erlang and you buy something more primitive than software libraries which you would otherwise have to build – and the cost-saving is so great that there is a compelling financial case for using Erlang almost without exception in certain classes of businesses.
In order to understand what ‘this primitive’ is, it is helpful to go back to the beginning of our trade, yes all the way back to the beginning, the “hello Mr Babbage, how’s it going with auld Ada Lovelace, nudge, nudge, wink, wink?” beginning.
In the beginning the computer was a hardware object sufficient to the task. Babbage conceives of a calculating machine in 1822 and Turing builds the first big computer farm during the Second World War to decode enigma traffic.
These early computers are monolithic, physical implementations of the solution to the problem domain – a single physical design shown schematically to the right:
It is worth recapitulating this timescale. Babbage conceptualises, but does not build, the first computer. 120 years later the first proper one is built and the ‘burning platform’ for this implementation is the Nazi hordes at the gate – one might reasonably come to the conclusion that building monolithic computers is not really that cost effective.
The critical point about Babbage computational machines was that a change to the business requirements meant a change to the physical design of the Bombe. The introduction of a 4-rotor machine in 1941 meant that a new version of the Bombe had to be designed and build.
Turing being the genius he was, was able to address the change management problems of computers. His Universal Turing Machine proved that any problem that was calculable could be decomposed into a hardware and a software element. There had previously been ‘software’ for ‘calculation-like’ machines (punch card driven looms and pianos, for instance) but Turing’s insight was that ‘everything’ was calculable, not just ‘somethings’. Turing postulated the ‘first great bifurcation’ – that between hardware and software, (figure to the left):
In a Turing machine multiple pieces of software can run on a single hardware platform and we see the birth of programming. Turing doesn’t stop there though, he also postulates the second great bifurcation – that between programme and data, shown to the right:
Now the hardest thing about the past is trying to ‘unknow’ things you take for granted –but once this was hard stuff. Again it is worth recapitulating the development cost of systems after the second bifurcation. Each ‘application’ was custom built in software – albeit running on a common hardware platform. In the epilogue to the 20th Anniversary Edition of The Mythical Man–Month Fred Brooks talks about the start to his programming career that is both comfortingly familiar and deeply alien:
The first computer I worked on, fresh out of Harvard, was the IBM 7030 stretch computer, Stretch reigned as the world’s fastest computer from 1961 to 1964; nine copies were delivered.
Yes, that is 9. Fancy running an application? Well let’s start by working out how we intend to layout the files on the DASD, clearly you can’t begin coding until you know how you intend to write your IO subsystem, can you? Gradually coding teams built up libraries of software that they used on application after application. The developers slowly split into programmers and sysprogs – the sysprogs taking care of non-functional aspects of the application.
From this gradual build-up of libraries emerges the operating system in the next great bifurcation, (right):
Voila! operating systems. Again the boundary between the two seems immutable and clear, an obvious natural boundary; as clear as maths. The operating system was always out there just waiting to be discovered. Except of course it isn’t. A gentle reading of the Posix standards [4] will soon disabuse you of any notion of clarity. The boundaries of the operating system are as definitive as a statement about the length of the coastline of Norway. Take the entry on vi. To comply with the Posix standard an operating system must provide a version of the vi editor, to wit:
The vi (visual) utility is a screen-oriented text editor. Only the open and visual modes of the editor are described in IEEE Std 1003.1-2001
It is pretty clear that you could exclude one of the modes of vi quite cheerfully without ‘breaking’ the operating system – it is more import that the boundaries of an operating system are drawn clearly than they are drawn at a particular point. The operating system is just another convention; but the reason the convention thrives is because it is bloody useful.
They say that man is a pattern-matching animal, and by now a number of you will be thinking – “he’s about to make the case that Erlang is The Fourth Bifurcation [5]”. 
Nearly right. I will now make the case that Erlang/OTP is the first implementation of an Application System – which is the product of the 4th bifurcation, (right):
So what are the characteristics of an application system? Well lets turn to the maestro, the father of Erlang, Joe Armstrong. His Doctoral Thesis is called Making Reliable Distributed Systems In The Presence Of Software Errors. The salient point of the introduction is [6]:
When we make a fault-tolerant system we need at least two physically separated computers.
An application system runs on more than one computer. Network up a Windows machines, a Linux machine, a Solaris machine and a FreeBSD one. Start an Erlang shell on each machine with a common shared cookie, a bit of the old-ping-pang-pong in the shell and shazam! an operating multi-machine cluster – and not a line of code written – you have started up an Application System. Its unwritten lines of code sit suitably lightly on the balance sheet.
Application systems address the essential complexity of the software environment by providing a set of ‘bought not built’ mechanisms for addressing the –ilities:
· scalability
· reliability
· manageability
· changeability
· securability
· performability
From a CEO’s perspective using commodity software to deal with the –ilities leaves the paid-for technical team concentrate on the –ality, the functionality. Given that under current circumstances only 10% of the cost of software is the development, with the rest being the post-live servicing, commodifying the bulk of the cost makes obvious sense.
This commodification needs to be seen in the context of the three previous bifurcations. Each bifurcation constitutes an episode of commodification. Each commodification causes a change in the underlying cost-structure of the software industry and each change in that cost base makes incumbents vulnerable and opens up opportunities for new companies.
The critical aspect of the Erlang/OTP that makes is a suitable candidate to be an Application System turns out to be exactly the same as the key aspect of Operating Systems:
· concurrency is provided by processes
Processes are independent units of concurrency that don’t share state and which communicate by passing messages. In operating systems (which run on one computer) a process is only addressable local to the machine. In an application system (which runs on many computers) a processes is addressable across many machines. It follows from this naturally that A/S processes can’t be O/S processes.
Historically the unit of concurrency in an operating system is the ‘user’ – a given process should think it is the only user on the computer. An O/S process should have access to full range of functionality of the physical machine. The O/S process doesn’t actually talk to the hardware – it talks to the O/S which handles the hardware on its behalf.
In an A/S the application system itself runs as an O/S process. If an A/S process wants access to the full range of O/S resources is asks the A/S to access them on its behalf.
Given that O/S’s have a user (or an anthropomorphic batch user) as their basic unit of concurrency it comes as no surprise to find that the upper limits on O/S concurrency tend to mirror the upper limit of concurrent sign-ons for timesharing machines – 4,000 to 8,000 concurrent objects.
In order to get around this limitation many applications use an alternative mechanism of concurrency – the dreaded thread. Unlike processes, threads share state and software errors propagate between them. The threaded model in applications has its analogue in operating system design. There was a very popular operating system that didn’t impose memory page writing constraints between applications – they all shared state. MS-Dos was enormously successful, for a while, but it had, a-hem, stability issues.
By comparison with O/S process, Erlang processes are a lot more lightweight with an upper limit of tens or hundreds of thousands of concurrent processes – if they want heavy lifting they just ask the Erlang Virtual Machine. Critically the Erlang Virtual Machine doesn’t rely on the operating system scheduler to time share between its processes. The VM is just another anthropomorphic batch user receiving one time slice at a time from the Operating System. Its own time-slicer/scheduler parcels that clock-tick amongst its own internal A/S processes. And the VM underlying the Application System has to replicate a whole host of constructs familiar from Operating System design like spawning processes and loading code. This can be of great benefit, as when, for instance, the Erlang code loader is used to hot load changes – running processes are invited to swap themselves out for newer versions whilst executing.
It may seem like this diversion into concurrency models shows a little too much interest in the technical side for a CEO, 
but it is critical to understanding how A/S programming and O/S programming should interact. It is worth showing this schematically, (right):
You can’t just yoke generic libraries written in an Operating System [7] language directly into an Application System language, or more precisely you shouldn’t. It is entirely possible to do so. But mixing programming between abstractions is never going to be a sensible idea. There used to be a popular class of operating system that didn’t prevent the application from writing directly to the hardware. The Win3x class of operating system kernel lasted all the way up to Windows 95 and its ‘flexibility’ brought with it the thrice-blessèd BSOD [8]. You could recompile the Linux kernel to allow your apps to write direct to hardware, and you could recompile Erlang to use your single- or multi-threaded
libraries in some other language – but you really don’t want to.
For good reasons there is a clear distinction between user-land Linux developers and kernel developers (or sysprogs as they used to be known). Typically these two sets of coders don’t share libraries. Unless your (typically) C library is specifically intended to be implemented in the VM you really shouldn’t link it in – ie don’t mess with the VM unless you are an asprog – a contraction I fear that will either never catch on, or already means something beastly on the far shores of the internet.
In Erlang the Application System is provided by a series of discrete VM’s running on a number of machines (typically more than one VM per physical machine). These VM’s expose a common cluster and communications semantics to each other. The ‘universal driver’ approach for encapsulating libraries and other non-Erlang add-ons is to wrap the hardware board or Java class or whatever in a set of libraries that expose those common semantics. In Luke Gorrie’s Distel a popular IDE (some techie doo-dah called Extendible Macros something or other) is so enabled: giving you a clustered development environment and multi-tier debugger.
Naturally the emergence of Application Systems will not avoid meeting some resistance. But, again, ‘twas always so. The world once pullulated with shoals of ‘knit your own operating system’ men but they have gone the way of all flesh.
One may well ask why if Erlang/OTP is so good as an Application System it is not better known outside the telecoms industry. It has a long pedigree and a code base of 1.25 million lines of code. The answer I’m afraid is to be found in the capital requirements of telecoms and web start-ups. In capital intensive industries like telecoms the techies don’t become rich. And rich is the only way geeky stuff gets sexy.
That’s enough of this romp around the history of computing. All the experience of the history of computing tells us that at the point of a bifurcation the cost base of the industry is transformed by an order of magnitude or more. The arrival of commodity Google scalability will change the cost base of the industry and when cost bases change, commercial opportunities arrive.
In Conclusion
Do I, as a CEO, recommend that you use Erlang? No, but I recommend that you use an application system. This particular Kool Aid currently only comes in one flavour, Erlang, but I’m sure new ones will be along soon…
[1]
Henceforth NSB http://www.cs.uu.nl/docs/vakken/pm/docs/no_silver_bullet.html.
There’s a good reason it is well worn and that’s because it is simply excellent. On re-reading it for this article one particular paragraph leapt out at me and made me laugh out loud. On trolling around various start-up offices in and furth of London I have observed the inexorable rise of the super large Apple monitor, 4, 5 or 6 times the screen acreage of a normal laptop screen. Back in 1986 Brooks wrote apropos of graphical programming:
…the screens of today are too small, in pixels, to show both the scope and resolution of any software diagram … the hardware technology will have to advance quite significantly before the scope of our scopes is sufficient to the software design task…
[3]
<autoblogger>Jeezo, he’s an idiot, in what way can IO be anything other
than essential to a computer language – muppet.</autoblogger>
<wiseoldhand> In the olden days we had ‘core’ as in ‘core dump’ and it wasn’t volatile – when you switched the machine off and then on again it came back with the memory state restored. This old volatile memory malarkey is a new thing, and it won’t be around forever. Back in the day the Cray-XMP at London University was the only machine in the UK with 1Mb of Ram; so Ram is much, much more available. Mebbies soon we will have so much non-volatile core that you’ll no need to ship stuff in and out… Io can go back to being a lovely wee moon of Jupiter again.</wiseoldhand>
[5] Sounds like a fillum to me…
[6] Chapter 3 Section 3.1
[7] an
O/S language may be defined as one in which you know at write time which physical machine the concurrent units of the code will execute on at run time. Think C, C++, Java, Ruby, PHP, Perl, Python, Lisp, Fortran etc, etc for each of which the code can only run on ‘this’ machine. You might call a library on ‘this’ machine that will talk to ‘that’ machine.
[8] In the olden days when I wrote the first undergraduate thesis at Bristol University to be ‘word processed’ the word processor I used was a printer. You wrote the text with embedded printer commands in and the ‘programme’ wrote natively to the peripheral hardware.
Kris Buytaert has been a prolific contributor to GMT for the last six months, bringing news of the goings-on from the heart of Europe, beside writing for his own blog, Everything is a Freaking DNS Problem. He is a long time Linux and Open Source Consultant doing Linux and Open Source projects in Belgium, Europe and the rest of the universe. He is currently working for Inuits, and starting up some new projects still in stealth mode.
Kris is the co-Author of Virtualization with Xen, used to be the maintainer of the openMosix HOWTO and is the author of different technical publications. He is a frequent speaker at different international conferences. He spends most of his time working on Linux Clustering (both HA and HPC), Virtualisation and Large Infrastructure Management projects, hence trying to build infrastructures that can survive the 10th floor test.
Kris was the obvious choice to ask for an overview of the tech scene in Belgium. He’s done us proud.
Let’s get started with discussing Belgium, if that even exists as a community… don’t get me wrong … I love Belgium, but the fact is that we are a multilingual country, and it’s mostly these languages defining the communities. Luckily there are some exceptions. But still localisation is one of the borders. Where the Flemish speakers will mostly only work either for a local audience in their own language or for an international audience in English, the French speakers often stick with French as they have a big enough audience with that language.
Is the tech scene in Belgium buzzing? Does Belgium have a history of encouraging technology in general and computing specifically? What is the scene like with regards geeks getting together?
Geeks get together … we have our GeekDinners, monthly events where people from different areas meet, have some fine food and chat about technology. The GeekDinners initiated by Serge van Ginderachter tend to attract an interesting mix of entrepreneurs and kernel hackers. The GeekDinners travel around – we started in Gent, went to Brussels, Antwerp, Leuven and now back to Gent. The GeekGirlDinners are more hyped (even made it to local TV and national radio) but given the fact that I’m male I can’t really comment on them as I’m not one of those wannabees that try to get a date with one of the geek girls. We’re up to the 5th Barcamp, we had 4 editions in Brussels organised by Peter Forret and now Thomas Bouve is organising an edition in Gent. Last but not least we have Robin Wauters organizing frequent OpenCoffeeClub, down in the center of Brussels.
Do you have a vibrant User Group infrastructure? Is there a cross-pollination between people in different areas of technology?
User Groups tend to be really local and small, usually due to the language issues, they tend to start … not gain a critical audience … then fall apart.
We had different LUG’s over the years, but even the bigger ones are now struggling to get an audience.
Our MySQL group is really small, the PHP group is just starting, the Ruby Group seems to be growing steadily…
What events are specific to your part of the world? And what are the benefits of live events? How do they affect the attendee’s creativity and output? What major conferences take place in Belgium? How do they affect the day-to-day life of a techie? Are they a source of inspiration or something that goes on in the background for a while but doesn’t really touch the grass roots geek?
Fosdem definitely … who doesn’t know Fosdem? Already in it’s 8th edition this year. Fosdem attracts geeks and FOSS developers from all over the planet. With figures up to 5000 visitors (no real figures as there is no registration required), Fosdem is starting to become the victim of its own success. Different development rooms are too small, the BeerEvent location is too small and there definitely isn’t enough time to meet everybody who is around.
And of course Javapolis, the Mecca for the Java Developer, one of the bigger events in Europe. I went to the early editions but I haven’t visited recently however.
Lots of new groups/organisations are started at Fosdem, most notably this year Fosdem was the founding point of the Postgres Europe organisation. Last year an Open Source Government workgroup was started.
Does the culture support start-ups? Are the government helpful in this regard? Is there a ready supply of venture capitalists eager to invest in the talent of a promising set-up?
I never really did much research on this topic so I’m pretty sure I’ll be missing some stuff here.
The government support depends on which side of the language border you are – things differ a lot.
On the Flemish side, we have the IWT and the IBBT. I’ve seen a lot of people trying to get funding from the IWT and few succeeding. The IBBT is doing a good job at getting people together.
On the French side, we have Agence Wallonne des Télécommunications (AWT): their people are also pretty active talking to new businesses, helping them getting started.
The fact is that we don’t have a good track record in trying, compared to other countries where you can become a serial re-starter: we have more of a try once, win or disappear mentality. Also we have what we call the “under the church tower” problem. Local businesses often think local, say local and don’t even dare to go away from their local church. Luckily also that mentality is changing.
Are other techies supportive? Do the best ideas come from the best techies or do they come from outside the pool of Belgian geeks? Is there a particular business model preferred round your way, eg do the start-ups build to sell, use advertising as a model, give the app away and hope that somehow money will follow, or do people develop purely for fun?
Ideas come from everywhere, sometimes it’s dreamers that have absolutely no clue on how to get started, sometimes it’s techies who launch before the market is ready. I don’t think there is a rule that defines our way of working. Lots of start-ups combine consultancy tracks to get the money rolling with internal development of the product they really want to build, the advantage is that they can invest their own money and still have bread on the table, the disadvantage is their time to market.
What part do key bloggers play? Is there a feedback loop that helps everyone keep in touch with what others are doing? What sites do you all read?
I’m not sure if there is a site we all read … the tech communities have planet.geekdinner.be and planet.grep.be. There’s gentblogt.be but there is much more going on. We have a couple of bloggers that are really talking about the bloggerscene (blogologie.be), others are reporting local startups and business ideas (web.2point0.be)
I’m personally blogging a lot but my audience is not the local start-up scene, I’m sure there are couple of them reading my blog since I’m advising them technically but the first Barcamp we had was more of a blogger meeting than a real techno start-up event where people came to meet and discuss fresh ideas. Lots of start-up folks are blogging, other are more focussing on their code. It’s a mixture. Robin Wauters is organizing Plugg which I’m really looking forward to.
Lots of interesting stuff is going to happen in the next couple of months, I’ve been talking to a lot of people with good ideas … and some of them are planning to launch soon I’ll keep you folks posted :)
Develer Srl was founded in 2001 by the young and dynamic Simone Zinanni and was joined later on by Giovanni Bajo. Develer’s offices are situated in the heart of Tuscany in a little town north-west of the magnificent city of Florence. Simone started developing at the ripe old age of 12 and became a C++ developer/programmer – he is now the CEO of Develer. Giovanni, a Python enthusiast changed all that and the company is now using mainly Python and C++.
Develer provides a consulting service with full development and support mainly using Open Source technologies but able to accommodate their clients wishes with proprietary software. The success of Develer is made by the ability to respond to customer demands rapidly and effectively. Growing at a steady pace, they now employ 20 people and are always looking for the best programmers Italy can offer them.
A few years ago Simone and Giovanni with friends started l’associazione Python Italia – a national association dedicated to the broadcast, teaching and support of Python – all done on a voluntary basis. From there was born Pycon Italia, now in its fourth year with Develer as the main sponsor and organiser but allegedly helped by the national association.
What people say about Pycon –
“The conference locations were great, the organizers went above and beyond to make things run smoothly, and the food truly transcended the conference food that I’m used to.” – Brian Fitzpatrick
“PyCon Italia is a great way to meet the top coders in Italy, talk with major vendors, and either recruite or be recruited.” – Raymond Hettinger
“PyCon Italia is the conference closest and dearest to my her out of the many I regularly attend.” – Alex Martelli
“I loved PyCon Italia! It has everything: a beautiful city, a great organization, and some of the brightest coders in Italy.” – Guido van Rossum
Develer has also created its own conference – Better Software for an audience of Project Managers and Programmers. In its second year, Better Software has over doubled the number of attendees and from the feedback is going from strength to strength.
Next year, Florence will not see Pycon in May but Pycon Italia will be hosting EuroPycon. I trust they will produce one of their best show for contents, organization, fun and tourism but more on that nearer the time.
It is cold, it is raining, I am in Roma.
It is JavaDay IV at the Engineering Faculty of the University of RomaTre on January 30th. The bad weather did not stop people turning up – over 1600 came to listen to 30 experts in their field. The talks were divided into 6 tracks – What’s Hot; Object-Oriented and Beyond; Managing & Turning; Mobile & Content Management; Web 2.0; and Scalability & Concurrency.
JavaDay is organized by JUG Roma, Java Italian Association and Java Italian Portal. The University lends their premises for the day. Over 20 sponsors – including big companies such as Red Hat, Sun, IBM, Cap Gemini etc, came to display their wares and to speak with the attendees, who were encouraged to leave their CVs with whichever company they would like to work for. This way everybody is happy – people have a chance to talk to the company’s HRDs and the companies get to meet a selection of the best guys available. 
I understand that students are asked to send their theses, which will be read and judged. The authors of the 6 best theses will go home with €1000 to help develop their projects.
Year after year, JavaDay is growing, which shows the importance of Java in the business world. I believe this year the number of attendees grew by over 12%. Unfortunately, I was too busy at our booth to mix with the attendees or talk to the sponsors but I was invited to a great lunch where I talked to several people. Once again, I am overwhelmed by the Italian hospitality and savoir faire. I hope to be back next year at JavaDay V, even if the weather is not agreeable.
Ignite London 2 was a roaring success. From Cory Doctorow’s polemic about the BBC introducing DRM to High Definition at the start to Tom Scott’s SciFi story Mob at the end, there wasn’t a dull moment the entire evening.
The audience - aware, ready to learn - I hope so, these are my O'Reilly colleagues - Helen, Caitlin, Sharon
Arthur, the Sound Guy
With my O’Reilly head on, I particularly enjoyed seeing a couple of the presentations. Andy ‘Bob’ Brockhurst did a fine talk on the Maker/Hacker Revolution. Bob will be Yahoo!’s emissary to Maker Faire (as I type, a week a way, Saturday 13th and Sunday 14th March 2010). And NK Guy did a fine talk on flash photography – NK (or Neil as he’s also known) is about to become a Rocky Nook author, Rocky Nook being a client publisher of O’Reilly’s. His book, Mastering Canon EOS Flash Photography, is out this month.
The view from the sound booth
My co-organiser, Richard Johnson, has worked like a madman to edit the first of the videos for public consumption in quick time. Here’s the first one, the first person on the bill and one of our two special guests for the night, Cory Doctorow:
How the BBC wants to break your television by Cory Doctorow
from hurryonhome on Vimeo.
Paul ‘Vagueware’ Robinson writes:
There is a real buzz all over Europe at the moment about “Open Data” initiatives where repositories of information are made available for exploration by the general public. Of course, it’s not sufficient to just make the data available – there is a growing need for new tools to explore datasets, perhaps similar to Google’s recently announced Public Data Explorer.
Whilst projects like data.gov.uk provide a starting point for national data sets, momentum is starting to build at local and regional levels for more detailed, higher resolution information held by local public bodies. By comparison, most European cities are some way behind the impressive local data catalogue published by the city of Vancouver.
Many public bodies might feel there is not a large enough community to make use of the data, or that it is of insufficient value to make available to the public. In Manchester, a new group hopes to change that.
Open Data Manchester is a new group that has as its stated aims to:
“[…] help developers and designers become more familiar with tools, datasets and other projects around the World; identify datasets of use to local data users and of interest to us generally; [and] to provide a catalyst for local authorities in Manchester and the North so that they might make data available for exploration, knowing there are established groups who wish to use it
The group plans to have regular meetings (probably at MadLab, the new home for most digital communities in Manchester), but there is also a Google Group for those further afield with something to contribute, or just keen to hear about what’s going on.
I went along to see Christian Crumlish speak at Covent Garden Wallacespace. Christian is the curator of the Yahoo! Design Pattern Library and co-author of Designing Social Interfaces, and he did a fine hour-long talk (plus questions) about how design patterns can be used for building social websites.
My notes, incomplete as they are, read:
Patterns originated in the Architecture Community. Once you discover rules and patterns, they can be used in other situations. Architects missed the benefit of patterns, but Computer Scientists got it. Patterns and wikis have a similar base, as they were originated within the same crowd.
Yahoo!’s Pattern Library, of which Christian is the 3rd Curator, is the sibling to Yahoo! User Interface.
Delicious and Flickr popularised tagging, both were bought by Yahoo! But both have different implementations of tagging.
People take the code and design on top of it. People didn’t think they needed the telephone – it took people a while to figure out how it could be used for business. Business is done through it. Blogs are like that – they don’t necessarily bring money in, but business is done through them. The network effect of having millions of phones connected escalates their value, just as it does with having many people connected via a website.
People used to phone a place – ‘is such-a-person there?’ – where now we phone a person, or even a pocket.
Christian offered up with 5 Principles – not set in stone, not without exceptions, but strong guidelines. These principles, were tested at Barcamps etc, and eventually became Designing Social Interfaces:
Pave the Cowpaths
Look at the behaviour that is already happening and facilitate that.
Talk Like A Person
Don’t do corporate speak etc, your web copy should be conversational to set the tone you want your users to use. These include:
- Conversational voice
- Self-deprecating Error Messages
- Ask Questions
- Your vs My (‘Your’ gives context – someone else is saying ‘Your’, it is inclusive – with ‘My’ you’re out on your own)
- No joking around (sarcasm, in-jokes often don’t translate online – your audience might not get the humour)
Play Well With Others
- Embrace Open Standards
- Allow data outside the bounds of your application
- Accept external data within the sphere of your app
- Support two-way interoperability
Learn from Games
There are things about games which apps can learn. For example, reward participation. Got to give up some control to your users. For example, the designers who build Never Ending Game went on to build Flickr – a game-like culture was transferred.
Respect the Ethical Dimension
Every social website has an ethic dimension. Don’t spam, don’t pester.
Other gems Christian doled out:
- Give people a way to be identified
- Social Objects – Things people have a common interest in, something to rally around. They keep people involved, they allow the site to grow organically
- Give people something to do – eg upload videos – let them be involved
- Let the community elevate people and content they value – 3rd party moderation doesn’t scale
- Enable a bridge to real life events
- Anti-patterns! – Things which seem like a good idea at the time but later on shows itself to be bad eg copying a successful website but not understanding what makes it successful
- Don’t break email – eg NoReply is a bad idea
Overall, this was a great evening (Yahoo! always put on great events) with a capacity audience stacked right to the back of the cafe. I have a few more photos here.