Filippo Diotalevi

(Review) Code, by Charles Petzold

Thu, 03 Jan 2013 00:00:00 +0000

I don't remember where I got the suggestion to read this book; it must be something related to my quest to find books that talks about the foundations of technology, rather than about the latest and greatest tool. Anyway, somewhere I found a link to this nice little book, and I'm glad I did.

You can think about Code: The Hidden Language of Computer Hardware and Software as an Introduction to Electronic university course, given by a professor that actually cares about his students. Weird, I know.

The book is actually much more than that. Its trajectory goes through two thousands years of discovers, starting from the concept of numbers to Object Oriented Programming; explaining, in the process, binary numbers, relays, logic gates, transistors, integrated circuits, Intel 8080 and Motorola 6800, assembly language and operating systems. Everything is plain and simple English, with a surprising attention to details.

The book is so simple, and convincing, you'll actually consider (for a short while) building your own adding machine, or going back to Assembly to code your next toy project. It is seriously that good!

I see only a couple of shortcomings in this book: the first, is that it feels, and is, a book written in the twentieth century. So many things happened in the past twelve years which are not reflected in here. Second, it's mainly a hardware book; it talks, in the last few chapters, about programming languages and operating system, but it's only a very brief overview.

So now, after enjoying Charles Petzold's Code, the question is: is there an equally good book about the story of operating systems or programming languages? For any suggestion, please give me a shout on Twitter.

Strata London Notes

Tue, 02 Oct 2012 00:00:00 +0000

A few useful (mostly for myself) notes after Strata London 2012.
As other people there noticed, quite a disappointing conference. Not just for the appalling location (note for organizers: hosting conferences in hotels is almost always a bad idea), but because all the topics of the conference felt, at least, three years old. Some of the things missing:

Graph databases, not a single mention
Real time analytics and machine learning: too little. Three related talks, no mentions to Storm and alternatives
No talks on R and related technologies?
Very little on data visualization (mostly driven by The Guardian and Tableau)

On the other side, too much of...

Hadoop everywhere! I do understand many of the sponsors were Hadoop "resellers", but so many talks?
Non technical talks.

Rant
I don't really know who started this trend of having non-technical, "cultural", "i will surprise you with my precious insights" kind of talks at technical conferences. I admit that at times it may be fun, even interesting, like when Bob Martin goes into details explaining how black holes work. However, isn't it boring when you see that happening at every single conference?
Do we really need three talks talking about Turing, Babbage and co, and what Big Data was like 50 years ago?

After the moanings, a list of the few discoveries that are worth a second look

Open Data and Open Government in UK

UK seems to be quite serious about opening up their data. Some examples:
Legislation.gov.uk offers the UK legislation, accessible via API and open to 3rd party applications
The Open Data Institute is the place to start to learn about the government Open Data policy

Pro Bono IT and Big Data

I have been looking for years (since I learnt about the inspiring JugAvis project) to find opportunities to connect charities and NGO with developers wanting to work pro-bono. Now DataKind proposes to do that. "Using data in the service of humanity" is their tagline. It's a US no-profit, but hopefully they'll expand their activities to UK soon.

I learnt that Google is producing tools for data analysis quicker than I can learn about

In no particular order

Google BigQuery

There's quite a lot of research going on around online (streaming) algorithms

... and it's not only limited to "summarizing" algorithms, but also to online machine learning.
Stream Lib implements some of them
Ted Dunning shares the details of a fast, large scale, k-means analysis

London Population Over The Years

Mon, 06 Aug 2012 00:00:00 +0000

After a few weeks of intense work sharpening my R skills, I figured it was about time to start pushing out some code. One of my favourite blog is the awesome Spatial Analysis, so why not starting with some maps?

The maps displayed in this blog post are the combination of two distinct datasets:

The UK "Boundary Line" dataset, downloadable (after a tricky and tedious procedure) from this website
The Historic Census dataset downloadable from the London Datastore website

The first dataset (Boundary Line) contains the polygons (shapelines) defining British counties, districts, cities, ... I used ggplot2 to display the shapelines (mostly because I wanted to learnt it better).
The second dataset contains the population in each London borough from 1801 to 2001. Merging the two dataset is immediate. The full code is on Github, and 36 lines of codes are all you need to load, merge and display the data.

The different images show, as expected, how the population has progressively increased in number, and moved away from the central boroughs (Westminster, City of London in particular).

After the second world war, there's the most noticeable increase in population of the external boroughs; Barnet, Croydon, Ealing are rapidly moving from average-low inhabited boroughs to some of the most populated; all the other external boroughs show similar increases.
The last 30 years of the 20th century don't show a significant change in how the population is distributed.

From a visualization point of view, working with R allows to create maps easily, and with a few lines of code. There are definitely a few things I'd like to improve, like the legend (which sometimes shows numbers in scientific notation, sometimes not, without an apparent reason), and the placement of borough names which sometimes make them difficult to read.
Unfortunately the general lack of documentation on R and its packages makes this fine-tuning procedure very hard and time consuming, but it has been surprisingly easy and quick to obtain some decent maps.

HttpCheck in node.js

Fri, 09 Sep 2011 00:00:00 +0000

One of the most useful tools when you develop mobile portals is a simple web page to check the headers your browser is sending to the server. Which User-Agent? Accept-Language? Accept-Encoding? Accept-Charset?

A good opportunity to write a tiny node.js application (with just a bit of Underscore magic).

var http = require('http'), 
    _ = require('./underscore'),
    port = process.env.PORT || 3000;

var FullResponse = function(req) {

  var blanks = "                                   ";
  var fields = {"url": req.url, "method": req.method, "httpVersion": req.httpVersion, 
              "headers": req.headers, "data": ""};  
    
  var format = function(key, value, indent) {    
    if (_.isString(value))
      var valuePart = value+"\n";
    else
      var valuePart = "\n" + _.reduce(_.keys(value), function(sum, it) {  
        return sum + format(it, value[it], indent+1);  }, 
        "");
          
    return blanks.substring(0, indent * 2) + key+": " + valuePart;
  }        
  
  this.addData = function(data) {
    fields["data"] += data.toString("utf8");
  }
  
  this.printOut = function() {        
    return format("response", fields, 0);
  }
  
}

http.createServer(function (req, res) {
  res.writeHead(200, {'Content-Type': 'text/plain'});
  var response = new FullResponse(req);  
  req.on("data", function(it) {
    response.addData(it);
  });  
  req.on("end", function() {
    res.end(response.printOut());  
  });  
}).listen(port);
console.log('Server running at port '+port);

Full source code on Github, the application is running at http://httpcheck.herokuapp.com.

Trying node.js

Wed, 04 May 2011 00:00:00 +0000

After a few months of Rails development, I start to really enjoy working with dynamic languages. Coming from a Java background, it takes some time to get used to work without a proper IDE with code completion and good refactoring tools; however, after a while the development experience becomes more natural, especially if you pair a dynamic language with a schema free database like MongoDB.

I was really curious to try some other (dynamic) way to develop web applications, so I decided to build a website to sell my old furniture, and I picked node.js because... well, a lot of people is talking about it. I understand that's not a sensible reason to choose a technology but hey, I just wanted to sell a few things. Have a look at the result, if you are interested (I still have a good wardrobe for sale, anyway).

At the end of the development, these are my notes/impressions.

Overall development experience
Developing the website has taken me about 12 hours. I estimate I spent roughly 30% of the time to google for information, and 20% to integrate the design. So development is reasonably quick.
Programming a web application in a event-driven fashion (the node.js way) feels a bit weird, but it's not a problem for simple webapps. I can imagine that if you need to write some complicate business logic that might be quite challenging though.

Node.js and Express
Node.js is a low-level framework, so if you want to develop a web application you better use Express. Express is a minimal web framework, similar to Sinatra. It's well documented in the website, and very quick to get started.

Jade
Since the Express documentation states that Jade is Haml successor, I figured that was the right choice. Wrong. For two reasons.
First, jade is basically undocumented. The Github page gives you some hints, but really not enough to write even a simple web application. Pro tip: the source code contains a examples/ folder.
Second reason: jade is very similar to Haml. Except that it is different. It feels similar, but some small details are different, so that (if you happen to work also with Haml like I'm doing) you keep making annoying errors.

MongoDB integration
I used node-mongodb-native and mongoskin to integrate the web application with MongoDB. Documentation here is still very minimal, and you have to figure out things via trial and error. But anyway, every MongoDB driver works basically in the same way.

Deployment on Duostack
One of the reason I chose the Node.js+MongoDB stack is to be able to deploy the application, for free, on Duostack. Duostack is the Heroku for Node.js, with some nice additions like the availability of MySQL, MongoDB and Redis to store your data. And a very reasonable free plan that allows to install your application with a 128MB database.
Duostack works exactly like Heroku; you push your code in production via git, and everything works. The application feels actually a bit slower than some of my applications deployed on Heroku, but I haven't really done any benchmark. It's free! so I can't complain.

In summary
Node.js and Express are enjoyable technologies to use, but really in their infancy. They're also in constant development, and incompatibilities are likely to be introduced in new versions. Their documentation is quite good, but the whole set of frameworks/technologies that you need to use to implement a full web application (templating, authentication, authorisation) is not. So expect to spend a lot of time to google for solutions, and to ask/answer questions in mailing lists and stackoverflow.
Also, I'm not convinced the event-driven approach can cope well with websites with complex business logic; it would help to have some books or at least a best practice collection to refer in this case, but I couldn't find anything.
On the other side, if you need to implement a network server, node.js is certainly worth having a good look.

Week 5

Sun, 06 Feb 2011 00:00:00 +0000

My bi-weekly update... a few interesting things happened.

Unexpectedly, I started working on my first iPhone/iPad application. That's something I didn't foresee because my objective-C skills are virtually not-existent, but my friends at The Dusseldorf Experiment are helping me out. My goal is to create a mobile application for Trunk.ly, the new bookmarking service I use. The iPhone version is ready (albeit minimal), I hope to be able to start the iPad one this week. And no, I don't want to become an objective-C developer; I already have my problems switching daily between Ruby and Java.

Last weekend I attended the startup weekend in Bruxelles; like my previous experience in Austin I enjoyed the event so much, but this time our team also went on to win the price for the geekiest project. That's a screencast of the demo if you are interested.

Unfortunately I had to miss Fosdem this year; I was planning to go on Sunday, but a bad cold convinced me to stay home.

I read Start Small, Stay Small by Rob Walling. I have mixed feeling about this book: while it offers a lot of useful suggestions, if you (like me) have already read all the possible books about SEOs, bootstrapping startups, lean startups and outsourcing to Virtual Assistants, this book won't tell you anything more. So, it's a useful summary, but nothing really new.

Usual contracting work going on, nothing new on that side. The three contracts I have are more than enough to keep me very busy right now, so I don't see any news in this area at least until April.

Week 3

Mon, 24 Jan 2011 00:00:00 +0000

Two weeks worth of updates today.

Three consulting gigs ongoing: Ruby on Rails work, an architecture/performance review and finally some work on a OSGi SaaS platform. Definitely busy times!

I started to collect some Ruby and Rails code snippets (and some wiki pages of explanation) at https://github.com/fdiotalevi/rails-dump. Nothing polished, just some code to copy in my next project, or step-by-step tutorials to remember certain solutions. Some things I posted there:

an improved standard layout for Rails projects (the default one is ugly)
a reusable module to create wizard controllers
how to setup Devise using mongoid as ORM (and MongoDB, of course)

For my traditional Sunday project, I decided to write a minimal blog engine in Ruby. I'm definitely sick of the terrible service offered by my current hosting provider (Servage), and have always played with the idea of writing something simple I can customise to my needs.

After getting the final inspiration looking at Toto and Marley, I wrote a minimal Sinatra-based (and Git backed) blog engine in 4 hours last Sunday. Deployed on Heroku, it's a 0-cost solution (because you can deploy it using the free plan), runs on a fast server and it's a completely customisable blog. I need some more time to add a couple of other features (in particular, a RSS feed), but it's already deployed here.

Finally, I started to define my idea for the Bruxelles Startup Weekend. It's one of my many ideas to improve the way people travel... more details in a few days.

Innovation and Large Companies

Mon, 17 Jan 2011 00:00:00 +0000

I had a few conversations lately about the difficulties of producing innovation in large companies (telco operators, in the specific, but that equally applies to many other businesses).

That reminded an excellent article Robert Scoble wrote a couple of months ago, Why Google can't build Instagram; that is, the main reasons why big companies can't innovate are

they can't afford to release early (minimal) versions of the product and iterate; people expects finished, quality products from big established companies

their products need to scale from day 1; while a startup normally have a few hundred users in its early days, a new product launched (even in beta or labs) by Google immediately has millions of users. Therefore more people or entire teams need to be involved to make sure the new service will properly scale

in big companies there's no complete freedom, even when it comes to technical choices. You are supposed to use one of the approved technologies, for instance; and you aren't suppose to use or integrate with competing services

So what's the solution? Are a vast majority of big companies doomed to fail within ten years for lack of innovation?

Robert proposes three ideas: innovate by acquisition, work on open source projects and keep teams small. I agree with the first proposal, but I think there is actually a better method large corporates can use to innovate.

YCombinator, TechStars, CapitalFactory and other similar programs have demonstrated in the past five years that there's a systematic and repeatable model to create innovation. The rules are simple (but correct execution is, as usual, the most important thing)

select the best aspiring entrepreneurs worldwide

constant mentorship

provide minimal funding to entrepreneurs to create their startups (not to work on your company and your ideas)

help entrepreneurs to manage company financials, funding and growth (but without interfering too much with their business)

What's the budget to create such a program? Less than the money spent in any mid-sized project in any big corporate. You can probably go a long way with 1 million $ (or euros) per year, and roll out a great program for less than two.

Noteworthy links, January 10th 2010

Mon, 10 Jan 2011 00:00:00 +0000

Many interesting news in the first 10 days of the year..

The Pleasure Principle: Not All Products Need To Be Painkillers How the usual "solve users pain" does not apply to all interesting businesses
In Defence of Copycat Businesses What about copying successful services?
Facebook hype will fade Interesting take on Facebook success
Ten business models that rocked 2010
Funding Lessons from a Successful Kickstarter Campaign Funding Lessons from a Successful Kickstarter Campaign
2011 may mark the beginning of a golden era for entrepreneurs "when it was the darkest, we saw the stars"
The 10 Coolest Kickstarter Projects of 2010 Crowdfunding at its best
10 TED Talks for Entrepreneurs

Week 1

Sat, 08 Jan 2011 00:00:00 +0000

Week 1 of 2010 has been particularly eventful. It seems like everybody woke up after the Christmas break and decided to contact me!

First of all, I finished reading Metaprogramming Ruby; more than a book, a great experience! I already suggested it to many people, but if you are interested in Ruby, read it by any mean!

A few people suggested me new events to add to my Startup/Tech European events list: now is more complete than ever! Have a look at it and choose your events for the first quarter of 2011.

I've started by Ruby on Rails development contract; I cannot blog much about that, but I'm having a lot of fun improving my Ruby and Rails skills.

I've also reached an agreement for another consulting contract for the first trimester of 2011; it's going to be more a performance and architecture review job than a coding one, but hopefully it will turn out to be interesting.

Finally, I've had some free hours on Sunday to put together a Trunk.ly Ruby client; it is still incomplete and possibly buggy, but it's already on GitHub. Give it a try. Trunk.ly is still a young service, but very promising