Hack Days

Last week we had our third internal hack day at STC Europe and I am still buzzing from the excitement. For those of you that are not familiar with the concept, hack day “is an event where developers, designers and people with ideas gather to build ‘cool stuff‘”.

I have been organising them at STC Europe since I first joined (after some gentle prodding by Ian Hegerty) and I have found them to be extremely fun and valuable.

Therefore, in this post, I want to share with you some of the learnings & benefits of organising Hack Days and to gently prod you in return to start your own hack days within your organisation or team.

Are you going as fast as you can?

What surprises anyone participating for the first time in a hack day, or attending the demo session, is how much can be created by very small teams – from idea to implementation in 48 hours – while so many traditional software project don’t produce anything over much longer time periods with much more managerial care and oversight. For me, having worked for bigger companies in the last 5 years or so, it is a good reminder of how lethal start ups can be: when you have a highly motivated team that fully buys into an idea,  combined with a strict deadline that forces a team to focus on the essence of the product, not much is impossible.

Although by its very nature, a 48h sprint will be faster than a normal development cycle, and hackers will take many shortcuts, hack days can illustrate shortcomings in your normal systems and processes. For example, are your release management systems causing too much overhead? Are you giving developers enough ownership of feature development to be fully engaged? Is planning becoming a drag instead of a tool? It can also be a call to arms for your leadership team: how can we bring the same energy, agility and excitement to our day-to-day work? Even if you are no longer a start up.

Your team may be better than you know

Remember that quiet guy in the corner that never seemed that interested. He will blow you away with something truly cool. This is not a Hollywood movie. Believe me: Hack Days will allow team members to demonstrate talents, skill sets and great ideas that they themselves may not even known they had. This opens up many new opportunities and improvements for how you structure the team, products or features you go after as a company but also makes for much happier team members.

Joint experiences & adventures make for stronger teams

Surviving release cycles together may create team bonds over time, but hack days will do that for you in much shorter time and will do so cross-team, maybe even cross-office. Hack day is basically one big rush: come up with a cool idea (buzz), find like-minded people (buzz), implement it together – quickly (buzz), demo it for your peers & management team (buzz). All lubricated by pizza, beer and doughnuts. In 48 hours. It’s like a road trip but geekier. You will create bonds, you will make friendships and you will gain respect for your peers.

Great invention ROI

I love this recent tweet from Kent Beck: “once i thought all my ideas were gold, then i realized most of them were crap, then i learned i had to try them to find which were which”. You can try to research if new products and features will work via a variety of means but nothing will demonstrate if an idea will work or not faster than a working prototype. And you will get many of them via hack days, far outweighing the cost of the 2 developer days.

And don’t forget about the ideas. Nothing like a nice mix of backgrounds and experiences to come up with new cool ideas… Much better than a blue sky thinking offsite with similar people just waiting for the open bar in the evening.

It’s fun

Finally, it’s fun. You are likely in the technology business because you love to create. Hack Day is it – close to its purest.

What are you waiting for?

How many things out there are fun, create better teams and have great ROI?

Go for it!

Reading “end of management”

I am really enjoying reading and re-reading The End of Management

It makes the argument that managers should become more like venture capitalists as the structure of a “corporation” loses its relevance.

To be honest, I do not feel I have fully digested all the points made in the article but one line that especially stuck with me was that “most of the greatest management stories have been not triumphs of the corporation, but triumphs over the corporation”.

Something to think about for today

The Sorrow of L&H… and Belgium

L&H was in the news again yesterday as its founders were convicted for fraud.

For those of you that have never heard of L&H, Wikipedia to the rescue: “Lernout & Hauspie Speech Products, or L&H, was a leading Belgium-based speech recognition technology company, founded by Jo Lernout and Pol Hauspie, that went bankrupt in 2001. The company was based in Ieper, Flanders, in what was then called the Flanders Language Valley (mimicking the Silicon Valley).”

I still remember the hype about the company so well. I was studying at the time and I cannot deny that its apparent success played a role in my focusing on computational linguistics and NLP in my licentiate years.

My cunning plan (to quote Blackadder) for future career success however did not work out as planned. Does it ever? Turns out I was not the only Flemish person with this career prospect in mind and other language skills were more in demand around the time I graduated.

I ended up joining Trados (www.trados.com) instead and by the time I did get the call from L&H, I was too invested and too in love with Trados, its startup culture and its prospects to switch. Even though it was still a fairly difficult call and I needed some talking to from the then CEO (who happened to be in Brussels at that time).

Regardless of my not joining them in the end, I was however very, very proud of the company and excited about what this would mean for Belgium. I still remember going into Microsoft Office and checking for the L&H copyright in its About Box. And feeling yes, proud.

And then, the demise, the ugly truth, lots and lots of investors losing their money and lots and lots of talented people losing their jobs. And, in a way, it feels like Belgium lost its place forever in a burgeoning field that could have been theirs due to yes, its language skills — with or without L&H.

Reading the news today, I wonder if the wounds are still too deep to support another NLP startup in Belgium any time soon or any “larger than Belgium” tech startup story overall.

And that is the saddest thing of all.

Build Your Own Search Engine in .NET: Step-by-Step

I still remember very well my dad bringing home a commodore 64 as one of the first families in our neighborhood (making us quite popular for a short period of time). I tried my hand at some BASIC but my rationalisation for spending more time on the games than on BASIC was that I had read that everything was evolving so quickly in this new world and that anything you would code now would probably be available shortly anyway or easier to do due to better computing power, graphics, etc. Why then bother to apply myself now (instead of playing games) as the landscape & tools would have changed & be easier when I enter the workforce… 

Not my smartest moment but at least I have tried to stay in touch with the programming world throughout my career (and inspired due to working with some incredible engineers). And one thing I realised as a kid is true, it is becoming easier & easier for people to create really cool applications in a short period of time. You just need some vision,  a long weekend and some takeaways (and/or an understanding spouse).

One of the amazing new APIs that makes it easier than ever to develop something really cool is Yahoo! BOSS. Basically, BOSS gives you an easy-to-use web service to build your search engine. So in light of Yahoo!’s upcoming hackday in London (which I will be attending as a representative for Y! Search in Europe), I thought I’d give it a go myself.

The result was a site I named Searchdoublr. Basically, Searchdoublr searches both Yahoo! Search and Yahoo! Answers at the same time. Rationale is that Yahoo! Answers really gives some great results for any query where experiential knowledge would be helpful as to your task intent. E.g. try “best Japanese restaurants in london” (really looking for recommendation) or ”heathrow” (you may really be looking for how best to get there, or how much parking costs, etc and hoping that the official site provides that information). 

In any case, let me quickly describe how you can build your own search engine in .NET using the free Visual Web Developer:  

1. call the BOSS URL with the correct parameters
2. load the returned XML into the standard ASP.NET XML control
3. transform the XML to HTML using a standard XML Transformation stylesheet
4. Use a short CSS stylesheet to make the returned HTML look pretty

Or the longer version:

• Get an AppID at http://developer.yahoo.com/search/boss/
• Add a textbox (tbQuery) and a button (btnSearch) to your web form (for the user to enter your query) as well as an XML control (which will be used to show your search results).
•  Create a button_click event by double clicking your button and create an XML document as well as an XML Transformation object to show the search result:  

System.Xml.XmlDocument srp = new System.Xml.XmlDocument();
// this is where you call the boss API where you specify your query and appid 
// and ask for xml to be returned                
srp.Load("http://boss.yahooapis.com/ysearch/web/v1/" + tbQuery.Text
                + "?appid=<yourappid>&format=xml");
// Create a XSL Transformation
System.Xml.Xsl.XslTransform transSRP = new System.Xml.Xsl.XslTransform();
transSRP.Load(Server.MapPath("SRP.xsl"));
Xml1.Document = srp;
Xml1.Transform = transSRP;

• SRP.xsl contains your XML Transformation which is pretty standard. Note that it is important to take care of the namespace. I use the prefix “op” here. Specifying “disable-output-escaping=”yes”" ensures that we show the bolding of the keyword in the result (title & abstract) correctly.

<?xml version="1.0" encoding="iso-8859-1"?>
<xsl:stylesheet version="1.0"
    xmlns:xsl="http://www.w3.org/1999/XSL/Transform"
    xmlns:op="http://www.inktomi.com/">
  <xsl:template match="/">
      <h3>Search Results</h3>
      <table>
        <xsl:for-each select="//op:result">
      <tr>
        <td class="title">
          <a><xsl:attribute name="href"><xsl:value-of select="op:clickurl"/>
            </xsl:attribute>
            <xsl:value-of select="op:title" disable-output-escaping="yes"/></a>
        </td>
      </tr>
      <tr>
        <td class="abstract">
          <xsl:value-of select="op:abstract" disable-output-escaping="yes"/>
        </td>
      </tr>
      <tr>
        <td class="url">
            <xsl:value-of select="op:url" disable-output-escaping="yes"/>
        </td>
      </tr>
        </xsl:for-each>
      </table>
</xsl:template>
</xsl:stylesheet> 

• You may want to specify a CSS stylesheet as well to display the results a bit nicer:

.title
{
   font-size: small;
}
.abstract
{
    font-size: smaller;
}
.url
{
    color:Green;
    font-size:x-small;
    padding-bottom: 15px;
}

And that’s really all you need to do. Let me know how you get on. I will make the full source code available shortly,

Wahanda

As I have stated before, I admire the people with the guts to go out there and start up a business and I recently ran into an ex-colleague who made that jump. Salim Mitha used to run Y! Answers in EU and after leaving the company he co-founded a brand new internet startup called Wahanda. Wahanda is a London based health, beauty and wellness community & marketplace and you should really check it out. Nothing like a good treat to forget the depressing news of late. But as time are tight maybe wise to stick to the offers page…

goalsetr

Just put up my first website. The idea is to bring goalsetting to the masses by making it social so users can share their goals, experiences and resources.

The result is goalsetr. Check it out.

Start Ups

For quite a while now, I have been thinking about starting up something of my own, but as of yet I haven’t found the courage to say goodbye to the security of a steady job. I guess moving around from country to country is enough adventure for now

Well, until that day comes, I am just going to live vicariously through those that have made the step. Here’s a shout out to my old colleagues who have dared to go into no mans land — such as my old boss who started up Metaversum, and to those silently working away in stealth mode for their big day, or joint a startup such as evincii.

For you fellow dreamers, I loved this book Founders at Work: Stories of Startups’ Early Days.