ReliefWeb's Early Days

Craig Duncan
Screenshot of ReliefWeb in 1997 © Internet Archive

ReliefWeb is now 25 years old. In a series of posts, we’ll be looking back over those years from the perspective of some of the people who helped make it what it is today. Craig Duncan, one of the first people to work on the project, remembers how it developed from a list of links – and how it found its name. Inevitably there are some technical details involved!

The ReliefWeb project had a few false starts before a new team started in the spring of 1996 under Sharon Rusu, the coordinator and project lead. I was the fourth hire and at that time, there was not really much of a website, just a canonical list of links online at a numerical IP address.

Dan Zalik, a bright young student from Brown University, was the brainchild for the new content-based design model and we started collecting content to publish in April 1996. At first we only tried to cover five key emergencies, but that expanded quickly.

The website was originally to be called ReliefNet, after Dan’s Master’s project called RwandaNet – but the domain name was already taken (by a band called Relief) so we went with ReliefWeb. Our disappointment was ameliorated when ITU, our technical partner and web host, provided an ‘int’ domain name:

At the time there was no such thing as a content management system for the web, or even an HTML editor, so everything was marked up manually in basic HTML. The menus too were created manually, and each new document had to be added to lists in three different places.

My first role was to input – and add to the three different lists – all the documents we’d collected – about 1,000 in two months. By the official launch date in October 1996, there was already an impressive amount of content on the site. From then on we added about 30 new documents every day. And the daily updates have continued now for 25 years!

One of our first challenges was the fact that Simon Fraser University in Vancouver was publishing Department of Humanitarian Affairs sitreps on the web the day after they came out – so we aimed to publish our own faster than they could. [In 1998, as part of the Secretary-General's programme of reform, DHA was reorganized into the Office for the Coordination of Humanitarian Affairs (OCHA).]

We had a substantial team from the beginning – eight people by May 1996. But unlike websites today, we benefited from low expectations: for the first two years, no one seemed to really have a clue what we were doing, but let us get on with it anyway.

We were very excited though, as we knew the site represented a paradigm shift in humanitarian response. In many cases, ReliefWeb replaced piles of fax paper.

Donors supported the project, as it represented a significant advancement in Open Source Intelligence, saving many agencies the need to compile this information; we were motivated by the fact that for the first time, this information was accessible to all. In addition, the site brought an unprecedented level of accountability to the aid sector as sequential and sectoral reports could now be easily compared.

Though we worked without an established content management system, we used a Lotus Notes database to store and classify all of the documents using sound library science principles from the beginning, thanks to Dennis King, our information manager in NY. But where a contemporary website uses code to take content from a database and display it on the web, we did all that part manually. Similarly, our first maps were made with PowerPoint, as we had no mapping tools.

The first major redesign in 1997 was a technical process to automate the Lotus Notes database to produce the updates automatically. This was way ahead of its time.

As the web was slow or non-existent in many field locations, we also developed ReliefWeb via email, giving those with only email and no web access daily updates on each emergency, and retrieve the actual documents they wanted. 
 Andy Andrea, the lead editor, had the idea of adding Jobs to the site. I was initially against it, but quickly realized that hiring was a key challenge in humanitarian response.

The transition from DHA to OCHA in 1998 was a challenging time and the project almost shut down. The team in Geneva was put on hold, and I took another job before being re-hired six months later as the technical project manager in Geneva, where I worked until 2006.

By 1999, ReliefWeb had become a standard and well-known information source in the humanitarian sector. So much so that at an OCHA staff meeting, the new ERC Sergio Viera de Mello quipped that when he visited a field duty station and said he was head of OCHA, the response was “Oh, you mean the ReliefWeb organisation?”

Since the beginning, ReliefWeb staff were based in Geneva and New York. In 2001, with the ambition of becoming a 24-hour service, we looked at a few locations in Asia, and opened an office in Kobe. This was chosen partly because OCHA already had a small office there, partly for its good connectivity and support from the Japanese government.

In 2002, we embarked on a major site redesign and had the good fortune of working with Adaptive Path, one of the world’s leading web agencies, based in San Francisco. The project was complex and challenging – and also pushed the limits of our technical platform, which was still, basically, Lotus Notes. After some painful delays, we officially launched the new site in early 2005 at the World Conference for DRR in Kobe.

We thought we could make a big splash, but just before the conference, a major tsunami took the world’s attention. In fact, Apple added a link to ReliefWeb on its home page, and the resulting traffic caused our servers to crash multiple times.

Of course, a major highlight was receiving the UN 21 award (now called the Secretary-General’s award) from Kofi Annan in 2004 for achievements in knowledge management.

All in all this brings back fond memories of great teamwork. Not many UN staff can look back on a project that is still thriving after 25 years. Personally I also had the privilege of moving on to start PreventionWeb in 2006, which is also going strong!