How does this system work?

The website that you are using is a good example of integration of strongly heterogeneous systems. Basically the whole structure is composed of three components. Two of them are "internal" components of the network where the server you're using to see these pages is located. The third one is completely independent and it is used only to retrieve the data.

• The website of the ICPSM (Istituzione Centro Previsione e Segnalazione Maree) of the Comune di Venezia: in a page published through a HTTP server the data about the last levels of the tide can be found. This page is in a standard format studied to make the computer elaboration easy but not suitable for the human reading. Its update is not in real time but it's about 10 mins late. The data are the ones given by the probe in Punta della Dogana.
• A PHP script is used to retrieve the new data from the page previously explained. Once fetched the new data are stored in the database. Initally this part was written as a Java program but after an important update it has been decided to use only the PHP language;
• The PHP engine that produces the pages you're looking at: it retrieves the data from the same MySQL database used by the Java server and it elaborates them to publish them with the graphical interface you're seeing. A proper library that uses the PHP GD libraries has been used to produce the charts. Also the data are retrieved from the database using a proper library.

The delay of the data published by ICPSM and the update made every 20 mins by the Java server, generates a global delay of even 40 mins. By the way, the given data are enough to have a good information about the current level of the tide.

The two "internal" components have been completely realized by Alessio Zennaro in every part during fall 2011.