How wierd. Many of the well-known chippers in Ireland are run by families from the same comune in Italy.
In the late 19th and early 20th century a significant number of young people left Casalattico to work in Ireland, with many founding chip shops there. Most second, third and fourth generation Irish-Italians can trace their lineage back to the municipality, with names such as Magliocco, Fusco, Marconi, Borza, Macari, Rosato and Forte being the most common. Although the Forte family actually originates from the village of Mortale, renamed Mon Forte due to the achievements of the Forte family. It is believed that up to 8,000 Irish-Italians have ancestors from Casalattico. The village is home to an Irish festival every summer to celebrate the many families that moved from there to Ireland.(via JK)
Think we’ll be watching some of these in work soon — Jez Humble’s talk (the last one) in particular looks good:
Amazon, Etsy, Google and Facebook are all primarily software development shops which command enormous amounts of resources. They are, to use Christopher Little’s metaphor, unicorns. How can the rest of us adopt continuous delivery? That’s the subject of my talk, which describes four case studies of organizations that adopted continuous delivery, with varying degrees of success. One of my favourites – partly because it’s embedded software, not a website – is the story of HP’s LaserJet Firmware team, who re-architected their software around the principles of continuous delivery. People always want to know the business case for continuous delivery: the FutureSmart team provide one in the book they wrote that discusses how they did it.
Currently, histograms are static structures: they are created from scratch periodically and their creation is based on looking at the entire data distribution as it exists each time. This creates problems, however, as data stored in DBMSs usually varies with time. If new data arrives at a high rate and old data is likewise deleted, a histogram’s accuracy may deteriorate fast as the histogram becomes older, and the optimizer’s effectiveness may be lost. Hence, how often a histogram is reconstructed becomes very critical, but choosing the right period is a hard problem, as the following trade-off exists: If the period is too long, histograms may become outdated. If the period is too short, updates of the histogram may incur a high overhead. In this paper, we propose what we believe is the most elegant solution to the problem, i.e., maintaining dynamic histograms within given limits of memory space. Dynamic histograms are continuously updateable, closely tracking changes to the actual data. We consider two of the best static histograms proposed in the literature , namely V-Optimal and Compressed, and modify them. The new histograms are naturally called Dynamic V-Optimal (DVO) and Dynamic Compressed (DC). In addition, we modified V-Optimal’s partition constraint to create the Static Average-Deviation Optimal (SADO) and Dynamic Average-Deviation Optimal (DADO) histograms.(via d2fn)