Frances Buontempo considers history, predictions about the future and how to shirk off writing an editorial.
Sometimes it’s good to look back over history in order to find inspiration, and to see how things have changed over time. Overload started with epilogues rather than introductory editorials, and occasional letters to the editor. If we were to revert to this format, that would let me off the hook. If any readers wish to send a letter to the editor, Overload@accu.org, please feel free. Several of the epilogues had brave prediction or questions about the future. How would namespaces work in C++? C++ is here to stay. Is there anybody brave enough to dismiss OO-COBOL? The first editorial appeared in April 1995. This considered the future directions of Overload and asked ‘ When are you lot going to stop messing around with the C++ standard? ’ [ Overload07 ] Seventeen years later, it seems the answer might be never. Sean Corfield also asked how many of the readers had email, stating ‘ Please use email, where possible, for submissions – I am allergic to paper ’ [Overload07]. As I explained last time [ Overload111 ], I am becoming allergic to emails, but articles in an electronic format are certainly easier to deal with than paper ones. How times change. If any readers don’t use email, please write in and tell us what you do with all your spare time.
Eventually paper crumbles away, old documents and code, on paper tape, punch cards, floppy disks and various other types of hard copies become unreadable. Either the medium itself decays, or we lose the means to read or understand the information. Taking a long view, I was struck by a BBC news article about proto-Elamite tablets [ Proto-Elamite ]; very old clay tablets with scribbles on. Though the clay tablets themselves have survived 5000 years, no-one knows what the inscriptions actually mean. It is suspected they might be some form of early accountancy, as many surviving writings from a similar era and area seems to be. I wonder if one day, no-one will be able to read a pdf or a Sage account. We shall see. Rather than writing our records on clay tablets, nowadays many people choose to write blogs on the internet. I suspect the internet will not disappear for a long time, but I wonder if we will lose these glimpses of the everyday at some point. This might lead to another ‘Dark Age’. Wikipedia describes the Dark Age as ‘ a period of intellectual darkness and economic regression that supposedly occurred in Europe ’ [ Wikipedia ]. The main reason seems to be few written records have survived from the time. Perhaps people in Europe were writing their own equivalent of blogs, not in the lingua franca of the time, Latin, and these have decayed away. Imagine that one thousand years from now, a historian tries to gather together evidence of how we live today. Will they find copies of Overload to use as a source? Or a blog? Of course, I am not suggesting I would rather you send articles in on clay tablets, or carved them into hillsides. I just wonder what now might look like, from the future.
Aside from the problem of using perishable storage media, the proto-Elamite tablets show the problem of communication. The Rosetta stone was a lucky find that allowed translation between Greek and Egyptian hieroglyphs [ Rosetta ]. For the proto-Elamite tablets, without a triangulation point, we may never know what they say. Rosetta code [ RosettaCode ] plays on the name to provide a rich resource of code challenges implemented in a variety of programming languages, allowing comparison and potentially is a great learning resource. They claim to have a total of 481 different programming languages, which is phenomenal. I wonder if they’ve missed any. How many different programming languages are there? I wonder how many different human languages there are. Recently I have been reading my bible, starting at Genesis and have just reached the story of the tower of Babel. It suggests originally ‘ The whole world had one language and a common speech ,’ [ Genesis 11 ] but God confuses peoples’ language so they no longer understand one another. Certainly, if you are confronted by a program in a language you don’t know, if may take a while to figure out how it works. Nonetheless, it is still possible to be bemused by a program written in a language you already know. My colleagues have recently written a tool to reverse engineer our config files, though that is another story. We have seen constant debates and considerations of the importance of naming variables and functions sensibly, in order to communicate our intent clearly. At the heart of this is avoiding the confusion of Babel. In August 2008, Ric Parkin’s editorial suggested, developing software is not so much a technical problem as a communication one. [ Overload86 ].
Technology has attempted to make in-roads in to automatic translation between languages to help communication. Various online translators exist, and seem to be improving. I have noticed a few recent news stories about live speech translation, not done by people, but by machines. Specifically, Google Translate has branched out and might now try to translate your spoken words live, presumably allowing you to communicate with colleagues distributed across the world over the phone even if one of you only knows English and the other only Japanese [ LiveSpeech ]. Had the Dark Ages never happened, and we all still spoke Latin, this wouldn’t be necessary. The live speech has grown from Google’s machine translation technology, which is a computer-driven pattern recognition algorithm, nudged by feedback from users. We shall see if the live translation takes hold. Technologies come and go. Recently, we have seen the death of Ceefax. Started in 1974, before the internet, it gave instant news, TV listing and weather forecasts on a television set capable of reading and displaying the information feed. The Ceefax pages were created manually – people monitored the incoming information and produced metres of punched tape to upload, after being carried up several flights of stairs to the ‘central apparatus room’. We are told, ‘It proved an invaluable service for the editor who used to alert his wife that he was about to leave Television Centre on his way home by using a back page on Ceefax. [ Ceefax ]
Watching previous technologies starting to grow and the predictions sparked by these is fascinating. I enjoy reading sci-fi, though I do wonder why these stories still tend to insist on the idea of flying cars. Sometimes such auguries are limited by a lack of imagination, and constrained by the current. As an antidote to ridiculous means of transportation, I have been reading The Last Man [ Shelly ]. Futuristically set at the end of the 21st century, it is free from flying cars. People still use horseback or coach to travel, the English monarchy has only just ceased, and wars are still fought with cannons and swords. The characters and story are played through with more conviction than many sci-fi books though. Heartily recommended for delicious gothic doom and cheer.
It seems that predicting the future is hazardous. “ Prediction is difficult, especially about the future. ” As either Neils Bohr or Yogi Berra once said: no-one seems to be sure who [ BohrYogi ]. See, predicting the past is hard enough. Would be traders will spend hours backtesting a new strategy, trying to see if they could make money from the historical data they used to form the strategy in the first place. And even getting the present right is difficult. For example, ‘nowcasting’ the weather is much more difficult than just looking out of the window. "These predictions are very expensive and not available to the public " [ Nowcasting1 ] and, I believe, frequently incorrect. To be fair, nowcasting isn’t trying to state what the weather is up to now, but rather what it will be doing in the very short-range, which does require accurate data on what is happening now, to predict rainfall, paths of tornadoes and so on [ Nowcasting2 ]. The met office gathers a huge amount of data and does some serious high performance computing to analyse it, producing thousands of forecasts a day. A variety of ways of trying to elucidate sense from data about now are constantly springing up. Twitter will tell you which subjects are currently trending, but not to be out-done ‘ Massachusetts Institute of Technology (MIT) associate professor Devavrat Shah has announced the creation of a new algorithm that can predict Twitter trends hours in advance. ’ [ MIT ] That will be hours in advance of twitter noticing, I presume, rather than the tweets actually being tweeted. That really would be something.
Sci-fi stories, along with letters to the editor, epilogues, and occasional stabs at editorials are all attempts to step back, and take stock of the now. They can draw on history, notice current trends, and try to make sense of it all. This is a time consuming activity, and as we have seen is increasingly being opened up to geeks armed with machine-learning algorithms. The next logical step is for the machines to write editorials for us. I have observed some automatic article generators of late. They seem to have started with an automatic Computer Science paper generator, [ SCIGen ] and sprouted new incarnations, such as a mathematics paper generator [ Mathgen ]. Some of these papers have been submitted and accepted by peer-reviewed journals [ ThatsMaths ]. A variant of this code this would get me off the hook. That does not let you, dear reader, off the hook. If you do feel the urge to submit an automatically generated paper, feel free, but rest assured, it will be read by our human review team, and we might just notice. Mind you, if it’s interesting, that is fine. I must stop for now, to brush up on my perl skills, in order to hack around the code from SCIGen and Mathgen, to get off having to write an editorial for next time.