WexQuers

11/01/2022

Why everyone should have to learn computer programming
News that numerous cathedrals are offering short courses in Latin is a reminder of the long decline of the language over the years. It was a core subject in the British education system until fairly recently – and not because anyone planned to speak it, of course. It was believed to offer valuable training for intellectual composition, as well as skills and thinking that were transferable to other fields.

It may have been the right decision, but when it was ultimately decided that these advantages were outweighed by Latin being a dead language we arguably lost that intellectual training in the process. This is why we want to make the case for moving another discipline to the centre of the curriculum that offers analogous benefits – computer programming. And unlike Latin, it is anything but dead.

Noam lore. Brian Talbot, CC BY-SA
There are many computer languages for different purposes. C and C++ remain the fastest to execute and are used by the gaming industry, for instance. In the internet era, much of the page design is done with the likes of JavaScript or PHP. Meanwhile Python has been rapidly gaining a reputation as a general purpose code that is easy to learn.

There are many parallels between natural languages and programming languages like these. You must learn to express yourself within the rules of the language. There is a grammar to comprehend. And what you write must be interpretable by another human being. (Yes, it must be interpretable by a computer. But just as Noam Chomsky’s example of “colourless green ideas sleep furiously” is grammatically correct nonsense, you can write obfuscated computer code that no one else can decipher.)

People who program can communicate with computers, which is becoming more and more important now that computers have a hand in almost everything. In today’s IT-literate world, we are all expected to be fluent in word processing and spreadsheets. The next logical step is to be able to program.

The younger generation are already exposed to computers almost from the day they are born, which explains for example Barclays bank’s recent launch of Code Playground, an initiative to engage young children in the basics of programming via a colourful website.

Problematis solvendis
There is a myth that only maths geniuses are suited to programming. It is more accurate to say you need a logical approach and an ability to problem solve. Just as Latin constructs reinforce communication, programming constructs reinforce problem solving. It teaches you to break a problem into achievable chunks and to think very precisely. And once you have mastered the basics, it opens up great potential for creative thinking.

Then there are specific workplace benefits, such as for businesses that are building a bespoke piece of software. Errors sometimes occur when documents outlining in English how a program should work are translated into computer code. Those who have an appreciation of a programming language can write these more clearly. Indeed, businesses usually have to employ specialist analysts as intermediaries to help with this translation process.

As computers become more dominant, those who don’t know how to think in this way risk being increasingly left behind. We can foresee a time when greater numbers of people become interested in learning to program for themselves, but in the meantime there is a great case for making the basics of computer programming a core skill at school.

One candidate language would be Python, it’s freely available and one of the easier programming languages to learn – compared, say, to C/C++. It has grown in popularity in recent years, initially for this simplicity but lately because it has been adopted by the big data community. It is likely to be around for a few years and not become a dead language any time soon. There are plenty MOOCs (online courses) to get you started.

Lighten your load with code … Mclek
If a teacher walked into a classroom and told today’s students they were going to study a dead language, you can imagine the reaction. Imagine instead introducing them to an easy-to-use programming language which is probably already installed on their laptops. It can allow them to automate many boring tasks such as checking email and sending out pre-written responses; or receive custom notifications by text; or download files or copy text from a website whenever it updates.

It’s time that those in charge of education policy recognised the shift in employability skills and the need for a new generation of problem solvers. We may have reached the point where the three Rs of education – reading, writing and ‘rithmetic – should become the four Rs, with the addition of programming. Or 'rogramming, as we would soon get used to calling it.

11/01/2022

Explained: why a reboot is the go-to computer fix
It’s the most common answer to our computing woes: when your PC or mobile is playing up, try turning it off and on again. Or, alternatively, rebooting.

To understand the concept of a “reboot”, it’s helpful to first understand what a boot means as far as computers are concerned. The word comes from the expression “pull yourself up by your own bootstraps”, which I’ve never fully understood, but apparently means “improve yourself by your own efforts”.

In a computer, the only program physically built into the computer hardware is a tiny one, called the “bootloader”. When the computer starts up, this program gets control and loads, or “boots” another, much larger, program which serves as the “operating system” for the computer. We know these systems by such names as Unix, Mac OS, Android and Windows 10.

The operating system does for your computer what your parents did for you during the first five years of your life. It organises the allocation of resources, fetches things and puts them away – and controls what the programs can and can’t do. However, sometimes the operating system can get itself into a bit of a state – like your mum or dad did when the doorbell rang just as the washing machine sprang a leak and your pet rabbit escaped.

A clean slate
If we give the computer too many tasks to run – or a set of physical events occur in a sequence that the software writers weren’t expecting – then tasks can get “stuck” in memory. Computer scientists talk about a “deadly embrace” that occurs when task A is waiting for task B to do something, and task B is waiting for task A to do something, causing them both to get stuck.

giphy.com
In addition, as tasks run, they fetch and use resources such as computer memory and, over time, the arrangement of these resources will become fragmented and harder to manage, just like it is difficult to find things in an untidy bedroom (which is probably why your parents made such a fuss about it). A reboot may also be a temporary fix for problems caused by hardware that is becoming unreliable, particularly if things start to go wrong when components get hot.

Modern operating systems are very adept at spotting and removing stuck processes and also work very hard to keep things tidy, but sometimes a computer can reach a state where the best thing to do is start again from scratch. A reboot removes every task and then restarts with a clean slate.

As a computer scientist I’m always looking for the easiest way to solve a problem and rebooting a computer is a good thing to try first, before looking for more complicated reasons why a system is running poorly.

Fix-all?
There are two flavours of reboot, which are often called “warm” and “cold”. You do a “cold” reboot by actually turning the computer off and on again. A “warm” reboot, meanwhile, just reloads the operating system. Sometimes a warm reboot will fix your problems, but if some of your hardware has got itself into a state where it is not responding to any signals from the outside world, you might need to reach for the power switch.

One thing reboots cannot fix, however, is malicious software such as viruses. These horrid bits of program usually insert themselves into the boot process so that they get control next time the computer starts up. The only way to get rid of these pesky intruders is to scan your system, find them, and remove them.

In my experience the need for reboots is decreasing over time. These days I find that the main reason why I have to reboot my machine is to install updates. This is because it is very hard for an operating system to update parts of itself while it is running – rather like trying to repair an aircraft in flight.

Some systems are never rebooted. Things like air traffic control systems and the programs that run our nuclear reactors are left running continuously. These systems have the advantage that they only run one particular program and their operating system can be built around this code. However, for general purpose machines like the ones on our desks and in our pockets, the need for reboots will remain for a while. For me, it’s just a necessary consequence of having such a powerful and flexible device at my fingertips.

To hear Rob’s reboot joke and more stories around the theme of “reboots” – from a Syrian rebel group’s rebrand to the resurrection of an old drug that could tackle brain tumours – check out the fifth instalment of The Conversation UK’s podcast, The Anthill.

11/01/2022

Happy 30th anniversary to .au domains – what comes next?
This year marks the 30th anniversary of Australian domain names – websites ending in .au. As of June this year, more than three million such domains have been registered. This is a lot for a country of only 24 million people, but especially notable when you think there were only a few hundred thousand domain names as recently as 2002.

This explosion in registrations is the result of the constant evolution of our system – changes that have not only opened domains up to more and more people, but will set us up for the fourth decade of Australian websites.

It began with one man
The story of .au, Australia’s Top Level Domain (TLD), really began in March 1986. It was then that University of Melbourne network administrator, Kevin Robert Elz – known to most as Robert Elz – was given authority to administer Australian domain name registration. Elz set the rules for the system and held this authority from 1986 through the 1990s.

There were a couple of different phases during Elz’s tenure. The first ten years were marked by slow growth and informal management. But from 1996 there was a boom in demand as the internet rapidly became a central part of our lives and businesses jumped on board.

1996 was the beginning of the commercial domain name industry in Australia. The boom in demand for Australian domain names led Elz to delegate responsibility for Australian second-level domains (2LDs). In her book on the domain name registration system, academic and lawyer Jenny Ng writes:

In October 1996 Elz granted a five-year licence to administer com.au to his employer, the University of Melbourne through Melbourne IT Ltd.

After this, registering a domain was no longer free.

A formal system
As commercial players moved in, more formal management of the Australian domain name system (DNS) was established. In 1997 the Australian Domain Name Administration (ADNA) was created, and in April 1999 its role was transferred to the .au Domain Administration (auDA).

In December 2000 the Australian Government formally endorsed auDA as the appropriate body to hold the delegation of authority for .au. In 2002, new policies and a new registry, operated by AusRegistry, came into operation. With this, the contemporary Australian domain name system (DNS) came into being.

Reliable data are not available for the early years, but when the first registry was launched in 2002, there were 282,632 domains, which grew to 500,000 domain names in 2005, and 710,428 on 30 June 2006. We surpassed a million domains registered by 2007 and three million in 2016; about 87% of these are .com.au.

This growth was facilitated by relaxation of the previous rules as a result of successive policy review panels.

Evolution of Australian policy
Domain name policies during the first 10-15 years reflected the beliefs and assumptions of the Internet’s pioneers. One guiding principle was that no one should be able to gain an undeserved advantage over other participants. So it was only possible for a registrant to register one domain name, and it was not possible to register a generic term (such as a common noun) or a place name. The principle of “first come, first served” was enshrined; there was no hierarchy of entitlement to a particular name.

Since the first formal policy was developed and implemented in 2002, most of these restrictions have been abolished, and the Australian DNS has followed the path of evolution in other countries, albeit more slowly.

Some fundamental principles remain: first come, first served; the requirement that the registrant for an Australian domain name must be Australian; provisions to reduce abusive practices; and a requirement for a link between the registrant and the name registered. As a result, the Australian DNS is regarded as relatively well-administered, trusted and stable. But we aren’t done yet.

Direct registration: the next step
In August 2015 I wrote about the work of the 2015 Name Panel established by auDA to review domain name policy and propose changes to the board of .auDA. Policy change proposals are adopted by a consensus of each panel, and the panels must be broadly representative of the industry and domain name users.

The panel made one major recommendation: in addition to registering second-level domains (like .com.au, .edu.au, and .net.au), Australians should also be able to register directly (such as example.au). Both the UK and New Zealand have already made this change.

The final report of the Panel was presented to the auDA Board in December 2015. The Names Policy 2015 website includes the text of all non-confidential submissions, panel minutes, and other panel documents; there was also a minority report by four of the 23 panel members.

The auDA Board considered and accepted the final report in February 2016 and in April agreed to implement the recommendation. Australians will soon be able to register domains directly.

Against this backdrop, auDA will soon consider the best way to manage one of its largest policy changes. This is the move from the strict hierarchy of names which has prevailed for thirty years using a TLD and 2LDs, to a dual system which permits both direct and second-level registrations. The changes will set a direction for the fourth decade of .au.

11/01/2022

11/01/2022