What is Google-fu? What does the code used by hackers in movies actually say? Should programmers really be worried about AI? And how to flush out broken code like a prowling hunter.
We learnt quite a lot in this edition of Source Code.
But that would come as little surprise to those who know Cameron Stevenson.
In the Cranmore office, he’s well regarded for his general smarts and mage-like coding. There’s not a lot he doesn’t know. At least, not a lot worth knowing.
A Queen’s alumnus (crushing his BEng in computer science with a first), Cameron’s trajectory had once been trained on a life in accountancy before a visit to his school that inspired a recalibration.
“My school introduced a computer course and people came in to talk about their careers. There was a software dev who showed us all this green code on a screen and I was just mesmerised. It was like when you watch 80s or 90s films about hacking – I needed to know how to do that.”
Today, however, Cameron finds himself mostly scoffing at the movie code that once seemed so mystifying.
“It drives my girlfriend crazy,” he says. “Every time there’s code on screen in a film or TV programme; I have to pause it and see what it says. It’s almost always nonsense and occasionally hilarious. But it’s every time, and I’ll be there explaining what it’s supposed to do and why it wouldn’t work.”
Cameron says he was “the typically techie kid” when growing up, eschewing footballs for consoles.
“Me and my friend built a PC when I was a teenager,” he says. “Even my subjects in school – maths, computing, physics – I was just always that way inclined.”
We wanted to know if, with technical adroitness seemingly in his blood, stories of AI replacing developers was something that concerned him.
“No.” he says flatly. “It doesn’t concern me. When you see some of the code it spits out, if you were to take that and drop it in to your own code, it just wouldn’t work.
“It’s not the cleanest code either and that’s because the data it pulls from isn’t updated. Some friends of mine have been discussing it but I think we’re pretty safe.”
Cameron’s view is that AI is “more a tool than competition”.
“It’s like a plumber worried they’ll be replaced by an automatic wrench,” he says. ”There will always be an engineer needed to maintain it.”
Working across Cranmore’s full stack, including the proprietary Agility365 software and Microsoft Azure, Cameron has picked up many developer tricks and tips along the way.
“First of all – Google is your friend,” he says. “A few of us here call it Google-fu. There are ways to use it to get more accurate results for what you want. Being good at search is a skill in itself.”
He explains there are little-known “search operators” to use when Googling, helping to zero in on the right results fast.
“Using quotation marks for example – that means ‘the result must include X’,” he said. “There are many others, like colons and other symbols.”
[A full list can be found in this article – Google Search Operators: The Complete List (44 Advanced Operators)]
Cameron also advocates the value of “fresh eyes” on a problem: “Never be afraid to ask someone to look at it,” he says.
“And never underestimate how often broken code is the result of the smallest thing – a capital letter, or a missing full stop. Don’t assume the problem is something complicated.”
Indeed, his approach to finding the ‘breakpoint’ is much like a hunter stalking its prey.
“When I’m debugging, I try to isolate the issue by establishing a start point and end point for the error. Then I keep moving the starting point forward and the end point back, closing the gap on either side of the problem. And, as you work your way in, you eventually isolate your breakpoint.”
Clever boy.