Can you fizz milk?
There was a piece on the one show this evening about the sodastream, and its association with the 1970s, which is when I mainly remember it from. If I had to list 10 objects that reminded me of my childhood in the 70s, the sodastream would be on the list (along with Rubik’s cube, meccano, lego, um …).
In my teens I remember one weekend wondering what would happen if I put a bottle of milk in a sodastream, rather than water. I put the question to Dad, who bounced it back to me, “what do you think would happen?”. I thought about it for a bit, and figured, hey, milk is largely water…
I filled the bottle and put it in the machine, taking extra care that the bottle was filled to the right level, it was properly inserted, the lever all the way down. I then gave one long press of the button. Milk came pouring out of the machine; I released the lever and opened it up: more milk came out.
I went and told Dad: “yeah, that happened when I tried it as well”!
When Mum heard she said Dad had to clean the machine out, as he knew that it would have to be disassembled to clean the milk out, so we wouldn’t get a machine stinking of milk gone off.
Geek’s guide to correct toilet usage
This appeared on the wall of the toilet at work:
Control-C: it makes your programs run faster!
As a postgrad in the late 80s I made some extra book money acting as a helper in the computing lab. A few of us would be posted there, for undergrads to come to for help. This tended to be focussed at the start of the year, when there were groups discovering Unix and programming for the first time.
One time an Irish girl called me over, saying that she couldn’t understand what was going on: she thought her program looked right, but for some reason each time she ran it she got partial output, and varying amounts of output each time. I don’t remember the specifics, but their assignment involved writing a program that was generating various values and writing the results in ascii tabular form to a file.
I had a look at her source code. Everything looked fine. She showed me the file generated by her last run, and indeed it looked truncated. Hmmm. “Ok, can you run your program for me, so I can see what happens?”.
She typed ./a.out and hit return. Her left hand darted to the keyboard and she hit Control-C. I was still mentally processing this when she cat‘d the output, and turned to me: “See!”. It did indeed contain partial output. Again.
“Um, can you just run it again please?”. I figured I must have not seen right. But once again she typed ./a.out, hit return, then whap! she hit Control-C. I asked her why she hit Control-C every time she ran her program: “I discovered that Control-C makes the % prompt come up quicker”.
In her mental model, the % sign appearing meant “the previous command has finished”. The undergrads were all working on a VAX-11/780, which at times was grindingly slow. She’d stumbled across a useful technique for making the prompt pop up more quickly. It hadn’t occurred to her that hitting Control-C was the cause of her problem, she just thought it was making the faulty program run faster.
This has stayed with me ever since, partly because it’s a funny story, but more as a reminder that when you don’t fully explain a system to people, they build up their own mental model about what’s going on, and as long as it fits their observations and experiences, it must be right!
Wii Tennis: it’s just like chocolate!
Friends came round today with their two children, Danny aged 7 and Lizzie aged 9. They’d recently been snorkeling while on holiday, so I asked if they wanted to play Endless Ocean (a diving exploration game) on my Wii. After an hour or so in which they all played, we switched to the Tennis game in Wii Sports.
Once they got the hang of it we were treated to a very entertaining half hour as Danny and Lizzie threw themselves around the room (the light fitting is still bent from our first evening playing Wii sports!). Danny won a match, then Lizzie, and they wound themselves up more and more. I got too close at one point, and caught a backhand. When it was suggested that they stop, as they were getting a bit wired, Lizzie announced “it’s just like chocolate!”
This is the children’s equivalent of saying something’s as good as sex, perhaps?
Pink box testing
Some years ago (1999′ish) I was in a job where I needed to bring in a lot of software contractors in a short space of time. As a result I ended up dealing with a good number of IT contracting agencies. To help the process I wrote very specific job / task descriptions, and sent these to all agencies.
Generally I would receive agency-specific CVs, where they had reformatted the candidate’s CV, and in places tweaked the content. Over time it became clear that some agencies certainly did a lot more than reformatting, adding things they thought I wanted to see. This was really annoying, as the fabrication would reveal itself at some point in the process, sometimes after wasting time on a candidate.
We needed a software tester, and while writing the task description I decided to see how far the agencies would go. The descriptions all had a section listing required skills and experience, and another listing desirables. In the required section I listed “Pink box testing experience”. There’s no such thing as Pink box testing (see below) – I made it up to see if I received any CVs with pink box testing listed.
Sure enough, a week or so after putting out the description, I received a CV for a software tester. He clearly had a lot of relevant experience, but the agency had added pink box testing as one of his strengths (it was an agency modification, not a claim of the candidate). It was fun ringing up the agent and explaining that he’d been caught in my trap, and listening to him trying to weasel his way out.
No such thing as Pink Box testing?
I can’t remember exactly when this was, but from my memory is was most likely 1999 or early 2000. At the time I did a quick web search to see if there was any such thing as pink box testing, and a quick skim of my testing books. Certainly didn’t seem to be at that time. As several commenters have pointed out, there are (now) several valid definitions of pink box testing, not all of them appropriate for polite conversation.
The term “pink box” was clearly relevant to phone phreakers at that time, so the term “pink box testing” made sense in at least that context, but in the context of software testing it didn’t.
