Do you think this accurately reflects the working conditions of a developer?

Who said anything about trying to simulate a normal development day? This is about finding out some of their thought processes and if they can solve an exceedingly simple coding problem.

They may blank, they may freeze and they may just be the type of person that reflects on an issue for a while before the set out to code.

Where did they say you can't reflect on the issue? That's exactly what they want is you reflecting on the problem and describing how you'd go about solving it.

If someone tells you to write a function that takes a number N and prints out 1..N in pseudo code(Which is basically what the pascal problem asks when they provide the formula for a row) and you can't even talk your way through a possible solution you're not going to be able to complete any interview process, no matter how gentle.

I personally would search for the mathematical formula and a text description that I could put on my second monitor as reference.

So you're complaining that you can't search for something that the interviewer is already providing?

I personally never work from verbal descriptions, I document the issue, and place it into a ticket system in which I work the tickets based on priority.

If you are this incapable of listening to a two or three sentence problem description, ask the interviewer to print it out for you and write "ticket #1" on it? I mean is your nitpick really "They told me the problem out loud instead of writing it down?"

This is even ignoring the fact that taking verbal problem descriptions is an incredibly useful skill as a software engineer. I mean I guess I could tell people to file another ticket and I could have the joy of e-mail tag as I have to send off clarifying questions further delaying the task.

I assure you I am not trying to be argumentative, I am trying to help people understand why I and the author of the article believe that it is a flawed hiring practice. I have raised some good points, you can reflect on them and utilize them or you can discard them as rubbish. But I can tell you this, I have a lot of success in recruiting and I have learned by committing some of the very mistakes that I now advocate against. I used to buy into the Cargo Cult hiring practices, until someone smarter than me taught me how to truly identify talented individuals.

That's not the goal of the exercise however, nor is it a desirable one. The point is to a very loose test of basic ability. You can test the basic ability of a runner by asking them to do a quick sprint; the fact that this is not an accurate simulation of a race is irrelevant.

> do you think that they will be comfortable making those small errors in front of peopl

They should be. Any programmer should be comfortable making errors and owning up to them.

> Do you think that they jump into problems and that small errors are OK to them In this situation it is. In a "real world" situation, those would be caught by the compiler or unit tests. Of course really bad mistakes in this test would count against you.

> They may blank, they may freeze

Of course, this can happen in any type of interview. The point of interviews is to try and get the best balance of weeding out unsuitable people and finding the best ones. No matter what process you use, you will occasionally discard perfectly viable candidates...however this is preferable to accepting unsuitable ones.

> just be the type of person that reflects on an issue for a while before the set out to code.

...and that's fine...I'm not sure why you think this is some sort of "gotcha" test. It's perfectly fine to take some time...nobody is timing you. Again, this is a basic test to make sure you have the simplest skills and to get a little insight into your thinking/working. If you are uncomfortable with some aspect of the test, it is fine to say "well usually I take some time to think" or "I don't like writing on whiteboards"...etc.

You seem to be fundamentally misunderstanding the nature and point of the exercise. Errors are ok and expected because it's a simple test of reasoning and logic ability...likewise jumping in is ok because it's a simple problem.

Do you as a programmer spend a lot of time considering a simple if/else statement? Of course not...you jump in and write it. Again, we are talking about a simple algorithm here...given the description it basically writes itself for anyone with a modicum or programming experience.

The point here is to get a basic assessment of raw ability. The rest of the interview is to determine how well they might fit in the company, and any other gaps can be filled in with training.

> I personally would search for the mathematical formula and a text description that I could put on my second monitor as reference

That's the point you're missing. You don't need to search for that, it's been given to you. The description of the triangle contains the algorithm for constructing it. It's also so simple that you should be able to hold it in memory for the short time necessary, but if needed you could ask for a refresher.

> But first I would search for a standard library and avoid the effort all together.

That would negate the purpose of the test. Again, this is the most basic of basic programming tasks. It tests your ability to write simple logic and flow controls...something fundamental to writing computer programs.

Even if you solve a problem using mostly library functions, you still will need a few simple loops and logic to tie them together, and that's what this is testing.

> I personally never work from verbal descriptions

The verbal here is incidental...the point is you're working form some description or set of requirements. In this instance it's so simple that a verbal description will suffice, perhaps with one or two reminders. If you really can't hold such a simple description in your mind for 20 minutes I'd say you're not suited for any sort of mental work and might need to see a neurologist.

It seems your objections are based on a flawed premise, a fundamental misunderstanding of what the test is, how simple and easy it is, and what it is supposed to be testing.

If you look at want ads for day labor and other physical tasks it will often say something like "must be able to lift 50 lbs". When you went to the job site, they might ask you to lift a bag of concrete weighing 50 pounds. This is an exceedingly simple task and anyone fit for the job will be able to do it easily. It will quickly weed out anyone who is simply unfit for the work required, anyone who is fit will be able to do it easily.

More specific to your answers I understand your position, and I truly believe that you use it in this manner, the problem I have is that for every one of you there are 20 of the former me's who get so wound up in code tricks and trivia that we create unfair and useless interview filters. As such there is a perception among developers about code test, and that perception can be summed up that I am going to be tested on what the company thinks I should know, not what I know. For some of us, this creates a horrible and humiliating interview experience. In which a developer stands in front of a whiteboard analyzing their every move. No matter how much someone reassures them, they still have everything but logic running through their mind. It only takes one interviewer with an ego to permanently affect a developer in this manner and no amount of reassurance from a person they just met will help someone believe that their every move is not being judged.

Many people with Asperger syndrome fail tech interviews horribly but a good deal of them turn out to be wonderful developers. My best friend is affected by it, and I know from experience that he would never land a job if it where not for me. In spite of that, he is always the star developer at every organization that I bring him into. Now if someone sat him down and said, show me something you built and walk me through a routine, like the article suggest, they would be fascinated by his depth of knowledge.

If you don't know the formula and they haven't volunteered it, you can simulate this search simply by asking, and any interviewer using this for a proper purpose will happily write it on the whiteboard for you and walk you through it if anything is still unclear. It's not what's being tested.

What is being tested is that you know enough about programming to understand pretty much nothing more than basic flow control, basic arithmetic and perhaps trivial IO.

A use of a question like this checks that you've at least read "programming for dummies", and roughly understand the overall concepts of programming not whether or not you're proficient enough to even fill an entry level position. It weeds out the candidates that are completely unable to program at all.

> Do you think they will be able to put there best foot forward in such an unnatural situation? They may blank, they may freeze and they may just be the type of person that reflects on an issue for a while before the set out to code.

I've had smart candidates blank on trivial problems, but there's a very noticeable difference between people who blank and people who just have no clue. With people who blank, you get them to calm down, you ask probing questions to get them talking, even if it's about the weather, and they will eventually start offering up small bits and pieces of an answer that makes sense and gets them to "unfreeze" and start solving the problem. Whereas with the type of candidates this type of interviewing should be used to get rid of, you get absolutely nothing and/or total bullshit.

> But first I would search for a standard library and avoid the effort all together.

That's something you might say as a "by the way", but it isn't what the interviewer will be looking for.

> I personally never work from verbal descriptions, I document the issue, and place it into a ticket system in which I work the tickets based on priority.

Fine, so consider the description on the whiteboard the ticket with your highest priority would be my response to that if a candidate were to raise this objection (though be honest, if they did alarm bells would already be going off in my head).