Solve : Re: Programmer Benchmarks?

1.	Solve : Re: Programmer Benchmarks?
Answer» Quote from: BC_Programmer on July 27, 2009, 07:03:38 AM what did you expect? I was pretty clear with my first few paragraphs. Just ignore the last part, that wasn't really meant seriously; () the rest of it, however, fairly adequately argues your idea. That's good, I don't want to start a pissing match or flame war, but I will say benchmarking has been around long before computers to mark a standard of elevation and that successful bodybuilding is no more 'flex muscle/relax muscle' than successful programming is 'bit on/bit off'. Quote yes, but the way you've DESCRIBED what you want is kind of vague. I general you can learn alot about a persons abilities with a programming language (and even programming & computers in general) by just talking to them about their previous projects. I understand, I had anticipated being around to be a little more interactive. And as you seem to be beginning to appreciate, it's a vague idea to begin with. What about people who have no prior projects? As a rather vague example, a fellow chemist wants to write some software to perform some elaborate calibration or optimization calculations, what kind of criteria or tests would you run down in order to tell them whether they should DIY it or buy it? Quote My main beef with your original post was your implication that a person that knows C knows every programming language; while it's true they will be able to learn the language easier, it is erroneous to state that if they cannot then perform a task in ANOTHER, possibly completely foreign (and assuredly different) programming language they are deficient. Again, I didn't mean to imply anything about the language in particular, more about the programmer. As you said, someone who has learned a language will learn a language easier. Would you go so far as to say that the amount of time someone takes to learn their first language is irrelevant to their aptitude or skill? Quote However if your original intent (which, now re-reading your posts, I believe it was) you want to basically measure the abilities of a programmer, then there aren't any "global" type of tests you can perform; experience is just as important as skill, and in fact they go hand in hand. And during the process you cannot forget personality traits, it's important that they are ready to be part of a team, I have seen several programming efforts fall apart after hiring a promising candidate who has been used to working alone. On the other hand, outside selecting a job candidate, for example, to test oneself, there really isn't a good way to do that. You can't just TAKE a test/benchmark or whatnot and find out how good of a programmer, overall, you are. It just doesn't work that way; how good a person is at programming is completely context sensitive. I understand, to a point. However, hopefully without getting too divergent, we Americans tend to go overboard with 'political correctness' hoping people really are equal and that some just have untapped potential or some spirit that will fuel them to their goals when really they should 'not quit their day job'. There are no "global" bodybuilding tests either, just anecdotes, rules of thumb, and some bottom-of-the-barrel ultimatems. I certainly wasn't looking for a COOKIE CUTTER template to judge others and boost my ego. More judge myself and opportunities I see for myself in addition to getting some appreciation for when someone asks me 'How'd you do that?' whether I should RTFM them, just do it for them, or somewhere in between.There again, it may be a matter of time an budget, as well as expertise (of which I have little or none!) Is there software out of the box which will work? If not, will writing software for the task at hand be cost effective? I don't know about benchmarks, per se, but I'm rather sure there are fomulae available to compare person work-hours to the costs of computation.

Answer»

Quote from: BC_Programmer on July 27, 2009, 07:03:38 AM

what did you expect? I was pretty clear with my first few paragraphs. Just ignore the last part, that wasn't really meant seriously; () the rest of it, however, fairly adequately argues your idea.

That's good, I don't want to start a pissing match or flame war, but I will say benchmarking has been around long before computers to mark a standard of elevation and that successful bodybuilding is no more 'flex muscle/relax muscle' than successful programming is 'bit on/bit off'.

Quote

yes, but the way you've DESCRIBED what you want is kind of vague. I general you can learn alot about a persons abilities with a programming language (and even programming & computers in general) by just talking to them about their previous projects.

I understand, I had anticipated being around to be a little more interactive. And as you seem to be beginning to appreciate, it's a vague idea to begin with. What about people who have no prior projects? As a rather vague example, a fellow chemist wants to write some software to perform some elaborate calibration or optimization calculations, what kind of criteria or tests would you run down in order to tell them whether they should DIY it or buy it?

Quote

My main beef with your original post was your implication that a person that knows C knows every programming language; while it's true they will be able to learn the language easier, it is erroneous to state that if they cannot then perform a task in ANOTHER, possibly completely foreign (and assuredly different) programming language they are deficient.

Again, I didn't mean to imply anything about the language in particular, more about the programmer. As you said, someone who has learned a language will learn a language easier. Would you go so far as to say that the amount of time someone takes to learn their first language is irrelevant to their aptitude or skill?

Quote

However if your original intent (which, now re-reading your posts, I believe it was) you want to basically measure the abilities of a programmer, then there aren't any "global" type of tests you can perform; experience is just as important as skill, and in fact they go hand in hand. And during the process you cannot forget personality traits, it's important that they are ready to be part of a team, I have seen several programming efforts fall apart after hiring a promising candidate who has been used to working alone.

On the other hand, outside selecting a job candidate, for example, to test oneself, there really isn't a good way to do that. You can't just TAKE a test/benchmark or whatnot and find out how good of a programmer, overall, you are. It just doesn't work that way; how good a person is at programming is completely context sensitive.

I understand, to a point. However, hopefully without getting too divergent, we Americans tend to go overboard with 'political correctness' hoping people really are equal and that some just have untapped potential or some spirit that will fuel them to their goals when really they should 'not quit their day job'.

There are no "global" bodybuilding tests either, just anecdotes, rules of thumb, and some bottom-of-the-barrel ultimatems. I certainly wasn't looking for a COOKIE CUTTER template to judge others and boost my ego. More judge myself and opportunities I see for myself in addition to getting some appreciation for when someone asks me 'How'd you do that?' whether I should RTFM them, just do it for them, or somewhere in between.There again, it may be a matter of time an budget, as well as expertise (of which I have little or none!) Is there software out of the box which will work? If not, will writing software for the task at hand be cost effective?

I don't know about benchmarks, per se, but I'm rather sure there are fomulae available to compare person work-hours to the costs of computation.

Solve : Re: Programmer Benchmarks?

Discussion

No Comment Found

Related InterviewSolutions

Reply to Comment