|Perl: the Markov chain saw|
I've been re-reading a number of old threads, notably the ones about Perl certification. There are (nominally) good reasons for creating a Perl certification process. These are similar reasons why most people want to see a 4-yr degree (or equivalent) and they boil down to "You can actually finish something." Now, personally, I think seeing someone manage a CPAN project (or be vouched by someone who has) is a stronger indication of usefulness, but that's beside the point. And, more importantly, that's not the reason everyone seems to be giving. Everyone seems to be talking about PHBs and what they do(n't) understand.
Recently, I started my own company and am working through how we're going to hire people in 6-12 months for our first development team. I'm also giving a lot of thought to what kind of company culture we're going to be creating, for both the IT staff and the (much larger) non-IT staff. And, given that most of the staff won't be in my competency (which is good!), how I plan on measuring their fitness to be part of my company.
What I've come up with (which has yet to be tested) looks to be a variant on the "What do you have on CPAN?" metric. I really don't care about what tests you've passed. I also don't care (much) about where you've worked (unless they also use this metric). What I really care about is what you can show me you've done. Because, at that point, I can now measure to see how you work will be helpful for what I need to have done. I want to see things like:
None of these things are deal-breakers. No-one is a perfect fit for a job. The key is for management to know where your weaknesses are so that they can adjust the job to you. If you don't let me know, I can't make your working experience better for you. Remember - I, the manager, am taking a risk by hiring you. You will be near-worthless for the first 8-12 weeks. You won't hit your stride and be fully integrated until the end of 6 months, period. At roughly $80/hr (the total of salary, benefits, your workspace cost, etc), that's a $25k-$40k investment up front with little, if any, return. I most likely don't recoup my investment until at least 8 months have passed. I don't make significant money until a year has gone by.
Given that, I have a huge incentive to make sure you're as productive as you can be by giving you tasks that play to your strengths and avoid your weaknesses. And, that also makes you happier, which means you'll stay longer. I can't do that unless I know what your weaknesses are. Certifications won't tell me that. Only your projects will. When I was in the market, I always used certification requirements as a measure of how much I didn't want the job. As an employer, I will likely use certifications as a discount on your employability by my companies.
(Of course, this doesn't count certifications required to practice the craft, such as CPA.)
My criteria for good software: