Programmers: Stop Calling Yourselves Engineers(theatlantic.com) |
Programmers: Stop Calling Yourselves Engineers(theatlantic.com) |
Engineer started as a word meaning one who is skilled in the design, construction, or use of, surprise, engines.
The word evolved in its use to later encompass other systems that can build: civil engineering (bridges and buildings), mechanical engineering (robots and mechanical systems), and computer engineering (actual hardware).
Software is merely the latest system. Many universities grant software engineering degrees and apply principles of engineering to software.
The whole point of this article is to treat "engineering" as a reserved and hallowed word and vocation. Its name and usage must be protected at all costs lest we let just anyone call themselves an engineer.
It's an argument of arrogance and appeals to maintaining the status quo of exclusionary system. You didn't take a certain test or wear an iron ring? Sorry, you can't call yourself this one title. The title of "doctor" has similar historical exclusion, with infighting common amongst multiple professions (e.g., medical doctors not thinking dentists or PhD's are actual doctors).
Language evolves. Stop being protectionist about a certain word.
I think what it important in this discussion is the fact that "software engineer" right now is an aspirational term and that we don't use formal methods to verify software behavior, and instead rely on changing best practices.
Egineering implies a more rigorous process than most software really gets.
How ridiculous that this is the introductory sentence. I guess leave it to the Atlantic to assume there are no engineers working for private enterprise.
Not to say that they have not achieved some amazing and useful things, but I remember one of my lecturers at Uni saying, "At many times during your career you may well encounter unqualified and/or inexperienced people who can seemingly deliver comparable results by taking extreme or unwarranted gambles, but it is the engineer who should be able to be relied upon to do these things for the benefit of society consistently and safely, to a known budget and schedule".
Now many people would (maybe rightly) think that is fantasy land, but there is an art to navigating the conflicting requirements and expectations within a defined budget up front when you are creating things with a very real potential to directly kill people if they go wrong. But this also brings a certain amount of conservatism that seems limiting at times of extremely rapid progress within a field of endeavour, such as we have seen in SV for the last few decades.
Every mature engineering field begins with failures. Henry Petroski [1], the archetype of the classical engineer IMHO, believed failure analysis was at the heart of engineering. Non-engineers, including scientists and health-care professionals, make the false assumption that the practice of engineering is deterministic. The goal of engineering is to understand systems well enough that they seem deterministic but engineering knowledge never starts off that way. The formal structural engineering standards represent a history of failures along with the best practice for dealing with each failure mode. The process is iterative, just like software development.
Neither the term "computer science" nor "software engineering" is an accurate description but as someone formally trained in engineering and computer science, I've yet to encounter a more accurate label. In my experience, the complaints about the term engineering are about credentialism and the solution is to use the term Professional Engineer for those who have earned the credential.
For those who don’t care about the engineering profession, you can call yourself “software doctor” or whatever. I really couldn’t care less.
"The traditional disciplines of engineering—civil, mechanical, aerospace, chemical, electrical, environmental—are civic professions as much as technical ones. Engineers orchestrate the erection of bridges and buildings; they design vehicles and heavy machinery;"
so what? Traditions change, names change, everything changes, software engineers will be called engineers and no one cares if they build bridges or not. I can tell you if i connect a xmpp service to some twitter account i also built a bridge. so software engineers do build bridges! And talking about "tradition" in a digital revolution is also weird, because right now everything is changing and nothing will be like before. This wolrd will never be the same as 1960, so why stick to old naming convetions anyway. This article is nothing but time-wasting, sorry if i offend someone with that but it is my honest opinion.
In the lowercase sense of engineer, I think it is fine to say, "I have engineered this solution." But to claim one is an Engineer because you engineered a solution misses much of the point of progressional licensure.
I, of course, come from the perspective of a PE not a coder. I have no doubt that influences my opinion on the matter. It doesn't mean I think the word engineer has one and only one meaning, context is important.
To use the doctor analogy, to say one is a doctor is vague and requires context. To say on is a medical doctor disambiguates the statement. Same applies to engineer. The inclusion of context is needed to make sure you are not talking about one who operates trains versus one that does structural design.
TL:DR - Context matters, make it clear what sort of engineering you perform.
There is a debate about certification but its not one that some one not working as an "engineer" - has much to say.
What would such a test look like? The NCEES sample questions appear to be paywalled. The SWEBOK link 404s, but I found a pdf with some info: https://ieeecs-media.computer.org/media/education/swebok/swe...
Given how vast the field of software engineering is, would the test be similar to FAANG interviews where you need to solve some algorithmic leetcode puzzle? Or would it require knowledge of a specific programming language and all its pitfalls?