Programming is Not Terrible
blog.m1cr0sux0r.com
blog.m1cr0sux0r.com
The reason open source software exists is because programming is terrible - if programming were the digital orgasm that these people pretend it is, people wouldn't have been bitching for years about how bad the open source ORM they're using is - they would have rewritten it already, twice.
Programming is a menial mental task for anyone with the capacity to properly abstract their real world problem into a solution that a computer can understand, with a few bits of crunchy debugging on top, that has no reward on its own.
We program, not because it is fun, challenging, or fulfilling, but because it is morphologically necessary for our continued existence as solution creators and problem solvers.
90% of the time spent programming is a slog. All of the enjoyment comes from the 10% of time spent solving cool problems, but more and more as open source technology gets better, sturdier, and more engrained into IT, the 10% problems are being solved before ldconfig is done doing its thing, and so you're left with 95% slog or 98% slog.
Are there domains where programming is still hard? Most definitely, but for most programmers programming is no longer drole.
Really? I have yet to meet one of these 25-year olds.
The majority of university courses teach you compilers(Parsing, Multi-pass optimisation, machine code output) and ML. I've built a fair few compilers in spare time, and I'm in no way special within my peers. Sure the stuff you learn probably isn't leading edge in terms of research, but unless you take a second job reading research papers, it isn't going to happen for most software developers.
The most enjoyable part of job is probably introducing better software processes, but once it is set up well, I get bored again.
They never solve the actual fundamental problem of computing - we still trust computers less than a four year old when it comes to understanding context and meaning in language.
"Does your enthusiasm permeate the walls of our office like a disease?" "In the tech world, you've got to be incredibly passionate about... really boring things." "They need to care more about our mission... than whether or not they're gonna get health care. 'Cause they won't."
That's the problem. If that's all you are doing 90% of the time after ten years it's time to move on to a different company.
Once a company gets large and has numerous engineers it tends to settle down into a environment where people have set roles as engineers where they work on simple boring projects.
But when you are building something from scratch in a small startup team you'll often have a more flexible role where you are writing core code, solving scaling problems, configuring servers and load balancers, tuning database queries, and numerous other things. From day to day you are learning new technologies and coding new things all the time.
That's what keeps software engineering fresh and enjoyable for me.
Trouble is, 90% of the time, it's very difficult to find an opening at a company that does more interesting things. :(
I'm just a newbie (started my programming job 9 months ago) but most of the stuff I do is kinda boring. I try to enjoy myself, and I believe clever people can find ways to enjoy what they're doing. Somedays are just a bore, with crappy documentation, crappy tools, and everything just crashes around you. Other days you find a new and interesting problem. This is when you must set your brain free and tackle that problem like an hungry wolf.
And in the end, it's just my job. I NEED to have a job to have my independence and enjoy some of life's pleasures. But it's just a job. It's a mean to an end. I need a job to buy videogames, go to the movies, travel, enjoy an expensive dinner, etc.
And I like programming. I program at home, making games and scripting stuff. I't just that 90% of software engineering is crap. You just need to wait for and enjoy the other 10%.
If most programmers want a challenge in their day job they could try learning how to communicate or how not to assume that everyone else is an idiot or even try and improve the processes their team use so that everyone benefits.
The biggest problem I find programmers coming up against is cummunication and understanding and yet 95% of them would rather point out how their immediate coding problem isn't challenging enough to their almighty brains.
I'm not - the social aspect of problem solving, and especially the social aspects of requirement gathering, is the best part of my job.
This is vile. You aren't semi-autistic. You are or you aren't. And what qualities are you referring to when talking about semi-autistic? Are you referring lack of communication? Stimming?
How the fuck does stimming have anything to do with slogging down coffee while programming?
Seriously, "semi-autistic"? People like you are vile.
The challenge is getting into a positive spiral instead.
I would do that, I just don't have enough time :(
In all seriousness, though, maybe the programming you've done has been a slog. I'm sure I speak for quite a few programmers here on HN in kindly asking you to not call us all stupid for liking something you don't.
Exactly. It seems that recently there have been a lot of young people who are treating the Software Engineer position kind of like the RN position. They think "There is high demand for this profession so I'm going to study for it in school and make big money, and this will be awesome."
But money !== happiness. RN is considered to be a good profession to study for in college. But not everyone who studies to be an RN can be happy taking care of sick and dying people all day, even if they are getting paid well. Likewise not everyone who studies to be a software engineer can be happy sitting at a computer writing code all day, even if they are getting paid well.
Some of us are happy to write code all day, and to "waste work" by rewriting systems frequently whenever the business changes directions because we love the challenge and the sense of progress as we get to see our coding skills improving month after month as we learn new technologies and techniques. I'd probably still be coding even if it was a low paying job, but I'm fortunate that because I enjoy coding so much and do so much of it I've become quite good at it and am able to make good money as a side benefit.
The thing is academic/personal assignments are probably novel and interesting, while 99% of business programming isn't. The jobs that have problems that are novel or interesting, are damn hard to get.
So if you enjoy it at school/hobby, that has no relation to if your going to enjoy it as a job. So you can easily end up hating your job.
I disagree. The monthly Who's Hiring thread here on HN is full of great jobs solving interesting problems. If you find those jobs "damn hard" to get maybe you need to learn some of the in demand skills or build a better portfolio of projects that you've worked on. A github full of open source activity can get you an interesting job quite easily.
A while back I put a very simple entry on an HN contractor thread and started getting two to three job offers a month in my inbox, many from very interesting companies doing interesting things. This was while I was still in college, not actively looking for a job yet. I accepted interview offers from a few of those companies and basically picked and choose who I wanted to work for, which eventually lead to me getting a job at the startup I work for now (storydesk.com) without ever needing to send out a single resume.
Maybe my experience is unusual, but I don't see it as being that difficult to get an interesting job as a software engineer if you leverage a community like HN.
Mobile Development Skills? I know that. The majority apps are boring though, interfacing with a web service, saving and displaying data etc.
Web Devs Skills? I know Django the best, but again 99% of people using web frameworks are just doing the normal inputting and outputting data out of a database routine.
ML? I know that, but most companies only hire serious academic researchers, or people with previous experience of it. I have MSc in CS with quite a lot ML in it, but most are looking Ph.Ds;
Compilers? I've built a few prototypes. But very few companies actually need this.
If you don't mind some completely unsolicited advice:
I just checked out your HN profile and found a link to your blog. I checked the blog and it is fairly empty with only two posts, and I see no Github link. That's the first thing a good startup company is going to look at if they think you might be a candidate for a position: your blog and your github.
I managed to track you down on Github via a Google Search and while two repos seemed empty for some reason, Harlan looks pretty cool and non trivial, showing a good level of technical skill.
The primary things I'd do if I were you is simplify and expand the message on your blog to explain who you are and what skills you have and link to Github prominently. Maybe on Github put up some more projects from a variety of different domains, and write a few more blog posts about those projects.
Additionally blog theme says a lot about you. Right now the ideal blog design if you want to come across as a "startup engineer" is something that looks like Medium or Svtble. Your blog theme looks like a Tumblr.
(No offense intended but another slight thing to adjust might be the photo on your blog. Obviously appearance isn't everything but in that photo it looks to me like you have a smirk instead of a smile for some reason. If you choose to include a photo it has to be one that makes you look like someone goodnatured and a great employee, because people hiring in behalf of a company are going to get one of their first impressions of you based on the photo, before they even look at the Github.)
Anyway, once again I hope you aren't offended by my well intentioned by completely unsolicited advice. For what it is worth I'm sure that you can get an interesting coding job at an interesting company.
Because 95% of startups I've come across are solving trivial problems.
* Code to render PDF's into images using asynchronous streams to make it extremely fast.
* The same thing for videos, to recompress them for delivery to iPads.
* Code to use AWS to create a service that can be fairly easily scaled up and down dynamically in response to demand.
* Creating a highly tuned Node.js server capable of serving thousands of requests per second. Even basic CRUD operations can be pretty interesting to code when you are fine tuning server system level settings, optimizing database queries, using caching, etc to make sure that response times are all around 10-15ms even at hundreds to thousands of requests per second.
The original article (which this article addresses) falls straight out of the most entitled kind of thinking ("the world owes me a living I enjoy"). I'd go further: even if you don't enjoy programming, there are far more horrible ways to earn a far worse living, and plenty of people have little choice.
Yesterday, I ordered pizza which was delivered by a guy older than me (and I'm no spring chicken -- I learned programming on an HP calculator) who wasted ten minutes finding my apartment because the delivery instructions were so poorly printed as to be unreadable.
http://www.calnewport.com/books/sogood.html
Enjoy
The important thing to keep in mind is that your day job is a way to earn an income. If you don't enjoy it, at least do yourself a favor and find one you care to show up to every day.
If you really want to work on "hard problems" and other groundbreaking and cutting edge technologies, you can always do it in your free time. Whether you want to write software at work and then do it at home is another story, but keep in mind that if you're successful, you can turn your side project into your career, one you actually do give a shit about doing every day.
The great thing about software is that you are only limited by your knowledge and determination. There are very few problems you need a massive expensive computer to solve (at a small scale, at least), so your home computer is usually just as capable as your work computer.
If you're bored with programming, why don't you start digging into the parts that really get you excited. For me, that's functional languages and dynamic typed languages, computer vision, and machine learning.
Don't be afraid to try new things and fail at them. If you just sit around doing the same mundane work all day, don't be surprised when you believe all of software is mundane.
I actually agree that programming sucks, but that's more to do with the fact that you have to turn an idea into logic and math before you can make a computer do it.
This is very true. It's easy to have some appreciation for being closer to the machine in prior eras, but it is a much better time to be a programmer today. There are many more tools, and much more collaboration available. The time from idea to completed project is much shorter. The ability to scale projects is much higher. The wait time for code to compile and run is much shorter. I can go on and on...
This post is exactly right -- and I've seen a few newly-hired programmers in their first job out of college who had a very difficult time adjusting to the "real world", where programming isn't about fun and games and self-fulfillment any more, it's about creating real-world value.
There's nothing unique to programming in this -- journalists discover they can't write about whatever they want, scientists have to follow the funding, etc.
However, I do wonder if programmers, on average, might have particularly inflated expectations of what their jobs will be like? Just like the original blog post expressed. And what makes this so?
It depends on what you are programming that makes it terrible or a joy.
Im right know payed to do another CRM in PHP and Id rather gauge my eyes out at the end of the day than continue with it. Im not solving any problems at all. Its just business/stakeholders/managers/whomever that decided a few months ago that their CRM needed a "facelift" and here we are. We use some new frameworks (slim), TDD, bootstrap/jquery, talk about agile/lean but in the end the problem we are solving is just some stupid shit like making a document or presentation better than what it was before. And the users are joyful. They got to look at something new while at work. They compete with each other with who will be picked to be among the pilot group, the new the cool. Like kids. A day looks like, fix some "bug" in on the server in an older application among arcane spaghetti-code, or some client-side jquery platitude (oh its you ie8 again?), write some new "functionality" which is fetch this data put it in there display it like this, take this data validate, serialize and send it over there. This is not even programming, it is bullshitting.
Then I go home and relax and read, after a while Im ready to play with python and C, solving minor and sometimes bigger problems. Trying out some machine learning at the moment. Ah the joy of hacking. It fills me up for next day.