How to not waste time on a side project when trying to get a job
hackcareer.com
hackcareer.com
Side projects are more and more being touted as a way to "stand out" from the enormous crowd of people trying to get tech jobs. A company choosing among 100 candidates, all things being equal, may bias towards the few with active interesting side projects. If you accept that, then starting some kind of side project would be worth it.
FTA: > Side note: learning a skill isn’t a side project, you have to ship something.
It doesn't imply that. When writing about anything, explaining all the things that you're _not_ talking about is a great way to never get to the point.
Determining that the post is about side projects aimed at landing a job is a task left to the reader. If the title didn't get you there, the second sentence is here to help: "For the distinct purpose of getting a job in tech, side projects are either a great use of your time or a great waste of your time."
"Distinct" as in specific — "The following may not apply if you're objective is not specifically to getting a job in the tech industry."
If you'd like to talk about other great uses of side projects, I think that's a completely reasonable discussion to have. Simi-related threads in top-level comments are common: "I know it's a little different form what the article is talking about, but I'd like to talk about…"
Everything that I have, my companies, my capital, my house, comes from my side projects being transformed into self employment, recurring revenue or reputation for me or my team.
You have to consider that side projects will start small, like a plant, they are very easy to squash and destroy, but takes an enormous effort, and your rewards come later, when you have a tree.
Even a company like Apple could not make something like the App Store earn more than a couple millions the first year. Larry and Sergei tried to sell google for just a million dollars and could not.
A side project takes so much effort that I would have never ever considered making the effort just to show off to companies, then abandoning those(as you will be forced as you have a job that takes most of your efforts). Side project is like your children if you create them.
It is a big error to try to make companies value your projects. They will pay you 0$ if they can(even if they know it is a great project, more benefit for them).
There are much better ways to show off in front of companies if that is what you want.
I also have another set of side projects with specific intent to learn something new, which I consider an educational time investment in myself.
Neither is "click bait" for employers, and I consider neither to be a "waste of time".
Of course, balance is important and I won't try to fool anyone -- it's something I'll likely always be working to improve.
No, I'm not planning on raising money for this side project.
No, I don't have a million users.
No, I don't know why people would want this.
No, it's not for a startup.
No, it's not related to my actual startup at all.
It's called having fun, damn it.
Anyway, I had a couple of off by errors. The interviewer asked me if I was sure I didn't have any, so I had a look through, stepped through the code, and lo and behold, I did. While I was doing this process, we ended up having a discussion about how unit testing would help avoid problems like off by 1 errors, and would make iterating through the code and refactoring a lot easier.
That's what a good tech interview should look like, not expecting someone to write perfect code, first time.
On a side note, I didn't end up getting the job, since they had apparently hired someone the day before I came in for the interview. Apparently I managed to apply for the job the day they closed applications, but they asked if I wanted to interview anyway, out of courtesy.
[1] https://www.reddit.com/r/taoism/comments/38h66v/liezi_book_8...
i did one "field of dreams" project using a stack that i was mostly familiar with already and several "effort" project to learn and demonstrate skills in a different language/stack (Clojure).
totally agree that once the interview screens started, that side-project time got scaled back to vanishing. however, i'm really thankful that i had enough time to do the side project milestone as well as the interview proper, because i wound up finding a role developing in Clojure. i don't think that would've been possible had i not invested the time in side projects before starting in on interviews.
It is like buying stamina when the next boss can be easily defeated using strength.
So, ultimately that process is going to be whiteboard coding, because it (let’s be honest) basically always is. That whiteboard coding tends to draw its questions, most often, from question pools like those on Leetcode.
This means if you’re resume’s strong enough for someone to talk to you, you probably should be optimizing to pass the whiteboard section.
leetcode is intended to help that type of interviewee get up to speed on a lot of those sort of technical interview questions that they may not have directly thought about for years.
I work on quite a few side projects in my own time because I enjoy it, but I've never been asked about my projects on Github or my blog.
When I hire, code samples (like Github contributions) are my first screen of a candidate before I decide to interview them. It is proof of ability to write code and a testament to what I consider to be quality. I do this in lieu of whiteboard challenges or Fizzbuzz tests.
I know the prevailing sentiment on HN is sometimes that hiring managers like making candidates jump through hoops to make themselves feel important, but I don't think that's the issue here. It's simply that someone who came to programming late in life is at a severe disadvantage compared to people who have been interested in it since fifth grade, and some kind of programming in your spare time is the best (only?) way to catch up.
Contra the article, I don't care if the project shipped, or if it's thoroughly unit tested, or how many iterations or whatever. I just want evidence that you're in the larval stage.
https://en.wikipedia.org/wiki/Object-oriented_programming
And I've never gone to a job interview where people knew I had a blog, even though I list the blog on my resume:
I have to assume that other people put a blog on their resume, but its a mostly un-used site, so the maybe the employers get in the habit of ignoring personal sites? I've written on hundreds of technical topics over the years, but no one is ever aware of this when I arrive in a job interview.
Likewise, no one has ever checked my projects on Github.
But where I have public-facing projects' URLs on my resume - interviewers seem to be completely clueless about them. They'll ask me questions about a topic, and I'll point them to that URL on my resume as an example of how I know that area, or did that task.
It seems like they only read the "skills overview" blurb at the top, that we put there to get through the search filters - and they don't read the work history.
Source: https://thehftguy.com/2016/10/24/heres-how-to-make-a-good-gi...
The thing that companies don’t generally care for are non-impressive GitHub profiles. I don’t care about seeing code that doesn’t stick out as not from a craftsman, or just docs contributions when I review resumes/GitHub profiles. I want to see code demonstrating standard industry practices, and smart architecture decisions. I want to see an understanding of testing.
It’s up to the candidate to make good use of GitHub to impress. Most fail at it, but that does not mean that no one looks at GitHub profiles - on the contrary, they usually are quite revealing.
Your experience suggests that one should contribute to the major open source projects used by the big companies to get noticed by them. That's very different from the usual advice to write side projects.
My experience is that side projects mean nothing to employers (as an experienced dev) and memorizing algorithm trivia is rarely needed. Usually doing a coding test with internet access is the norm.
Because we don't have professional organisations like doctors and lawyers do.
> What's the value of my degree if it doesn't tell employers that I know what I'm talking about?
Well there's an open ended question. For one, assuming you went to a reputable university, the degree says that you understand the foundations of computer science, which isn't particularly relevant to what most of us spend our days doing. Understanding computer science fundamentals isn't a huge help when your building a CRUD web form and sticking data in the database.
Then there's the profit driven centers that universities have become, it's in their best interest not to fail anyone, that would decrease their revenue. I've had interns that are freshly graduated with masters degrees that would fail a CS101 test if you put it in front of them.
> Why do I have to build side projects and memorize algorithm trivia to prove to employers that I know enough to do a job?
Because employers are trying to compensate for the issues I mentioned in the cheapest, quickest way possible. Sure they don't do a great job of it, but neither does anyone else in the pipeline.
Comparing yourself to doctors, with the implication that it's easier for them, is pretty silly. Doctors have to memorize perhaps the most shit to get a job, and are frequently reevaluated on it through examinations to keep their license.
Not in the US. Keeping your license, whether for medicine or engineering, is just a matter completing continuing education units. There are nearly worthless activities that count towards CE units, and I have yet to see any that are actually graded (they are all just butt-in-seat things).
People here seem to have anloftier view of licensure than it deserves. Sure, it's better than nothing. However, licensed != competent and should never be used as a substitute for measuring competence.
For juniors it could be a positive signal though, if you're the kind of person that always looking to learn new things. Depends on the situation if you should show it or not: who are you trying to come across as?
If you have the resume, don't show it. Otherwise, do show it and let me know what parts you like and why some other parts are WIP (comments are great here).
Another thing to take in consideration is your own criticism. You might be really critical of the project since you're an exceptional programmer, and great programmers tend to move forward quickly and look back at their own old code with distaste. :)
- Programmers love good code
- Entrepreneurs love good products
So it's not always the quality of the codebase that matters when showing your side projects. Sometimes the product itself is more valuable.
Another recommendation I have is to pick side-projects that are small enough in scope to get finished. There's more satisfaction in that and also more useful for interview situations, arguably.
If I see commented code, rough violation of coding standards etc., that is bad and I will ask about it.
So if you share side projects, remember that it gives the interviewer "attack surface" to disqualify you. However, good interviewers won't look 20% for your weaknesses and 80% for your strengths.
If you are a regular open source contributor, I might use this information to argue that you are be better than other engineers who don't regularly code in their free time.
Obviously, if I really take this as a main parameter, I will be hiring probably 22 yr old kids with no relationships. I only see an indirect relationship to diversity. Still, good hint to keep that in mind.
1. Feross never got the job he was "offered" and reportedly "accepted"[1]. He went on to do other things.
2. He is on Patreon, and you can support his Open Source efforts here: https://www.patreon.com/feross
3. Or... you can try and hire him. He was excellent to begin with and seemes to have only gotten better through the years.
4. 2010. That was an era. "Must have been love, but it's over now."[2]
[1] https://techcrunch.com/2010/09/24/youtube-instant-instant/ [2] https://youtu.be/s-XolL_1dN0
My take is that this is targeted toward people without any sort of traditional/university CS background trying to get into tech from scratch and in addition to taking a basic javascript course or whatever are following the standard advice you hear of "you need to have some side projects to get employers attention." Think its just trying to give some pointers to people going down that path with the very specific larger goal of it leading to a full time job.
So ya, I would bet the author would probably agree that side projects are almost valuable in and of themselves, but some are going to be more likely then others to lead towards a full time position, etc.
Semantic Image Search: takes high dimensional images vectors from a Deep Convolutional Neural Network and searches for similar images using Locality Sensitivity Hashing.
Search is O(1) + O(n/B) where the last term is a small fraction, B, of the complete search space n.
The goal was an E2E machine learning project: modeling, training, validation, deployment and serving. Using VGG-mased CNN w/ tensorflow, redis for database, Flask for web frontend , tensorflow serving and docker+kubernetes for orchestration.
It all works on my local machine but when I went to deploy it to my prod website the tensorflow serving part broke down and I haven't had time to fix it yet :(
It was just supposed to be a one weekend project after I got the inspiration from the AlexNet paper, but it took over a month. Most of the time was fighting k8s.
So another data science project doesn't make it to production...
> Make sure your side project isn’t a waste of your time
A side project is never a waste of your time. It's a way to build something new, to exercise your creativity and passion, and to (when it's all said and done) ship something! I'd wager that 90%+ of "coders" have never built something from zero to hero. That is inherently a valuable thing, regardless of what your future boss thinks.
> Your side project, by itself, probably won’t get you a job
Okay, so? I'm pretty sure no sane person is going to go through the hard work of building a side project to impress during an interview. Not only that, but for a side-project to be successful, it also needs marketing, PR, networking, and so on. You probably won't have to do any of those things at your next programming job, but you most definitely need to do it when launching a side project.
> 1. Is your side project relevant to the job you’re applying to?
Again, who cares? Side projects should never be contingent on what jobs you're applying to.
> 2. Does your side project have depth?
Doesn't matter. A side project can be as shallow as Snapchat 1.0 (disappearing images? Sign me up). Run with it. Build something awesome and win at life.
> 3. Has your side project at least had 2 iterations?
I actually agree with this. Most of my MVPs are kind of shitty until I show them to a select group of friends and family that more often than not rip them to shreds. Fail fast, fail often.
> You have to bring attention to your side project
Yes, yes you do. But guess what: looking for a regular 9-5 programming job is completely antithetical to "bringing attention" to your side project. The latter of which involves emailing zillions of people, setting up mailing lists, getting featured on blogs, product design, going to talks, giving talks, and so on.
That can be your rock to hold onto as burnout approaches...
> I'd wager that 90%+ of "coders" have never built something from zero to hero
I think that's probably very true! Shipping something is such a useful exercise. You learn so much by going through the full process, rather than just starting and then later abandoning something.