Hacker News Reader

Except that the vast majority of programmers essentially say "I'm very good at taking pictures, but I can't actually show you any of them because they're all property of my previous employers, and I never take pictures just for myself".

> I think programmers should create their portfolios of personal or paid for projects that showcase their technical and artistic skill and companies should take those seriously.

100% agree, and Github in particular is great for this. Sadly, in my experience very, very few candidates have such a thing. Even non-code content such as issues they've filed or stackoverflow questions and answers is a very useful way to judge someone's technical abilities, but many candidates don't provide links and either use separate identities/aliases or don't participate.

I have all that stuff

No problem getting interviews

Still have to do the "brainteaser of the hour" technical interviews. Fortunately it is often the same question I realize, so this gets easier without needing to go through the whole book on this.

I imagine the vast majority of employers would be pretty pissed if the vast majority of programmers shared their private code, and some might even wish to take legal action. As an employer interviewing and hiring candidates, one should respect and expect this confidentiality to be exhibited and adhered to.

For example, I build things for clients that are proprietary to their businesses. Some things are publicly accessible, like marketing pages, but the real stuff is locked behind employee-only auth. Other things run entirely on internal networks, and can't be accessed from the web. They would be rightly upset if they discovered I shared their code with a third party, or placed it on Github.

Moreover, just because Github exists, doesn't mean programmers should feel obligated to "participate". Outside of developers who maintain or participate in truly used projects, it's mostly noise. The only exception I can agree on is looking at issues a candidate has reported to gauge their depth of understanding, clarity of communication, offering of solutions, and other such valuable insights.

Many companies don't bother looking at things like that. Even if you have a Github account with some amount of non-trivial, original code on it, they'll put you through a "phone screen" and have you do a longest increasing subsequence/subset sum/longest palindrome substring problem for a front end position. And yeah I'm talking about Silicon Valley/SF City/YC type startups here, not BigCorp Inc.

100% agree, and Github in particular is great for this

No, it's useless for this, because companies don't look at it. That would require the recruiters and interviewers to spend more than fifteen seconds prepping for the phone-screen call, and we can't have that.

Incidentally, after my last job hunt one of my new policies was that I hang up on any company that throws me into their fizzbuzz phone-screen after being the one to initiate contact with me. If they didn't think I had the skills they shouldn't have reached out in the first place, and if they do think I have the skills they shouldn't need to waste my time and an interviewer's time on the fizzbuzz question.

(I say this after multiple companies tried to recruit me with a "we want a Django committer, that's why we're contacting you" pitch and then put me into their fizzbuzz-esque phone screens)

I agree completely, and this resonates with me particularly because I'm one of the many programmers also serious about other creative endeavors (in my case, the visual arts).

But it's worth considering that in larger companies, management is often terrified of the creative aspect of programming, because it messes with their primary goal of predictability. Thus they seek, many of them openly and explicitly, to make all technical roles fungible.

And as much as that's hurtful to us creative programmer types, it's not an irrational position for them to take.

The fact (or hope) that fungible programming roles can be more easily outsourced, especially to cheap-labor countries, is of course seen as a bonus.

This situation presumably gives a big advantage to any organization willing to engage with the creative aspects of programming, but I am starting to think it's in the nature of organizatins themselves to stop embracing that risk after a certain size is reached.

(Not that this is an argument against the portfolio idea. Even if you're going to work as a fungible cog, you'll probably get a better cog-job faster with a decent portfolio.)

There's definitely a place, probably the majority, for programming jobs that are relatively 'turn key'.

There's probably a decent analogy in product photography. Unless you're going high end, bashing out well lit, well edited and well presented product imagery lies firmly towards the technical end of the photography technical - art spectrum.

Trouble is, many companies think the want these 'product photography' programmers but in reality they need the equivalent of 'stock photography'. And stock photography is one of the most difficult and challenging technical/artistic balances in the industry.

Both turn out similarly neutral end products that can be used almost anywhere and replicated, but getting to those final products requires very different mindsets and approaches.

That's likely stretching the metaphor more than it can handle but hopefully illustrates the point!

And as you mentioned, having that 'personal portfolio' that allows you to sell yourself (another important distinction I made in another comment) will get you the job ahead of someone who doesn't have one.

> it's not an irrational position for them to take.

It's about as rational as decreeing that it must not rain on Fridays, and then proceeding to fire the nearest person every time it does.

They do have some very good and rational reasons why they wish programming was not a creative job. When they assume they can change reality by wishing it, they completely destroy their people's (outsourced or not) productivity, throwing the cost of everything into the stratosphere, and the quality way under the floor.

I think the key -- the point at which the position becomes (to me) rational -- is to knowingly accept lower quality, less innovation, etc. in order to achieve predictability.

It's a continuum, so of course nobody ever gets 100% predictability with software. But most of us, if we really needed to, could ratched down our productivity to a level where we almost always delivered exactly what we said we would in each sprint or other schedule unit.

And I think that is the crux. Most people who write software don't want to live like that, but there are a whole lot of managers dreaming of a shop floor full of "individual contributors" who are not very individual, and plenty of CEO's who will happily take that over cat-herding their way to the next iPhone.

That should give a big advantage to any company capable of embracing a little chaos and giving a happy home to talented and creative programmers.

But when I look around and see even, say, Facebook -- a company that needs to innovate and seems to know that -- building developer-hostile open-plan offices, it makes me think we may all be doomed.

There are also different types of programmers. Some will do well with portfolio like web developers. But how you would approach creating portfolio as embedded dev? When you work on proprietary stuff and github repo does not really show that you do well with hardware, because customer might not have time/knowledge to run this stuff to check if it is actually working.

A portfolio isn't something you just have and it does all the work for you. It's a living and evolving thing that is there to show that you can solve problems.

A portfolio says "I encountered X problem and here's what I did about it to make this other thing and therefore solved X problem."

A web dev is a very easy job to create a portfolio for because the product can be your portfolio but when I say 'portfolio' I'm not talking about a GitHub repo of some code you've written, I'm talking about something bigger than that.

I'm talking about having examples of your code, things you've made, perhaps technical writing, perhaps a personal project, perhaps video or other 'evidence' of the programmers work/approach. If that involves making a personal embedded hardware project to show your approach then that's what's needed.

Perhaps another important attitude change needs to be established...

In a creative technical art you are not really selling your 'product' to customers. You are selling yourself. You're selling your ability to solve problems and it's incumbent on the individual to prove their ability. Web devs aren't 'selling' their previously created websites, they're 'selling' their ability to create those past projects.

It's a crucial distinction and programmers would do well to adopt this distinction - they're not 'selling' their GitHub repo for instance, they're using that as an example of how they (their mind?) work to solve problems.

For sure there will be times you work on proprietary/secret stuff or on systems of such scale you can't possibly make an equivalent scale of personal project.

But you can demonstrate your ability and sell yourself, and that's really the key to it all. You have to get to the point in a creative career where you realise you aren't just selling your final product, you're selling yourself, your approach and your way of solving a problem.

Finding a way to do that is how you get the job ahead of someone else.

Could you make a Photography portfolio without any photographs?

You're allowed to include sketches, because you have time to make those for personal projects. But no photographs because they're all proprietary.

This analogy makes no sense.

I often encountered situations where clients wanted to keep photos 'private'. I couldn't use them in my portfolio. To deal with that I took my camera and lenses (analogous to a computer) and made some equivalents that allowed me to demonstrate my skill in a similar way (analogous to writing some code in a personal project).

Suggesting you should sketch something to show your photo making ability is like writing pseudo code to say you can write python - it's an unrealistic analogy that doesn't happen and has no need to happen.

You can very easily use a camera or a computer for a personal project and make very real photographs or write very real code to demonstrate your ability.

Think of "the photographs" in the analogy as the code that a full-time developer with a wife, kids, and non-coding hobbies has produced for his/her employer over the last five years. This person may produce a script or a basic website for a portfolio, but those 40+ hours a week spent coding at a full time job is where all of the real problems get solved.

> This person may produce a script or a basic website for a portfolio, but those 40+ hours a week spent coding at a full time job is where all of the real problems get solved.

That's it right there. That's the bit that makes the difference. It speaks to an attitude of 'I'm doing it, I don't need to prove it'. It's wrapped up in 'I'm doing it and I'm too busy to prove it because I should have a life and other people should just recognise my ability so I don't have to sacrifice my lifestyle to put the effort in to get the job I want"...but it really comes down to thinking I shouldn't have to show it because I'm doing it.

There's no can't show it inherent in your original statement. There's a large dollop of don't want but no can't.

You can find ways to prove you ability, to use work you do day to day, to make high quality personal projects or contribute to those that are already running (beauty of programming is it can be easier to collaborate like this and still show your skill).

It can be done. Plenty of people in all industries work one job and study, prepare or work towards another all whilst supporting families and trying to have a life.

Programmers aren't special snowflakes, they've got to work just as hard at selling themselves as all other people in the creative industries do. That's why I think programmers really need to grasp that they actually are creatives and adopt an attitude that deals with all the burdens that moniker brings with it.

Can you have a decent, up to date, programming portfolio without investing tons of time in it?

Photography is IMO a flawed comparison since it's much easier and less time consuming to have some nice looking photos than it is to set up an equivalent software project. More than that, the photos you took in 2010 are likely to still be relevant in 2016, for the average user. However, for programming, your PHP4 project from 2010 might not cut it in 2016 when you're supposed to join a Scala project.

> Can you have a decent, up to date, programming portfolio without investing tons of time in it?

Why wouldn't you invest time in it? It's investing in yourself and could get you your next pay rise. It's not wasted time.

We're back to the point I made above about putting the effort in.

> Photography is IMO a flawed comparison since it's much easier and less time consuming to have some nice looking photos

Sure, as I already said it's not a perfect comparison, it's there to illustrate a point.

And I reject your second point very strongly. Even something simple like a single headshot can take hours of planning and setup. Something like a documentary project might take months or even years to complete.

Something like a commercial shoot might take a team of 20 a week to complete and cost hundreds of thousands of dollars - how do you get to put that sort of work in your portfolio? Sometimes you work for free, sometimes you create a whole shoot yourself at great cost. You work for it essentially. That's how.

And a single photo might be no more the 'whole' than a single file might be the 'whole' programme for instance.

Anyway, the exact vagaries of the industry aren't important, it's the assertion that programming is a creative discipline that is important.

> More than that, the photos you took in 2010 are likely to still be relevant in 2016, for the average user. However, for programming, your PHP4 project from 2010 might not cut it in 2016 when you're supposed to join a Scala project.

Possibly, but a portfolio isn't trying to sell a product it's selling 'you'. Stuff that goes in the portfolio shows how you do what you do, gives a window on how you work, how you approach a problem, how you solve it.

Many photographers will create 'behind the scenes' shoots and videos to show how they work and show clients how they solve problems for instance.

These can be independent of the exact example and just because a programming language might change you can still use an older project to show your skill and approach to a particular type of problem.

Anyway, I've laboured the point long enough, feel free to grab the last word but I think I've outlined my point well enough to be understood.

> feel free to grab the last word

I hope you realize you're having a discussion with three different people. It's not up to you whether the community continues this discussion without you.

Of course I realise, the usernames are literally right there. I was just saying I'm out of the conversation as I've laboured the point too long. As you say, a post here is a post to the community, so I was addressing the community - saying someone else can take the last word if they wish, I've gone on too long.

>made some equivalents that allowed me to demonstrate my skill in a similar way (analogous to writing some code in a personal project).

Except that doesn't work in programming. I can't take the core bits and re-release them in my own projects. In fact that's literally what I'm prohibited from doing.

While in photography you can make the same photo with different people, in programming you absolutely cannot simply rewrite the same code.

Did you miss the word equivalent in the statement you quoted? That's kind of the key word.

OK. Imagine this...

You photograph an event at an exclusive venue that you won't get access to again, with a setup you can't replicate (ever), with famous guests you'll never see again and a world famous VIP as the main subject.

Oh and to boot, the client wants total confidentiality. No photo use for portfolios whatsoever.

Seems just like the restrictions a programmer might face right? An infrastructure you can't replicate, code you can't publish, resources you won't have access too etc etc...

This was some years ago now and I've left photography behind but how did I ever get to do that job in the first place? Created stuff myself, off my own back, unpaid and at great time expense that showed I had the level of skill and ability to do it. I chased other photographers, contributed time and talent to their projects and made stuff myself.

And when the clients came looking I could show from my prior work and portfolio that I could do what they needed.

And what did I do afterwards? Took inspiration from what I saw at this event and made additions to my portfolio that showed I could solve the kind of challenges I faced there, proving to subsequent customers that new skill. I didn't just say 'oh dear, I can't use this so that was a total loss, let's not do anything about it.' I went out and made new stuff that showed equivalent skill, not recreating the photo, proving the skill.

Maybe I'm not being clear enough or maybe there's a misunderstanding of what type/level of photography I'm referring to so I'll state it plainly:

Photography was just an analogy. The key is you're not taking the original work that you did if it's restricted, the key is creating something that shows equivalent skill. Remember, I already said that as a creative you're selling yourself, not the end product, so you're not trying to recreate the particulars of the exact photo/code but demonstrate you have the skill/ability to create something of equivalent difficulty.

Need to prove you can run a big project? Well, run one. Do one yourself or join another. Find a way. Don't throw up a barrier every time something seems difficult or hard. It's meant to be hard. It might take a long time. Deal with it and get it done. That's what I'm saying.

People will tell you you're getting taken advantage of and doing too much for too little right up until you're working for some global mega corps running some incredible team, earning massive bucks and all that stuff. They'll keep telling you it until you believe it. Or you can ignore them and just keep going, keep making something to show what you can do and keep chasing the jobs you want with that big backing of proof of your ability that you've got.

I think http://rossumblog.com/ is a fairly impressive portfolio of embedded development.

I've only done very small embedded projects, like http://canonical.org/kragen/light_sensing/ and https://github.com/kragen/ar_bytebeat.

That is actually a really great point.

Its like hiring a writer based on their grammar. A writer could have perfect grammar but their books are boring as hell.

Or even hiring a writer for adhering to the techniques of connoisseurs. The book "Ready Player One" was heavily criticized as amateur work (it was, after all, the author's very first book), but it nonetheless was hugely successful (and even has a movie in the making).

Would you hire a writer with bad grammar?

If they had a killer hook, yes. An editor can always review and fix the grammar. :)

A coding analogy: A compiler or debugger can catch bugs.

If they can't form a competent sentence, they probably would fail to write a convincing hook. But some margin of error is acceptable.

Is the "technical" part really analogous to grammar? I don't think many interviewers care about that. They don't care about missed semi-colons. They want to see that you know how to solve a problem. And a github portfolio is not going to give them that.

As an aside, I think debugging is hard even with a debugger.

I disagree. A github portfolio combined with time spent via screenshare reviewing and discussing some of that work is far superior to a technical quiz IMHO.

Thats true, probably a poor analogy, but the overall point is that:

Determining a programmer's worth probably takes at least a month in my experience. Maybe even two months. I almost got fired within the first month of my first job, but quickly ramped up and turned around to be the highest value-adding employee by virtue of completing the highest paying contracts at a fast pace (the company was later acquired).

Most companies can't spend a month, but a day long take-home test is a great alternative.

There are so many facets to a programmer's worth, and so many ways to miss them or ask the wrong questions, that I just don't believe in short, high-pressure interviews. Take home tests have been highly successful in my experience (and a few others, from what I have heard).

how would a GitHub portfolio of problem solving projects and contributions not show that? I can't think of any better way.

While true for writing books, I wouldn't consider code that doesn't pass a linter or compiler a useful part of one's portfolio. There is no editor equivalent who will fix your syntactic and semantic errors for you, and getting that right is an important part of your experience with a language/tool.

There is no editor equivalent

It depends on what you are working on. A non-trivial part of my job involves working with domain expert scientists and taking their very clever, but badly written, code that solves some hard problem and turning it into fast, efficient and usable code.

The point though, surely, is that having good grammar shows you can write not that you're a writer. The corollary being that passing a technical test means you can code not that you're a good developer.

I think it would be a step forward if it became standard practice to expect an online portfolio of a developer's work.

It is only a certain sphere of devs that have such a portfolio. All my work is done for money and I don't have the rights to show it to anyone else.

The unwritten hand-waved expectation is that you should develop side projects in your spare time.

I find that this is incredibly discriminating against people who have other hobbies and don't like to spend 80 hours a week coding with only 40-50 of those being paid.

Where the programmer/photographer analogy fails is that usually, the photographer retains the right to the vast majority of their creative body of work, whereas a developer might not have anything client-facing to show at all, even if they have side projects.

Perhaps http://www.huffingtonpost.com/2014/02/20/grammar-rules_n_476...

Depends what you're hiring for, grammar never troubled James Joyce.

As an example of a brilliant book with awful grammar, there's Heinlein's The Moon is a Harsh Mistress.

I would hire a writer who can communicate effectively. Knowing the differences between a gerund phrase and participle phrase is irrelevent imo most of the time.

I have a portfolio (it's a bit dated now, though, and only includes the games I've worked on). It's always clear in interviews that my interviewers never bothered to look at it, though.

Still get the pop quiz interviews almost every time (with or without coding) and it's pretty clear my chances of getting an offer is at least 90% based on how well I do on their pop quizzes.

Portfolios are a good heuristic, but you can't be sure the person actually wrote the code/took the pics, which is why you need some other way to verify they're actually capable of producing that work, not just copying it. Hence technical interviews.

Even for a programmer with a portfolio, it's vital that you be able to probe them with questions about design decisions and expose someone who didn't actually write it.

Does photography have some sort of solution to this problem? Can you ask a photographer about how they took a picture to easily reveal whether they actually took it?

I'm guessing there's less of a need to verify skill with photographers since most of the work is on a gig or piece-work basis where it's easy to terminate the relationship or refuse payment in case of obvious frauds. OTOH, what stops an unskilled from stealing a portfolio (or using with their consent) to get work? What happens if they produce usable work, but which doesn't look anything like the quality suggested like their portfolio?

A lot of photographers will now video behind the scenes stuff to show them at work and have that ready for clients if questions arise. Also, during meetings (interviews) with potential clients invariably I'd talk through and show a collection of work that was similar to the types of problem the client was trying to solve.

Might have been a wedding or a commercial shoot, I'd show them stuff that would solve what they needed and talk through with them.

Similar things have happened with editing.

I suppose an equivalent would be not just code but maybe wider community contributions. As I mentioned a portfolio goes beyond just 'code' or some 'product' that you made. It's there to sell 'you' so that includes things like bug fixing, visible issue solving, maybe mailing lists, maybe run a community platform, maybe clear contributions to open source projects or whatever. Anything you wnat I guess.

It's relatively easy to become 'public' enough to be obvious who you are, or use some kind of cryptographic proof etc.

There will be problems of due diligence, those are universal, but I think individuals and companies would both be better off if they took portfolios more seriously.

Thanks for the thorough answer to my concerns.

Even for a programmer with a portfolio, it's vital that you be able to probe them with questions about design decisions and expose someone who didn't actually write it.

Repeating what I said in another comment: during my last job hunt I had multiple companies which contacted me, unsolicited, asking me to apply to work with them, because they wanted an "extremely senior" Django person, preferably a committer (I got my commit bit in 2007). Every company which contacted me like that also put me into a fizzbuzz-esque phone screen. There is no sense in this; my work is publicly verifiable and vouched for (sometimes in an extremely literal sense, as it's my PGP key that provided the signature on a bunch of release packages of Django), so either they believe I am who I am, in which case no need for a fizzbuz, or they don't, in which case they shouldn't have been contacting me in the first place.

My policy now is to hang up on any company that reaches out to me for a skill they know I have, and then fizzbuzzes me in spite of that knowledge.

I both agree and disagree with this as many of the other comments took this in a very different way than I had.

You can still test for creative ability in the traditional coding interview construct through systems design questions. It requires you to have a deep understanding of various technologies, but at the same time focuses more on how you view various trade offs which are crucial in designing complex systems.

When discussing art with its creator, you could have the same kind of discussions about various tradeoffs they made in regards to various technical topics. For a writer, use of alteration in a certain way, or a photographer trying to draw focus to a certain aspect of the larger photograph, in particular to various editing decisions.

The issue here is that the vast majority of new-grads/young programmers are fresh out of school and haven't spent any time focusing on systems design so companies focus on what they know best, asking them to be creative in the domain of algorithms. I also think its the case that since so many people's first interviews were algorithms based the process just continues to be that way.

But, at the end of the day, to say choosing a particular data structure to solve a certain type of problem has no creative aspect to it is wrong.

Having been through interview processes at big tech companies recently, I invested some time in preparing a portfolio to try and improve my chances of success. What I encountered was that nobody is really interested in your portfolio (or even your resumé really). I appreciate your advice that a portfolio should improve your chances, but nobody from any of the companies looked at it. I'm not sure it's a good investment of time, considering how little importance is attributed to it.