There are, in the simplest terms, two hurdles you have to overcome to get hired.
1. Get an interview
2. Interview well enough to get an offer
There’s various approaches to either, and it sounds like (1) isn’t a problem?
For (2), if you interview and test as well as you say you do, it’s likely the case that you are recognized as somebody who would be a good fit, but they have more viable candidates than open positions, and you’re basically losing a coin flip.
If this is happening a lot, it might mean that you come across as unlikeable. Given two equally competent candidates, they are choosing the more likeable person.
Key metrics, for sadists so inclined, might be—
A) Number of applications
B) application-to-callback ratio,
C) Interview-to-offer ratio
You could make this more fine-grained, but that’s it.
I was part of a mass firing in April last year, and found work in August. I had a similar experience; I had enough on my resume to have several interviews per week, so I never put much effort into optimizing that.
I, in retrospect, was usually failing interviews that were open-ended, and the interviewer had a long list of questions to fit into a short time, and I believe we weren’t getting through all the questions. I was talking too much—I get excited about Software Development—and they were content to let me. Even if they really liked me, what were they supposed to do when they have half a blank form at the end of the call?
I was getting into late stages with many of the companies, definitely completing all stages before you would expect an offer. This happened a few times too. It happens. There are lots of good candidates, I guarantee that the reasons would seem very petty and capricious if you could see what was said when the hiring committee met. I’ve been on both sides. That’s life.
Two months is not a long time in a job search, unfortunately. Many companies you’ve applied to are barely organized enough to start screening candidates within 30 days.
As for myself, I eventually had a friend/colleague from my former company refer me for an interview at the company I ended up at. And after I got one offer I found myself with three the same week, after months of searching. And no income, pregnant wife, mortgage payments, etc. Fun times!
It’s nerve-wracking and frustrating to look for a job and constantly be interviewing. I often say that job hunting is the most humiliating thing someone can do. It’s exhausting and awful. It’s hard to be likeable when you do something that is exhausting and awful all day. I set some target for daily applications, and made sure I had free time for interviews, and was technically ready. I trusted my skills, so I actually spent far more time trying to stay sane and personable and energetic for the calls I was getting.
I think committing to the numbers game is sound advice. If 100 applications gets you ten interviews, and you need 10 interviews to get an offer, it’ll be somewhere between 0 and some hundreds of applications to get an offer.
The good news is that you only need one offer, and your search is done.
The bad news is that the only variable you can really affect without improvements to your applications or interviewing skills, is to increase application volume.
The worse news is that a lot of really terrible applicants and bots are flooding employers with not-good applications, which makes this whole thing more awful for everybody. Which increases the value of a referral.
System76 used to be rebranded Clevo/Sager laptops. I’m not sure if that’s still the case. You can buy them in various places, customize them, and have them put Linux (or nothing) in it. Or build your own. I paid to have mine built, but got out the screwdrivers to replace a hard drive, and it wasn’t too bad.
The build quality is worse than Mac, definitely, but you can get a really nice spec for cheap. And they’re serviceable, definitely on-par with the Lenovo and HP models that we’ve had in our house.
I bought a custom build from Xotic PC, using the “no operating system” option, and slapped Fedora on it. I’ve been Linux-only in my personal laptops since 2012, and windows-free much longer than that.
Hardware support is excellent, I’ve never had an issue. Everything just works. There’s a lot of FUD out there, and I’m skeptical of how accurate it is. My experience has been extremely low-touch, and extremely positive; it’s a tool, not a hobby, for me—I don’t tinker.
We write away the status part of our syncs. Just write it down, we don’t need to hear it.
One thing we used at the last two companies I worked at was what we call SWATH—something I invented for reasons of corporate judo.
It’s 5 simple true or false questions that gives you a score. Answer them in order, atop when you answer false. Your score is the number if true answers.
It takes most people 2 seconds to do, and is a kind of health check. They’re very basic questions.
Other than that, we have a shared agenda that people can add items they want to talk about to. Work related or not.
I’ve been through this a few times over the last many years. In one case, joining a team with 3–4 people who have been with a product for 10+ years. And the product was much, much older than that.
I would treat it the same as onboarding to a new job, which is essentially what you’ve done.
One question I would have is, Is being “the best” developer on the team the most impactful way that you can contribute? If you could snap your fingers, and have all the minutiae transplanted into your head today, by magic, what would that get you? My guess: not that much.
I know nothing of your situation, but it sounds a bit like you understand the absolute importance of being trusted and empowering people, but are perhaps uncomfortable with the complement—you have to trust your team, and be empowered by them as well.
Simple, technical things that you can probably contribute to right away, while learning the product and codebase are usually build, deployment, and testing. You could lump doc in there too. All are extremely important, all will save everyone on the team hours per week (or even per day!) and all are usually areas where there’s rot and neglect and you can have a big positive impact, without blocking real work or needing all the context for larger features or nasty heisenbugs.
You can/should absolutely still do code reviews. I don’t know how organizations you’ve been a part of work, but I would expect a new grad developer to start doing code reviews on basically their first day in the industry, even if it’s not super valuable immediately, or authoritative; this is no different.
Focusing on outcomes and asking questions, “how can we improve X by Y% in Z days?” is as valuable than being able to offer a solution for the same. If they know what they’re doing, let them!
As a general heuristic, if you have a team that is a lot better than you at something, you should find ways for them to be doing more of that, rather than trying to meet or exceed them in every area. Assign yourself the tasks that they are bad at, the impact on the team and org is going to be a lot easier and quicker, and people mostly like to not have to do stuff that they’re bad at. It’s win-win-win-win.
There’s a lot of nuance IRL, I’m happy to chat if you want to catch up and talk over a few ideas. The good news is that earnestly trying to do a good job is like 80% of the job, so it sounds like you’re on a decent path.
I have a small web project that I could use some help with that’s due before the end of the month. If you can do it mostly by yourself, I can offer you $1000CAD. I expect that it would take me about two days, someone else a little longer. If you want to help out, it’s all yours.
Not the GP, but I read "Thank You for Arguing: What Aristotle, Lincoln, and Homer Simpson Can Teach Us About the Art of Persuasion" by Jay Heinrichs this year, and consider it an excellent intro to classical rhetoric.
I moved from OSX to Linux-only laptop in 2012. I'd intended to dual-boot, but re-installing Windows[1], even by itself, was much, much more pain than I expected; I eventually gave up on it. Go figure.
Linux having poor compatibility with laptops is, in my experience, a myth. Ubuntu worked fine, but I've since moved to Fedora, and extremely happy with it. Everything "just works". I can't really see my self going back at this point.
For me at least, the Mac's original value-proposition being a no-hassle Linux-like machine. Now that the real thing is hassle free[2], there is no point in anything else.
Finally, I've heard that it's gotten much better lately, with the "new" Microsoft (which I applaud) but, as a developer who did so for a few months this year talking about "just works", the amount of friction involved in building anything outside the narrowest "MSFT-only" niches on Windows is crazy. I'm not a ragey guy, but it got to me, and not in a good way.
[1]: The then-new Windows 8 installer had a forever-grey "next" button at one point in the install-wizard that no amount of hair-pulling seemed to fix.
[2]: I would still recommend spending a few minutes online looking for red flags when choosing a model to use.
Joostena, Van Den Berga and Van Der Hoeven "Teaching functional programming to first-year students"
Chakravarty, Keller "The risks and benefits of teaching purely functional programming in first year"
Felleisen, Findler, Flatt and Krishnamurthi "The Structure and Interpretation of the Computer Science Curriculum"
You can get these all off Google Scholar. The papers listed do a good job, in my opinion, of discussing both the good and the bad of a functional approach.
When I went through school, we were taught with a myriad of languages, and when I was finishing up, they were transitioning to a more "practical" java/c++ monoculture. I think that's a bad deal for the students, personally.
Having been a tutor for a first year class teaching (a subset of) C++, I have anecdotal evidence against doing that. There are too many gotchas, and too much rainmaking to make it worthwhile, in my opinion.
That different artists have "signature" styles seems trivially obvious; it's interesting that nobody was able to analyze paintings rigorously enough to reliably apply this in the past.
Though I do worry about its accuracy. Picasso, for instance, painted in a number of vastly different styles over his career. It'd be interesting, and somewhat surprising, if his brushstroke signature was the same throughout his long career.
I applaud the effort, but this is far from complete. As far as I can tell, you can only do standalone expressions, to the exclusion of anything interesting.
Type Haskell expressions in here.
> let x = 10
not an expression: `let x = 10'
I understand that mueval has this limitation too? It's more defensible in the irc context that "state" would be disallowed. Here, it's just crippling.
I don't need the mtl stuff, but being able to define a basic function would be nice!
http://codepad.org/ will run full programs for you ... I guess that'll continue to be my tool of choice for online Haskell prototyping.
Note that within the first few minutes of that video, they show the brain already sliced up, and speak about it in the past tense; there's also a cut-and-died slice in a photo of the already-dissected brain on the first link...
I first watched that feature a few months ago, and from what I could gather, the episode was originally aired on 2009 August 25; anybody know what gives?
https://newsletter.posthog.com/p/the-hidden-danger-of-shippi...