I lowkey judge any developer who is noticable slow at typing as I can't imagine they're using a computer effectively at such a pace given how much keyboard hitting needs to occur during regular use alone.
Not that it's a high bar but I'm surprised more companies don't test wpm when hiring over rote crap like LC.
So the saying goes, but this article, and many people's honest experience, suggests otherwise.
Yes, developing complex systems is about thinking, not typing. But human head only fits so much, and once you pick up a tool to externalize your thinking, be it pen and paper, or text editor, or a whiteboard, you're limited by I/O. Reading is by far the fastest part, but if you're typing (or sketching) slower than you think, your entire thought process is now I/O bound.
There are many ways to improve this - with editors, you can use shortcuts, compose complex commands, use autocomplete, etc. - but the nice thing is, most of these improvements are purely additive. So it really doesn't hurt to learn to type faster than you think, and it'll definitely help some.
Instead of just disagreeing, I find I actually agree! At least with the premise of the I/O bound system.
At some point I realized that linear writing (typing or writing on lined paper) does not have the flexibility to express the thought processes that I want to commit to writing.
Now I write on blank paper (or an iPad if the thinking needs to be active for a long period of time).
I might be I/O bound but at least this way I don't need Serializable thoughts.
Maybe, but I often think fellow developers would get more mileage out of Mavis Beacon than the latest flavor of training from Pluralsight/AWS/etc
When effective comments, appropriate variable names, and the like can flow effortlessly from your fingertips, you're more likely to use them. Plus you're not expending your precious mental energy on the mechanics of typing out your ideas.
you're right. but the ability, given a good idea, to spit out a substantial amount of code that compiles and basically runs in a single session removes alot of the consternation and back and forth in those discussions
despite industry motion to the contrary, the truth is that we really can build anything we want. if we just weren't such cowards about it
Do you find that typos are the biggest barrier to quickly writing code that compiles and runs in a single session? I'm an absolutely _atrocious_ typer (in terms of both speed and accuracy), and yet I still genuinely don't think that actual typing takes within an order of magnitude of the time it takes me to get out a quick prototype compared to debugging.
Small note, when programming, there's other keys involved which allow for hierarchical navigation (ctrl, alt tab, up/down/left/right keys), which adds both some technical challenges by broadening the keyset and cognitive challenges (but of course code itself is highly cognitive.
I try to avoid the mouse, but I usually queue (mentally) a lot of keypresses at max speed, while the only bottleneck is the loading speed of the computer.
For example: alt tab (change window to browser), fn+F5 (refresh website , Ctrl Shift (change brower tab), fn+f5, ctrl shift tab (back to original tab), alt tab (back to editor or command line), etc...
Also, if I leave WASD for home row I quickly feel pain. Seriously, place the middle finger on W (alao used for S), ring on A, and indicator on D. It's so much more ergonomic it's insane.
> it'll just be 'one man startups' for better or worse.
Not necessarily. The reality is, whatever some people can do individually, if they team up, they can do more together. The teams and small startups will remain for now, and so will big companies.
I do imagine however that the internal structure will change. As the AI gets better and able to do more independently, people will shift from pair programming to more of a PM role (this is happening now), and this I imagine will quickly collapse further.
Even today, LLMs seem more suited for project management than doing actual coding - it's just the space in-between that's the problem. I.e. LLMs can code great in the small, and can break down work very well, but keeping the changes consistent and following the plan is where they still struggle. As that gap closes, I'm not really sure how the team composition would look like. But I don't doubt there'd still be teams.
....have you seen how much code and work vates has contributed upstream to xen? It's more than citrix at this point IIRC. Everything they do gets pushed back to upstream projects so I'm not sure what point you're trying to make
No, I don't follow legacy hypervisors but fair enough perhaps my initial impression was off-base... still you can appreciate the irony of complaining about Rocket Company getting free stuff :/
My guess is money laundering, given that the product is pretty vapourware-y (as a game dev in a past life: Vircadia looks more like 'how a web dev thinks multiplayer games work' aka basically unusable in a serious title).
I think you fundamentally misunderstand the stack. Their main usage of the platform has to do with a E2E solution with avatars, audio, etc. all synced without issue. These features ship with the client and other private repositories wrapping the core.
However, for usage to HN users they would be (likely) more interested in the SDK, the core, the underlying system, and how it can fit their use cases.
If you want to understand a small part of the scale of this project, you are welcome to check out:
I lowkey judge any developer who is noticable slow at typing as I can't imagine they're using a computer effectively at such a pace given how much keyboard hitting needs to occur during regular use alone.
Not that it's a high bar but I'm surprised more companies don't test wpm when hiring over rote crap like LC.
reply