Ask HN: How do you handle performance anxiety?
7 comments
"Quand je me regarde je me désole, quand je me compare je me console." Talleyrand
When I look at myself, I feel dismayed. When I compare myself, I find solace.
---
No one follows on every single trend. It's hard to not feel overwhelmed especially when everyone in those verticals try to create this fear of missing out because it drives wages up, attracts investors and pumps assets values.
The inflow of almost weekly breakthroughs / amazing demos also takes its toll on insiders. I remember seeing a post on HN from an AI researcher saying that those days everybody in their field was both enthusiastic and close to burnout. Because the rule to make any significant contribution is be first or be best.
Personally, I am trying to ignore it. 99% of what hits the headlines will be irrelevant within a couple of months, context and general understanding are easier to get and more relevant than knowing the very last things that was in the news. I am using the "news feed eradicator" extension to prevent this hectic stream to even reach me.
I also try from time to time to dedicate some hours to toy or to read more about some things that interest me and dive deeper.
Ultimately: pick you battle, you can't become on expert on ten things at the time. You will be bad or superficial at the rest. It's life, be ok with it, we all are.
You don't need to be on the forefront of a technological race to have a fruitful contribution to the world.
When I look at myself, I feel dismayed. When I compare myself, I find solace.
---
No one follows on every single trend. It's hard to not feel overwhelmed especially when everyone in those verticals try to create this fear of missing out because it drives wages up, attracts investors and pumps assets values.
The inflow of almost weekly breakthroughs / amazing demos also takes its toll on insiders. I remember seeing a post on HN from an AI researcher saying that those days everybody in their field was both enthusiastic and close to burnout. Because the rule to make any significant contribution is be first or be best.
Personally, I am trying to ignore it. 99% of what hits the headlines will be irrelevant within a couple of months, context and general understanding are easier to get and more relevant than knowing the very last things that was in the news. I am using the "news feed eradicator" extension to prevent this hectic stream to even reach me.
I also try from time to time to dedicate some hours to toy or to read more about some things that interest me and dive deeper.
Ultimately: pick you battle, you can't become on expert on ten things at the time. You will be bad or superficial at the rest. It's life, be ok with it, we all are.
You don't need to be on the forefront of a technological race to have a fruitful contribution to the world.
Well said, I think most of what you said hit right in the spot. I should maybe train on ignoring the hype and try to focus on the long run.
I am slowly realizing the fact that specializing in an area is the true answer, there is only so much information my brain can handle and I would prefer being an expert in an area rather than just being good in general things.
I might just list everything that interests me and allocate some hours per day or week to play around with it, that's actually a good idea! Thanks.
I am slowly realizing the fact that specializing in an area is the true answer, there is only so much information my brain can handle and I would prefer being an expert in an area rather than just being good in general things.
I might just list everything that interests me and allocate some hours per day or week to play around with it, that's actually a good idea! Thanks.
At any given time I have three side projects, I am usually making progress on two of them, the other one is asleep. Sometimes they change places, sometimes I start a new and drop an old one.
I'd say "learn a new programming language" is the worst of them all. Somehow people never really learn the basic result of computer science
https://en.wikipedia.org/wiki/Universal_Turing_machine
which is that they are all the same. "Create a new programming language" is a better way to learn. If you've got a project which has a large amount of discovery (don't understand the domain, don't understand product-market fit, ...) it's particularly deadly to be wasting your time looking up if it is len(s) or s.length or s.length() or whatever to get the length of a string, having fights with your package manager, etc.
Every time I learned a new programming language it was either because I needed it for a side project (e.g. learn shader programming to do a graphics project) or because there was paying work. (So often I have started a new job programming in language X and learned language X on the job)
I'd say "learn a new programming language" is the worst of them all. Somehow people never really learn the basic result of computer science
https://en.wikipedia.org/wiki/Universal_Turing_machine
which is that they are all the same. "Create a new programming language" is a better way to learn. If you've got a project which has a large amount of discovery (don't understand the domain, don't understand product-market fit, ...) it's particularly deadly to be wasting your time looking up if it is len(s) or s.length or s.length() or whatever to get the length of a string, having fights with your package manager, etc.
Every time I learned a new programming language it was either because I needed it for a side project (e.g. learn shader programming to do a graphics project) or because there was paying work. (So often I have started a new job programming in language X and learned language X on the job)
You're not wrong, but it's hard to ignore all the interesting posts regarding Zig, Go & Rust. They are really faschinating and attracts me a lot, but learning hundred different programming languages will not benefit me at all.
So I will note it down, just be really good at the one I already know. At the end of the day, most languages implements the same concepts but are designed different.
So I will note it down, just be really good at the one I already know. At the end of the day, most languages implements the same concepts but are designed different.
I wanna believe in Rust but I can’t.
I was excited about the predecessor Cyclone developed at Cornell but it was hard to code in.
I see people working very hard to do very simple things in Rust.
People forget that garbage collection was the key to software reuse in the large. Memory management is a global concern of a system and you can never really get it right when the library and the caller don’t really know in general what other code still wants to do or not do with the memory. With garbage collection you don’t need to coordinate memory management across all your dependencies. You pay a small price in performance but gain complete scalability.
Now, my aspirational fun programming language is AVR8 assembly but talk about “not scalable”. The obvious answer to “need a bigger board” is to go to ESP32 or ARM and some of those boards offer 5 volt convenience, WiFi and other great stuff. I guess you can run a soft AVR8 on an FPGA and make the FPGA do the hard stuff but it is not the easy way out.
I was excited about the predecessor Cyclone developed at Cornell but it was hard to code in.
I see people working very hard to do very simple things in Rust.
People forget that garbage collection was the key to software reuse in the large. Memory management is a global concern of a system and you can never really get it right when the library and the caller don’t really know in general what other code still wants to do or not do with the memory. With garbage collection you don’t need to coordinate memory management across all your dependencies. You pay a small price in performance but gain complete scalability.
Now, my aspirational fun programming language is AVR8 assembly but talk about “not scalable”. The obvious answer to “need a bigger board” is to go to ESP32 or ARM and some of those boards offer 5 volt convenience, WiFi and other great stuff. I guess you can run a soft AVR8 on an FPGA and make the FPGA do the hard stuff but it is not the easy way out.
You hit the nail regarding GC
Wear two rubbers. Take one off when it's showtime.
There is so many different topics I want to dig deeper into and be good at. It might benefit me in the future or in my next job.
For example, there are a lot of different programming languages, concepts & tools I want to explore.
We are in a period where machine learning, ai, blockchain (excluding the scams) and now LLM have gained a lot of traction.
It feels to me like I am missing out of so much.
What should I do? How should I approach this feeling? How does HN handle this?