HNHacker News
TopNewBestAskShowJobs

michaelrpeskin

883 karma · joined September 4, 2019

submissionscomments
michaelrpeskin··on All New Java Language Features Since Java 21
> I don’t think Java makes people into bad programmers, but I do think it selection-biases for intellectually unambitious engineers. They learn exactly enough Java in college to pass their courses, and then get a job at a BigCo that doesn’t strictly require ever learning anything more than what they were taught in their “intro to data structures” course.

I think that's a fair comment, but also there's this perspective: I first touched Java 1.1 in 1997 in college, and only for a semester. Then for the next 22 years never looked at a line of Java, working mostly in C++ and Python plus dabbling in FORTRAN for high performance stuff that needed to be written there. I generally consider my self not intellectually unambitious.

Then I moved to a Java shop who specifically needed high performance math (well at least as high performance as you can get in Java, which is actually pretty good now). But we stick to Java 8 compatibility because we have some BIG customers who require that because of LTS on Java 8. There are some nice constructs that would help make the code more readable and modern, but when you need to support people and actually make money you do what you need to.

michaelrpeskin··on Electricity prices are climbing more than twice as fast as inflation
Obamacare/ACA has this too. The "Medical Loss Ratio" or "80/20" rule says that 80% of premiums have to be paid as claims. There's no downward pressure on claims payments because they raise rates and take 20% of a bigger number.
michaelrpeskin··on Claude Code is all you need
I attribute my barriers to entry as things that forced me to really learn. All my family could afford was a 386 with 16MB of ram when 486s where pretty common. I had to really hack to make things work. Working under constraints meant I was exploring limits and understanding efficiency.

I still carry that in my day job which I _think_ helps me write better code - even in unconstrained systems, thinking in terms of memory and instruction efficiency can still help make better/faster code.

michaelrpeskin··on Ask HN: What trick of the trade took you too long to learn?
yes - I often think about what I would do if I "won the lottery" or similarly got a significant amount of cash. Even though I have a low mortgage rate (refied in the middle of COVID), I'd probably still pay off my house. I know it makes absolutely no rational economic sense, but not having a mortgage payment and knowing that I own my house would give me a piece of mind that as you say is hard to put a dollar amount on.

I know there are people out there that like maximize returns with leverage and all of that - and the math all makes sense! - but I'd prefer the simplicity of not dealing with it.

michaelrpeskin··on Read your code
You haven't been in the industry long enough :)
michaelrpeskin··on 6 weeks of Claude Code
Yes, this!

A couple of week ago, I had a little down time and thought about a new algorithm I wanted to implement. In my head it seemed simple enough that 1) I thought the solution was already known, and 2) it would be fairly easy to write. So I asked Claude to "write me a python function that does Foo". I spent a whole morning going back and forth getting crap and nothing at all like what I wanted.

I don't know what inspired me, but I just started to pretend that I was talking to one one of my junior engineers. I first asked for a much simpler function that was on the way to what I wanted (well, technically, it was the mathematical inverse of what I wanted), then I asked it to modify it to add one different transform, and then another, and then another. And then finally, once the function was doing what I wanted, I asked it to write me the inverse function. And it got it right.

What was cool about it, is that it turned out to be more complex linear algebra and edge cases than I originally thought, and it would have been weeks for me to figure all of that out. But using it as a research tool and junior engineer in one was the key.

I think if we go down the "vibe coding" route, we will end up with hoards of juniors who don't understand anything and the stuff they produce with AI will be garbage and brittle. But using AI as a tool is starting to feel more compelling to me.

michaelrpeskin··on If you're remote, ramble
Yeah, I don't like it in "channels" either because it's too random. What I started years ago at a previous company who was using Confluence (yuck, I know), as the knowledge base was to have a "Personal Space" where I created internal blog posts. I still do it at my new place using their tools. What I do is more "bloggy" and than "rambly".

When I'm considering a refactor, I write up my thoughts in English (which often helps clarify things rather than focusing in code at first). And then I point the rest of my team, to the post and say something like, "I'm going to tackle this next Wednesday, let me know if you see anything wrong with the approach". People who care and have options can chime in, if they're too busy they can ignore it. But everyone is given the chance to comment.

But where I find the real value is when I'm working on a new algorithm or analysis approach. Our internal blog software natively supports LaTeX math blocks (like GFM), so I can write out my algorithm ideas using formal math notation. I've pre-found a bug or bad idea many times just by translating my English into LaTeX. I actually find the expression of those ideas in a blog post the key tool to solidify ideas before I code them.

I'm under no illusions that most of the team even reads what I write, but the work of formalizing it for semi-public consumption really clarifies my thoughts and keeps me from spinning too much while I'm actually writing the code.

These aren't super formal academic quality publications, more like semi-formal ramblings, but I think the difference between hitting "publish" vs just typing in a channel slows me down enough to really think through things - and those who do end up reading them are reading slightly more thought out idea than a stream-of-consciousness rambling which means they'll get more out of it too.

michaelrpeskin··on The Who Cares Era
This parallels an article that someone I follow wrote a couple of weeks ago. His way of describing this is the "second world effect". The article is better than I could write (link below). But basically, the "third world" is a low-trust society and everyone understands that and behaves in a defensive way. The "first world" is a high-trust society where things work. There is a discontinuous jump from third to first world once the culture has enough high-trust built in. But if the first world devolves back to low-trust, it doesn't go back to third world, it transitions to "second world" where things look like the first world, but nothing works any more because it's low-trust, but the society hasn't really recognized that.

https://boriquagato.substack.com/p/the-second-world

michaelrpeskin··on Ask HN: Anyone struggling to get value out of coding LLMs?
A little snarky but: In my experience, the folks who are 100x more productive are multiplying 100 times a small number.

I've found great success with LLMs in the research phase of coding. Last week I needed to write some domain-specific linear algebra and because of some other restrictions, I couldn't just pull in LAPACK. So I had to hand code the work (yes, I know you shouldn't hand code this kind of stuff, but it was a small slice and the domain didn't require the fully-optimized LAPACK stuff). I used an LLM to do the research part that I normally would have had to resort to a couple of math texts to fully understand. So in that case it did make me 100x more effective because it found what I needed and summarized it so that I could convert it to code really quickly.

For the fun of it, I did ask the LLM to generate the code for me too, and it made very subtle mistakes that wouldn't have been obvious unless you were already an expert in the field. I could see how a junior engineer would have been impressed by it and probably just check it in and go on.

I'm still a firm believer in understanding every bit of code you check in, so even if LLMs get really good, the "code writing" part of my work probably won't ever get faster. But for figuring out what code to write - I think LLMs will make people much faster. The research and summarize part is amazing.

The real value in the world is synthesis and novel ideas. And maybe I'm a luddite, but I still think that takes human creativity. LLMs will be a critical support structure, but I'm not sold on them actually writing high-value code.

michaelrpeskin··on ASK HN: How to engineer a JavaScript to Python migration?
Was it Larry Wall that said "it's easier to port a shell than a shell script"?

I've done similar inter-language changes, and I have always found it easier to not change the language of the business logic. (Unless the change gives you something really big - for me I often port stuff to numpy because I need the vectorized code, but that's only for very specific problems).

If I had this task, the place where my brain would go would be to find a way to compile JS into C and then use C calling conventions to call the functions from Python. Keep the JS code around so that if you need to change anything, you change it in JS and then recompile to C.

I don't know the JS space very well, but can you get a JS interpreter that lives in Python? That way you can call JS functions from Python?

I don't like transpiling, there's always enough differences between the languages that something bad happens. When I've run into issues like this, since I'm an "old guy", I tend to try to get everything into C calling conventions and use that as my base interface.

Worst case, there has to be some good JS interpreter that can give you a C interface that you could call from Python. So you'd have Python -> C -> JS and your business logic can still live in JS (if your port is because of efficiency and you need compiled code, then you can ignore me.)

michaelrpeskin··on Show HN: I analyzed 1500+ job ads to find the most wanted skills by recruiters
Yes! I came here to say this.

It's like saying my skills in "English" are equivalent to Maya Angelou's skills in "English"

Skills are things can do regardless of the technology/language. My skills are in computational math. I could write you a good algorithm in JavaScript with maybe a day or two of ramp-up/review of the nuances of the language. I could not (quickly) write a (good) web application in the same language.

I can deploy complex distributed computational code using EC2, I cannot even begin to understand the IAM. Do I have "Skills" in AWS?

It's such a pain when I'm reviewing resumes. I don't care about the language, I care about what you can do, and that's so hard with the checkbox system we've devolved into.

Much of my career has been based around low-level C and bit-twiddling optimizations for micro optimizations of math for speed and floating point accuracy. My current job is at a Java shop. When interviewing for that work, I hadn't touched Java since CS101 in college. I was lucky that I was interviewing at a small company that didn't have recruiters, but had me talk directly to the team and they could understand that the skill was "Math" not "Java". I still can't understand how Gradle or the rest of the Java build system we uses works, so it's good we have someone with "Skills" in DevOps to do that.

I hope the industry recognizes this soon - I think you're right, it's probably the biggest impediment to finding good candidates.

michaelrpeskin··on Fewer students are enrolling in doctoral degrees
Regarding Opportunity Cost...

For a long time, I felt stupid for getting my PhD during the buildup before the 08 bubble. I could have socked away a lot more money than my measly stipend. And afterwards, I always had decent jobs but not SV style salaries. That made me feel like it was all a bad decision.

But now that I'm approaching my 50s, I feel a bit differently. I traded variability for steady consistent growth. When SV lays of 80% of the work force and a bunch of people lose their jobs and their fancy SV salaries go to 0, I've luckily (knock on wood) never had that experience.

I bet in the long run a person making SV salary right out of college and invests smartly will still out perform economically than a steady growth after a delay for the PhD. But mentally the lack of variability has been good for me. YMMV.

michaelrpeskin··on Ask HN: Trying to find a post about some OS developer in the 80s coding by hand
I have a old guy that I work with - PhD in Math (because CS didn't exist then) - who does lots of algorithm development with me. I often get Word docs of pseudo code from him. I'll do a search-and-replace on things like "LET" and "ELSE IF" and a very high percentage of the time if I run it in Python it works on the first try. Kind of amazing to me.

The hard part for me is then translating his ideas into vectorized numpy for speed, but at least I get the right answer to check against.

michaelrpeskin··on Ask HN: How can I grow as an engineer without good seniors to learn from?
I've been in similar positions throughout most of my career (early promotion above my current skill level with no mentor). If I look back, I think only my first internship ever did I have a good mentor to learn from. So I think I know how you feel.

There's lots of good advice in this thread, but I think the most important thing to keep in mind is to know that you will make mistakes and do stupid stuff. Always be evaluating yourself and use that to adjust how you do the next thing.

Anecdote: in my first real job after school, I was essentially give a one person R&D project that reported directly to the CTO to design a framework for "the future of our data analytics". I ended up designing this crazy over-engineered highly abstracted OO framework that was a nightmare to document and use. But since it was fast and met the requirements, the CTO pushed it through as the framework for the rest of the engineering team.

I spent the next couple of years justifying it and supporting it. Once I recognized that I made many dumb (well maybe not dumb, but novice) decisions. I used that as a learning experience for my next project.

I've had many of those kinds of learning experiences over the years, and as long as I keep self-examining and being self-critical, I learn from them and don't make the same mistake twice. Now 25-ish years later, I have a long list of internalized knowledge of what not to do.

I guess that's a long way of saying, don't be afraid to make mistakes, you will. Just pay attention to them and learn from them.

michaelrpeskin··on The EdTech Revolution Has Failed
Yep, we found the google-youtube url issue too.

Right now, I have youtube filtered by MAC on my Roku. I would like to add granularity so that they can watch some youtube, but they make that too hard. So we all just share the living room TV right now so that I can just keep an eye on things.

I'm still getting outsmarted though. I have locked-down parental controls on Minecraft so that my two kids can play together but not invite or be invited to stranger's worlds. But somehow, I see other people in there. Microsoft parental controls leave a lot to be desired. I don't know how non-tech-oriented parents do it.

michaelrpeskin··on The EdTech Revolution Has Failed
That's "equity" for you. We can't be unfair and give someone something that makes them better. It's easier to keep the top kids down than it is to lift the bottom kids up.
michaelrpeskin··on The EdTech Revolution Has Failed
^^ This!

My 4th grader has some sound sensitivity issues, so he uses that as an excuse to go out in the hall and read a book or listen to an auto book on his school-supplied chromebook when the classroom gets too loud for him

Well, guess what he's been doing? Watching unfiltered youtube! He's clever and sneaky, so it's been an arms race at home to block youtube enough for him but still allow it to work on my Roku for me. Why does the school allow 4th graders to have unfettered access to youtube?

michaelrpeskin··on Family medicine is in decline
My take on it is that ACA set up the insurance companies as essentially a cost-plus system. I forget the actual name for it in the ACA, but the general idea is that insurance needs to spend at least 80% of the premiums on health care. Sounds good at first pass, but there's also no limit to costs in the ACA. So the way an insurance company makes more money is to pay more for service. And the providers are happy to comply. Costs go up, insurance explicitly doesn't fight costs, so that premiums go up and they keep 20% of a bigger pie now.

I've essentially stepped out of the system. I get the cheapest, biggest deductible policy through my employer to cover a big emergency. But for my personal care, I do Direct Primary Care where I pay my doc $175/mo for my family of 4 and any time I need him, he's there "for free" no office visit fees, etc.

If I'm sick, I can always get same-day telehealth or next day in-office visits. For routine care, I may have to book out a week or so. He's never rushed because his incentive is consult me to keep me healthy so that I _don't_ need to see him. His incentives are not to keep me coming back for more visits.

During my last routine yearly visit, we had a hour long conversation about all the little nagging things that were happening has I get older, and we made a good plan on what to do.

I love this model and I could see it scaling like this:

Rather than forcing (individual mandate) ACA-specified insurance purchases, you are "forced" to put that cost into a HSA-like account (government supplemented for if you can't afford it), and then use the HSA to buy the insurance and/or DPC memberships you want and work for you. That would put downward pressure on price since you're making the decisions not the insurance/government. And you adds competition on service since you choose your providers.

michaelrpeskin··on The Friendship that made Google huge (2018)
I think that PP is generally talked about backwards. Everyone says that the least experienced programmer should drive while the senior programmer talks. When I was an intern 25+ years ago before pair programming had a name, I sat beside the grey beard who did the typing. I watched him and tried to keep up. For the first couple of weeks, I contributed nothing, but once I started to understand the code and his style, I was able to start seeing simple things, like reminding him to do a null pointer check, and then I started to see more and more stuff and I was able to contribute in real time.

I think this worked because 1) he was comfortable with me staring over his shoulder with neither of us talking and 2) I was comfortable with just watching an learning. I've tried this style with my junior programmers, and I'm too anxious being quiet and coding while someone watches me, and most juniors are also too anxious to prove that they're paying attention so they have to talk, and that breaks the flow.

michaelrpeskin··on Mermaid: Diagramming and Charting Tool
The lack of control on the layout can sometimes be a feature. If something starts to look bad, I often use it as a sign that my architecture is wrong. I’ve found that many times it’s helped point me in a better direction. If it’s easy for the graph layout engine, it’s probably a better architecture.
michaelrpeskin··on Can a product with "0g sugar" contain lactose?
I remember when I was in grad school, I was out with drinks with a bunch of students from the biology department. One of them was talking about finishing up a paper to submit to the FDA on a drug she was testing. She called the FDA "the department of legalized rounding". Another author on her paper was proposing that they could round ".445" to ".45" and then to ".5" so that they could be above a threshold to go on to the next phase. She was horrified, but they were going through the regulations because it wasn't clear that it was prohibited.
michaelrpeskin··on Repair and Remain (2022)
Ha! When we were putting in the flooring in the kitchen, I kept interrupting him "optimizing" the layout and he snapped at me like one of his workers. He immediately apologized since I was paying him, but I told him that he was supposed to yell at me because he is the boss when it comes to building stuff. He knows what he's doing and I'm just trying to learn. We have a great relationship and I know when to back off and let the expert just do the right thing.
michaelrpeskin··on Repair and Remain (2022)
Agree - I have a great handyman who also lets me "help". I can do the grunt work (carrying heavy stuff from the truck, demoing stuff, etc) saving him time, and he can do the stuff he's experienced at (he can eyeball a measurement and do a perfect cut much faster and better than I ever could) saving me time. Plus I get to learn how to do stuff. I get to be involved so I feel ownership, I get much better work done than if I did it myself, and I learn so that I can start to take on more projects myself.
michaelrpeskin··on A wonderful coincidence or an expected connection: why π² ≈ g
> A unit system is not just something that matches objective reality but something that has some cognitive ergonomy.

Beautifully stated!

And that's one reason why I like the US units of measurement better than SI. I mean, the divide-by-ten thing is nice and all. But _within a project_ how often are you converting between units of the same measurement (e.g, meters to centimeters)? You pick the right "size" unit for your work and then tend to stay there. So you don't get much benefit from the easy conversion in practice.

But if you're doing real hands-on work, you often need to divide by 2, 3, 4, and so on. So, for example, having a foot easily divisible by those numbers works well. And even the silly fractional stuff make sense when you're subdividing while working and measuring.

Of course it all finally breaks down when you get to super high precision (and that's probably why machinists go back to thousands of an inch and no longer fractions).

I think there's a little bit of academic snobbery with the SI units (though, it is a good idea for cross-country collaboration), but for everyday hand-on work the US system works really well. I always love the meme: There are two kinds of countries in the world, those who use the metric system and those who've gone to the moon.

I'm an AMO physicist by training and my choice of units are the "Atomic Units" where hbar, mass of the electron, charge of the electron, and permittivity are all 1. That makes writing many of the formulae really simple. Which is what you say: it has cognitive ergonomy (and makes all of the floating point calculations around the same magnitude). Then when we're all done we convert back to SI for reporting.

michaelrpeskin··on Weak Soft Skills: Why you are stuck at the Senior engineer level
I think it's even MORE important to have good "soft" skills to graduate above a senior engineer. If you go to a management track, you kind of have the prestige of "being the boss" that means you can be pretty bad at soft skills and still have people listen to you.

My company does it great, we have a full engineering-not-managing track.

I'm a principal engineer, and because I have "engineer" in my title, I'm responsible for engineering. Though, that doesn't mean I'm an IC. I probably only commit 100 lines of new code a year - just the stuff that's clearly my expertise (and we use that as a learning experience by having the juniors responsible for reviewing that code).

I do need to lead the engineering effort of the team. I'm not their manager, they don't need to do what I said just because I said so. So, I need to be able to communicate well which means both being technically correct and persuasive so that the engineers who actually do write the code, right the correct code.

It's almost all soft skills needed to shape the direction.

michaelrpeskin··on Is a 'slow' swimming pool impeding world records?
Yeah. I’m an aged mountain biker. Very mediocre. But when I travel to low altitude, I feel like Superman when I ride.

I live at 5000 ft now, but in my 30s I lived and slept at 9000 but did all my rides near 5000. That was an amazing time for my fitness. I could go hard at “low” altitude but recover at high which is what my body adapted for.

michaelrpeskin··on Making my own wedding rings
I did a similar thing using some titanium out of the scrap bin. (Yes, I know that it will be nearly impossibly to cut off in an emergency, but we just take them off before we go climb or do other dangerous stuff.) It’s been 15 years and they still look great. I left the tool marks on them, I liked how it looked like it was machined.
michaelrpeskin··on Show HN: Edna, note taking app for developers
For me, I like that HeyNote auto saves to buffer.txt and I can have a crib job auto commit that to git at regular intervals.
michaelrpeskin··on Show HN: Edna, note taking app for developers
Is desktop on the roadmap? I use HeyNote (desktop) every day and the only feature I _really_ want is multiple notes. You added that, but I need desktop because of corporate rules.
michaelrpeskin··on Ask HN: What skills do you think will be most in-demand in 10 years?
Roughly it can be viewed as a change of coordinates and then recognizing the symmetry. Say you had a 2D function of x and y and transformed it to r and theta and then noticed that it was independent of theta. You go from an 2D problem to a 1D one.

In this case we’re transforming a 7D problem into a sum of a discrete set of 2D problems. With more algebra we could get it down to a 1D problem but that would take more human work that wouldn’t be paid off later in CPU time. If their project ends up scaling and the 1D transformation makes sense, we’ll do it.

← PreviousPage 2 of 8Next →