gommm
gommm on Hacker News
As my eyesight is getting worse I do end up appreciating bigger phones. Even though I miss how smaller phones felt in my hand, it's just hard to see small text on it and that's even though I don't use my phone all that much since I have a kindle for reading and a computer for everything else.
I've been running my own email server with exim and dovecot for 20ish years, I like the freedom it gives me but I have to say it's a significant investment. Not so much keeping things up to date but instead making sure that my emails are actually delivered. I've eventually switched to using fastmail just for sending emails and even with that, I don't get perfect delivery.
What would you recommend to increase international equity exposure? Index funds ETF like VWRA?
We use luuni which is similar (except that it also enable choose your own story with audiobooks). Even then, we limit it because otherwise he would want to listen to it every time before sleeping (and it prevents him from sleeping)
True for children under 3, however I see plenty of 8-9 years old glued to their tablet in restaurants.
For me open source means that the entire training data is open sourced as well as the code used for training it otherwise it's open weight. You can run it where you like but it's a black box. Nomic's models are good example of opensource.
> My personal headcanon: this tooling works well when built on simple patterns, and can handle complex work. This tooling has also been not great at coming up with new patterns, and if left unsupervised will totally make up new patterns that are going to go south very quickly. With that lens, I find myself just rewriting what Claude gives me in a good number of cases.
I've been doing a greenfield project with Claude recently. The initial prototype worked but was very ugly (repeated duplicate boilerplate code, a few methods doing the same exact thing, poor isolation between classes)... I was very much tempted to rewrite it on my own. This time, I decided to try and get it to refactor so get the target architecture and fix those code quality issues, it's possible but it's very much like pulling teeths... I use plan mode, we have multiple round of reviews on a plan (that started based on me explaining what I expect), then it implements 95% of it but doesn't realize that some parts of it were not implemented... It reminds me of my experience mentoring a junior employee except that claude code is both more eager (jumping into implementation before understanding the problem), much faster at doing things and dumber.
That said, I've seen codebases created by humans that were as bad or worse than what claude produced when doing prototype.
I just run CC in a VM. It gets full control over the VM. The VM doesn't have access to my internal networks. I share the code repos it works on over virtiofs so it has access to the repos but doesn't have access to my github keys for pushing and pulling.
This means it can do anything in the VM, install dependencies, etc... So far, it managed to bork the VM once (unbootable), I could have spent a bit of time figuring out what happened but I had a script to rebuild the VM so didn't bother. To be entirely fair to claude, the VM runs arch linux which is definitely easier to break than other distros.
That's been my experience as well. Claude code does better with Elixir (plus I enjoy working on the code better after :) )
That's not helped by a recent change to their system prompt "acting_vs_clarifying":
> When a request leaves minor details unspecified, the person typically wants Claude to make a reasonable attempt now, not to be interviewed first. Claude only asks upfront when the request is genuinely unanswerable without the missing information (e.g., it references an attachment that isn’t there).
> When a tool is available that could resolve the ambiguity or supply the missing information — searching, looking up the person’s location, checking a calendar, discovering available capabilities — Claude calls the tool to try and solve the ambiguity before asking the person. Acting with tools is preferred over asking the person to do the lookup themselves.
> Once Claude starts on a task, Claude sees it through to a complete answer rather than stopping partway. [...]
In my experience before this change. Claude would stop, give me a few options and 70% of the time I would give it an unlisted option that was better. It actually would genuinely identify parts of the specs that were ambiguous and needed to be better defined. With the new change, Claude plows ahead making a stupid decision and the result is much worse for it.
Personally I love rust for agents because of types. In the ruby world there's sorbet and rbs so would be interesting to try that.
When I was an exchange student at RIT and had just arrived from France a month before, one of the admin staff invited me and a friend in the same situation for thanksgiving because she didn't want to leave us by ourselves for a major holiday.
I have fond memories of that kindness.
I've usually not been impressed in AI's implementation of math heavy rules so I wouldn't trust it much and I tend to find it easier for me to write them myself and then verify :) Yup, it's always interesting to see the different usages.
Oh yeah, that's why I added the parenthesis. I consider lisp macros to be a dsl and that's exactly what I tend to like using. Similarly with ruby and some meta programming tricks.
I tend to find that for things like this that are really math heavy, it's usually better to create a DSL (or create easily readable function calls, etc) that you can easily write yourself instead of relying on AI to understand math heavy rules.
Bonus points, if the rules are in an easily editable format, you can change them easily when they need to. It seems that was the path the author took...
And yes this kind of use-case is exactly where unit tests shine...
Oh I mean alternative would be a well regarded university/school in Germany or France... I'm French but we live in HK and most kids here (even the ones who go to the French International School or the German Swiss School) end up trying to go to UK or US universities. French and German international schools tend to not be that well ranked in the most well known rankings despite being very good technically (which is annoying when trying to get a visa to certain countries).
Part of my bias is that I was an exchange student at RIT and while I appreciated the experience, I was not impressed by the CS courses or the level of maths of the students going there.
That's what the French government paid per year per student at my engineering school in the early 2000s. Tuition fees paid by the student were 540 euros a year, but the cost to the government was quite high.
France is the same, the better universities are all public. But I know that the government spent an average of 35,000 euros per students at top public engineering schools in the early 2000s, not sure nowadays, so they do have funds it's just that the way of bringing money depends on actually being great academically.
This is why when my son is old enough to choose a university, I'd probably try to advise him against doing undergrad in a UK or US university if he's studying STEM. Based on interviewing CS graduates, it doesn't seem that the level is that high in most UK/US universities compared to other countries (of course with the exclusion of the very top) and that seems partly due to a culture of pushing for profits over education and making it very hard to fail.
I've been using it since I got it. It's been working great with one small issue that I haven't been able to solve. For some reason when I use plasma on Arch linux (but not ubuntu), the display outputs garbage. I'm guessing it's not detecting the EDID correctly and setting a weird resolution or refresh rate. It's not a major issue since other desktop work well so I haven't spent much time looking into it.
Since you seem knowledgeable about knives, Do you know any great knives makers? And are those custom steels or carbide blades worth it?
So far, I mostly sharpen my knives on the back of a plate. So definitely could be doing more :)
To be fair, tipping the cook makes more sense to me than the waiter. I come to a restaurant for the food, I don't particularly care about the service beyond a certain baseline. It never makes sense to me that waiters can earn more with tips than kitchen staff.
I end up using Perplexity a lot too, especially when I'm doing something unfamiliar. It's also a good way to quickly find out what are best practices for a given framework/language I'm not that familiar with (I usually ask it to link to examples in the wild and it find opensource projects illustrating those points)
> The notch has been around for 4 years now, and Apple still hasn't provided a solution for the problem they introduced.
As a lot of people told you, you can just disable it. I've been doing that for 4 years, just set your resolution to a 16:10 ratio and you're good to go. The resolution is exactly the same as it was before they introduced the notch
Personally I like the fact that Apple gives us the choice. I dislike the notch and prefer my menu bar below because I use apps like intellij. My wife likes the notch and keeps it. So, both of us can have what we want.
Maybe Apple could have made it slightly easier to disable it by having an option instead of choosing a 16:10 resolution but, to be honest, most of the people who dislike it tend to be power users who can figure it out.
They pushed the U2 album a few years ago and had that blow over their face when they realized that people actually did mind.
And when it comes to doing the analysis, how do they determine that users don't mind? Focus groups?
This looks like LLM blog spam. To be taken seriously, they'd need to publish the implementation in each language of each benchmark which they didn't.
Instead they show pseudocode with very vague descriptions of failure mode that do not really make sense: "Under our error cascade simulation, some low-level failures in unsafe code regions propagated in ways that eventually caused deadlocks in resource management." That doesn't give any details nor does it sound like a realistic failure case to have "failures in unsafe code regions".
Outside of charities, I donate to a few projects:
- Openbsd foundation (I use openssh all the time :))
- Corejs + babel (donated after reading this thread https://news.ycombinator.com/item?id=34859766
- Magisk . I root my Android phone because I believe that if I buy a device I should actually own it.
- Scummvm I believe in preserving computing history and scummvm does great work making older games accessible
- MisterFPGA. Instead of giving my 3 years old son a tablet, I've decided to set up MisterFPGA for him with an Amiga core for him to do kidpix and scummvm to run Adibou and the humongous games. I like the idea of him having access to a relatively simple system like the amiga where he can learn how computer work without having access to internet and learning to passively consume.
- Valetudo. I love having a robot cleaning my house. I do not love having something with a camera that's not purely local. Thanks to valetudo, I can use it without worrying about my privacy.
- Not directly a donation but I buy a yearly license to crossover to support wine development
- Calibre
- Syncthing
- The developer of Karabiner
- Internet Archive
- Free Software Foundation
I used to give to the following:
- Wikipedia (but stopped when I realized that they have more than enough donations and that I should focus on other worthy causes)
- Mozilla (but stopped when I saw the CEO increase her salary while firing and stopping important projects. I do not want a repeat of the IE monopoly with Chrome and as such I want Firefox to succeed but I have completely lost trust in Mozilla's management)
Exactly, it's not an either/or situation. I've worked on a codebase that did 3-4k/s jobs through sidekiq and in this case sidekiq is obviously the best option (although we also used a go sidekiq client for certain tasks that didn't need the rails app context) but using sidekiq from the beginning is premature optimization. You can easily start with Solid Queue and switch over to sidekiq when you actually have data that tells you that it'll potentially become a bottleneck. Premature optimization is a complexity cost.