HN user
Lots of people take LLM feedback as mostly trustworthy because it doesn't _directly_ try to sell you stuff, yet. Once it gets more in your face, the trust average people place in the quality of answers should see some erosion. In terms of the influence the answers have, this should have a much needed sobering effect.
Let the tiger show it's stripes so that we are reminded what it is actually after.
I run Jellyfin on Raspberry pi and access it with Android app. The main issues I have with the app are:
1. Subtitle management is a total pain. Sometimes it doesn't show any subtitle toggle, sometimes it does. There isn't an option to switch between different subtitles. Adding an external subtitle file on the fly via the app never works for me. One approach that almost always works is to download the subtitle file, rename it to match movie filename, trigger a metadata refresh for the file/library (this requires switching from library view to dashboard view and back so certainly has friction). Also, if you start a movie directly on Chromecast, subtitles don't show even if they have been chosen. However, if you start the movie with subtitles "ON" on the app and THEN stream it via Chromecast, subtitles show up on the TV.
2. Sometimes it has issues with Google Chromecast (puck connecting to my LG OLED TV via HDMI port). I have had better luck with the in-built Chromecast feature of the TV.
The Pi isn't mega snappy when it comes to the lag between button press to actual playback start/end of seeking but it generally works fine once the stream starts playing.
I didn't pay a dime for it so this is not throwing any shade on the Jellyfin team - I am grateful for their work.
You'll have so much more time now that you aren't AST wrangling.
So much time because you were fired and have nothing to do. Nothing is as good as oodles of quality time worrying about funding your family's needs. It's amazing.
I fully agree. This isn't my usual pattern. I have a Claude pro subscription + some money loaded into DeepSeek which is super cheap. Just wanted to try out Kimi K3 and sort of got into the flow with the personal project. I am definitely not going to be burning more cash on this once the current recharge amount runs out.
Damn some timing, I just signed up few hours ago and within 3-4 hours spent over 50 bucks on my personal project (didn't realize that the launcher script I copied from an example had set effort to max). This is API use that consumes from your recharge amount. It found tons of issues with the code generated by Sonnet 4.6 (yeah I know, older model). Also didn't seem to burn through context window that fast (1M). The generated code works but quality could be better. Works fine with Claude code cli. The standout aspect was strong analytical abilities and a much more sobering and cautious tone (which is a good thing).
Will have to watch cost carefully though. If only I could get it at DeepSeek adjacent prices.
What I said was specific to work and it is not entirely clear to me why the cost goes up so much the moment I switch to medium with just Sonnet 4.6 (newer models would be even more I guess). There is a monetary budget allocated and I've blown it easily in the past with Claude. Cursor has been a life saver.
That's what I have been doing with Cursor with auto mode. Let the model manage the cache through compaction but staying focused on the topic. It has been quite efficient cost wise and quality wise too. It found issuesbat cheaper cost that Sonnet 4.6 medium effort had missed while costing more. I would like to repeat this test with higher tier Anthropic models like Opus effort medium to see the results.
arbitrary numeric values which are defensible
Defensible on account of discharging their relentless duties of generating shareholder value within legal limits... i.e. profit.
The fact that it is done in a lopsided way at the expense of everything else is just sign of societal decline due to deification of money and letting the tiger of capitalism roam without boundaries.
Where has all the empathy gone? And common sense? :(
Contrary to popular belief, I don't think common sense has gone away. It just has been deprioritized at the altar of profit. It'll be visible loud and clear the very instant if/when the lives of the assholes chasing profits was on the line.
If I can expect to be penalized for not paying my legitimate bills, companies should also be penalized for failing to implement common-sense reasonable safeguards that prevent them from slapping their consumers with such absurdities.
In that case, they settled with Robinhood and FINRA fined Robinhood 70 million dollars.
https://edition.cnn.com/2021/07/01/business/robinhood-lawsui...
I am not certain about Windows but on Linux, aren't applications supposed to use PATH based resolution provided by shell for this? Why go to the root of the project directory at all looking for binaries? Because it has been become their favorite place due to readme / skills etc being there?
One thing that I have noticed somewhat cuts through the haze of prompt->change->prompt is to drop into the editor when I notice something that is not exactly bad but isn't exactly what I want. It doesn't even have to be a big change - infact smaller annoyances are better in this case because they itch more.
Don't let the agent scratch that itch - it is your hungry inner programmer expressing the need to eat i.e. do it correctly as per your taste. That morsel of work is sustenance.
It also activates the thinking gears that only kick in when I am writing the code. In the same process, the code sinks in better and becomes part of the mental model, even if I may not have paid attention to every single line.
I have also noticed that the mode switch puts me in a marginally better mood - sort of a recharge for the next bout of prompt->change->prompt.
FDEs generally work with customers, long-term.
In these days of mass layoffs every month, talking about "long-term" sounds like a cruel joke.
It's okay to not do something for a minute.
Not doing anything and being totally idle for a while is nothing to sneeze at. Especially in this day and age.
Seems like if the experiment didn't work, nothing to lose. But it did and he capitalized it in the most in-your-face possible. it would be interesting to line up the timelines of the Rust experiment vs him joining Anthropic.
This is true in Singapore and Malaysia, as well, where Filipino or Indonesian cooks and housekeepers are extremely common, as are separate entrances--typically into the kitchen.
Live in housekeepers are very common indeed in Singapore. However, majority of Singapore lives in Housing Development Board flats that do not have any separate entrance into kitchen.
Anyway, reading this reminds me of the old chestnut from networking about choosing UDP over TCP for its low-latency and simplicity and then eventually adding nearly all the reliability facilities of TCP to the app (automatic retry
Sometimes you can build a successful business out of doing so: https://aeron.io/
It would be if it suddenly became popular and they could show more ads to people for using it.
Because you're probably come from a high trust culture where you've been taught reciprocal trust...
That is just a long sentence for "us" vs "those people".
Having said that I don't entirely deny the effect of society on people's behavior. But at the same time, I have seen people from so called high-trust society being all polished and nice on the surface while being assholes and people from so called low-trust society being genuinely decent people despite not having the right name or the surface polish.
Also, assholes tend to attract assholes and people of the same tribe/clan/race tend to form groups.
I did the side by side between Claude code (effort medium) and Cursor (auto). Asked Claude to prepare a plan and asked Cursor to review it and it found tons of gaps in the plan. Cost-wise, it came out better too. I have been using Cursor daily (along with Claude) and the former has been 20-30% cheaper despite me spending more time on it.
This is like repeatedly trying to train a dog with amnesia to not poop in the bedroom. Despite the dog repeatedly doing so and moreover being particularly easy to be fooled into doing so.
It can't reliably learn so stop trying to teach it. Lock the bedroom instead.
- I am shell-shocked at the lack of quality assurance! Can you at least apply a patch?
Employee as Kubernetes pod.
Maybe I am being naive but doesn't that seem like a basic and effective approach that others should have been able to implement too? Why doesn't Xbox or Microsoft do it? Optimistically download updates so that if the user wants, it is ready to go. If low on space, sacrifice those cached updates because they can be downloaded again.
this is a good lesson for founders. The culture you build into your company will likely outlast your tenure.
Good founders already know this. Bad ones don't care.
The biggest casualty of that is probably stored procedures.
Not much can beat stored procs when it is dealing with multi-step heave volume stuff. But I don't miss not having to do hacks for logging and debugging compared to the flexibility offered by non-db side.
For pretty much everything else, the poor ability to log and debug makes them a headache to manage. I
That was one of the needs we had during my initial days - dynamic DDLs/DMLs. It was basically a bash + SQL stack which is fairly low level. I remember discovering Perl was installed on the Sun Solaris boxes, learned it and soon everyone jumped on it and boy what a massive step-up from bash that was!
That is exactly what I was thinking. There is such a low barrier to entry with an outsized payoff.
As someone who started their programming journey with SQL, it just feels so odd hearing about learning SQL being presented as an useful option. I get it, it just feels odd. SQL was considered table stakes in the financial IT world - if you said you didn't know SQL, people would look at you funny.