Hacker News: jhgg

New comment by jhgg in "Show HN: Utilyze – an open source GPU monitoring tool more accurate than nvtop"

jhgg — Mon, 27 Apr 2026 17:20:53 +0000

We just track power utilization.

New comment by jhgg in "Discord: A case study in performance optimization"

jhgg — Sun, 15 Feb 2026 02:11:02 +0000

Cool article :)

New comment by jhgg in "Gleam OTP – Fault Tolerant Multicore Programs with Actors"

jhgg — Mon, 20 Oct 2025 05:13:19 +0000

Yes, agreed, hence rarely a problem in practice ;)

New comment by jhgg in "Gleam OTP – Fault Tolerant Multicore Programs with Actors"

jhgg — Mon, 20 Oct 2025 02:30:05 +0000

Risk of deadlock is real if you have processes calling each-other in a cyclic way. e.g. process A sends GenServer call to process B, that then sends a GenServer call to process A to in order to handle the original call. However, process A is busy waiting on B to reply to it's initial call.

This is rarely a problem in practice however.

New comment by jhgg in "Users claim Discord's age verification can be tricked with video game characters"

jhgg — Sat, 26 Jul 2025 06:39:54 +0000

    > government passes law that requires companies to age verify users
    > said government provides no way to actually verify a human's age
    > hilarity ensues

New comment by jhgg in "Show HN: I wrote a new BitTorrent tracker in Elixir"

jhgg — Fri, 20 Jun 2025 01:35:57 +0000

Really cool! You looking to write Elixir as your main job?

New comment by jhgg in "Eleven v3"

jhgg — Thu, 05 Jun 2025 21:26:34 +0000

This is not a model issue - you just have not properly implemented acoustic echo cancellation on your end.

New comment by jhgg in "Updated rate limits for unauthenticated requests"

jhgg — Wed, 14 May 2025 21:49:21 +0000

The truth is this won't actually stop AI crawlers and they'll just move to a large residential proxy pool to work around it. Not sure what the solution is honestly.

New comment by jhgg in "The Cost of Being Crawled: LLM Bots and Vercel Image API Pricing"

jhgg — Tue, 15 Apr 2025 00:56:37 +0000

Sweet! That's much more reasonable!

New comment by jhgg in "The Cost of Being Crawled: LLM Bots and Vercel Image API Pricing"

jhgg — Tue, 15 Apr 2025 00:42:19 +0000

$5 to resize 1,000 images is ridiculously expensive.

At my last job we resized a very large amount of images every day, and did so for significantly cheaper (a fraction of a cent for a thousand images).

Am I missing something here?

New comment by jhgg in "Query GPT: Transform Natural Language into SQL"

jhgg — Sat, 05 Apr 2025 02:28:02 +0000

IMO, it's not. All the UI boils down to: https://github.com/zhangchenchen/query_gpt/blob/main/src/api...

New comment by jhgg in "IO Devices and Latency"

jhgg — Thu, 13 Mar 2025 18:07:17 +0000

Metal looks super cool, however at my last job when we tried using instance local SSD's on GCP, there were serious reliability issues (e.g. blocks on the device losing data). Has this situation changed? What machine types are you using?

Our workaround was this: https://discord.com/blog/how-discord-supercharges-network-di...

New comment by jhgg in "Datadog acquires Quickwit"

jhgg — Fri, 10 Jan 2025 21:55:31 +0000

Hoping this leads to datadog launching a logging solution that does not cost an arm and a leg at scale.

New comment by jhgg in "How to make LLMs shut up"

jhgg — Sat, 21 Dec 2024 20:05:31 +0000

Except this article is about how their efforts to prompt the LLM didn't end up working and how they used embeddings / vector search to filter out comments that the LLM generated based on user feedback.

New comment by jhgg in "ChatGPT Pro"

jhgg — Thu, 05 Dec 2024 21:11:52 +0000

Does that matter when you can just swap it for gpt-5-whatever at some point in the future?

New comment by jhgg in "ChatGPT Pro"

jhgg — Thu, 05 Dec 2024 20:13:19 +0000

Are you implying this technology will remain static in its capabilities going forward despite it having seen significant improvement over the last few years?

New comment by jhgg in "How much memory do you need in 2024 to run 1M concurrent tasks?"

jhgg — Fri, 29 Nov 2024 06:55:28 +0000

I did this measurement, and using time -v, the maximum resident size in KB comes out to 440,424 kb for 1m tasks, 46,820 kb for 100k, and 7,156 kb for 10k.

New comment by jhgg in "Using Erlang hot code updates"

jhgg — Tue, 19 Nov 2024 23:43:09 +0000

When I worked at Discord, we used BEAM hot code loading pretty extensively, built a bunch of tooling around it to apply and track hot-patches to nodes (which in turn could update the code on >100M processes in the system.) It allowed us to deploy hot-fixes in minutes (full tilt deploy could complete in a matter of seconds) to our stateful real-time system, rather than the usual ~hour long deploy cycle. We generally only used it for "emergency" updates though.

The tooling would let us patch multiple modules at a time, which basically wrapped `:rpc.call/4` and `Code.eval_string/1` to propagate the update across the cluster, which is to say, the hot-patch was entirely deployed over erlang's built-in distribution.

New comment by jhgg in "Show HN: Lyceum – An MMO game built with Zig and Erlang"

jhgg — Mon, 11 Nov 2024 16:04:27 +0000

Having a potentially untrusted client connect to the erlang node as a c_node (which seems to be what zerl does) is not a good idea generally, as connecting that way essentially allows the client to execute arbitrary code on the server.

New comment by jhgg in "Zigler: Zig NIFs in Elixir"

jhgg — Thu, 24 Oct 2024 19:14:04 +0000

https://github.com/discord/sorted_set_nif