News for hackers. Stuff that matters.

snizovtsev

It was ~3–6 months ago in our org. Now management has decided to trust AI reviews, and we ship to production without a human in the loop. Now my job is to manage tokens and invent new ways of AI-guided testing of already shipped code after release. It's better than review fatigue, but I still feel bad.
Re:WTF Ubuntu? on Monday November 02, 2026 @12:52PM (#50340696)
The patch itself isn't related to MPTCP, but the bug is hard to exploit. AFAIK, Anthropic is the only ones who claimed exploitation using their unlimited frontier models without safety restrictions. So non-MPTCP configurations may also be vulnerable, but exploiting them could require even more resources. I've tried to trigger KASAN using Qwen 3.8 27B @ local RTX3090 - no success after 96 hours.
Re:Backups Aren't Simple on Wednesday September 16, 2026 @10:14PM (#49735644)
You can exercise it against regular, non-Glacier S3. Same flow; the only difference is that you skip one attribute on upload and one request/wait step on download.
Re:Backups Aren't Simple on Wednesday September 16, 2026 @09:24PM (#49735302)
I'm polishing my backup system right now, and found that `tar | zstd | gpg` plus a small vibe-coded wrapper is pretty much all I need.

GNU tar has its own incremental index via `--listed-incremental=FILE`. Unlike Borg or Restic, which have their own more complicated repository formats, this leaves me with just one additional file (the `.snar`) alongside a dumb, portable full-disk tar archive.

The nice part is that, unlike repositories that require both read/write access patterns, tar can compute deltas using only the small `.snar` file, while the main `full-disk.tar` can remain buried in write-only Glacier storage. This makes it a much better fit for Glacier's write-once model and 180-day minimum retention.

My current plan is to upload a full-disk tar to an external HDD + Glacier on a weekly cadence, and the `.snar`-based deltas daily. That gives me a pretty simple cloud backup solution for a few bucks a month (after burning through $100+ of free credits).

Re:RAM Has a Design Flaw from 1966. I Bypassed It [video] on Sunday April 12, 2026 @06:01AM (#47737871)
Not really. FPGAs run some simple proven strategies, but hardware guys can't keep up with how fast quants iterate on new strategies - so CPUs stay relevant.
Re:Introducing tmux-rs on Friday July 04, 2025 @01:46PM (#44466488)
I started working on this feature about a month ago: https://github.com/snizovtsev/tmux (it's in early experiments, not ready for a try).

Luckily tmux has a well-established "control mode" protocol designed for iTerm2 that can serialize internal state updates into a stream of text messages. So I do write client-side of this feature right inside tmux.

When it's ready, you will able to embed remote sessions inside local tmux instance by calling something like "ssh example.com tmux -CC attach". It will auto-detect "control mode" escape sequence and create special "remote session" you can switch into.

If you are interested, connect with me ([email protected]) so I will notify when feature is ready and ask for early testing before making upstream proposal.

Most likely gas stations will have to buy these systems to keep their business running.
Re:Discovery Coding on Wednesday January 29, 2025 @05:19AM (#42863410)
No, it will be "somehow working solution for case A" and "somehow working solution for case B" but not together.
Re:My failed attempt to shrink all NPM packages by 5% on Monday January 27, 2025 @11:09AM (#42842584)
Yes, but it was expected. It's like prioritising code readability over performance everywhere but the hot path.

Earlier in my career, I managed to use Zopfli once to compress gigabytes of PNG assets into a fast in-memory database supporting a 50K+ RPS web page. We wanted to keep it simple and avoid the complexity of horizontal scaling, and it was OK to drop some rarely used images. So the more images we could pack into a single server, the more coverage we had. In that sense Zopfli was beneficial.

Re:Intel's Tofino P4 Software Is Now Open Source on Thursday January 16, 2025 @10:53AM (#42726931)
I think half of Tofino's complexity was in their compiler. So it may inspire new hardware vendors to reuse it in some contexts.
Re:Engineer eats efficiently for $2.50 a day (2016) on Monday January 06, 2025 @04:54AM (#42609177)
Optimal in theory but totally uncookable.
Re:Engineer eats efficiently for $2.50 a day (2016) on Monday January 06, 2025 @04:51AM (#42609161)
Lower labor cost and heavily subsidized agriculture.
$999 is for soldered 8GB/256GB crap. Add $400 for 16GB/512GB models.
Re:Freenginx: Core Nginx developer announces fork on Thursday February 15, 2024 @03:48AM (#39380449)
Living in Russia could be very cheap compared to other countries. If you own a flat and you don't need cars or travel, then it's possible to live a few years just on money saved from your previous software job.
Re:Ask HN: V2 – Laid off folks, are you getting hired yet? on Saturday March 11, 2023 @11:37PM (#35116725)
I had to quit in September due to burn out. Given how many better engineers now on market I feel there are no space for me anymore. Now burning my saving money on rent and food and going to commit suicide few months later.
Re:GNU Make Guile Integration on Tuesday March 17, 2020 @10:24AM (#22605067)
make -p -f/dev/null | grep FEATURES
Re:H-1B in 2020: Immigration Visa Faces Further Restrictions on Tuesday December 03, 2019 @09:38AM (#21692557)
I think foreign remote employees are much more risky in terms of IT security and law protection. Companies need to take this risk into their business model or develop countermeasures beforehand.
Another online bomberman clone with massive multiplayer from Russians: http://bombermine.ru/