Pritam

time is relative

[Dmitrii Kovanikov @ChShersh]

You’re responsible for how you use AI.

You can use it to explore novel domains and learn new things and become smarter.

Or you can use it to delegate challenging work while you’re getting dumber and dumber.

[Dmitrii Kovanikov @ChShersh]

When I was doing competitive programming, I constantly faced problems I couldn’t solve.

On a daily tech job, I’ve never faced a problem I couldn’t solve.

[Arpit Bhayani @arpit_bhayani]

Cloudflare has been building Meerkat, their new global consensus service, built on a protocol called QuePaxa instead of the usual Raft or Paxos.

To be honest, I did not know about QuePaxa before this. I learned about it from the post itself, and it is genuinely interesting.

Most consensus systems, like Raft, rely on timeouts to detect a dead leader and trigger a new election. That works fine until your network latency fluctuates a lot, which is exactly what Cloudflare deals with every day.

Too short a timeout, and replicas panic and block writes. Too long, and the system just sits there while something is actually broken. So instead of Raft, Meerkat runs on QuePaxa. Here is how QuePaxa avoids the leader problem.

A client does not need to go through a leader at all; it can contact any replica, and that replica can drive consensus for the current slot on its own. A leader still exists in the system, and it has one advantage: if it is the one proposing, it takes just one round trip to reach a decision.

A non-leader replica needs three or more round-trip to do the same thing. So the leader speeds things up, but it is not a requirement for progress. If the leader goes down or slows down, any other replica simply picks up the work, and writing keeps flowing.

Also, concurrent proposals do not conflict destructively. The replicas coordinate and converge on a single agreed value regardless. Interestingly, Cloudflare is planning to build a full key-value store and a leasing system on top of it.

I have started reading up on QuePaxa after this post, and wanted to share what I found first.

Hope this helps.

[Pritam @pritam_01]

@railcomplaint Now delayed by 8 hours

@railcomplaint

[Pritam @pritam_01]

Train 02569 is running 7 hours late, was scheduled to reach Delhi at 4:30 AM, now expected only at 11:30 AM.
What's worse: despite this massive delay, the train is still being halted at multiple non-stoppage stations just for crossings, delaying it even further.

@railcomplaint

[Netflix India @NetflixIndia]

Everyone knows who owns these colours

[pulkit mittal @pulkit_mittal_]

for HFT interviews, i had to prepare for computer science round even for a senior role. and all companies ask these questions from freshers.
here's the list of most commonly asked computer science questions.

Operating Systems (OS)
1. Process vs thread
2. Virtual memory, paging, thrashing
3. Fragmentation: internal vs external
4. Semaphore vs mutex
5. Producer consumer problem
6. Deadlock and its conditions
7. Banker’s algorithm
8. Scheduling algorithms: FCFS, SJF, SRTF, RR
9. RAID and its types
10. Direct mapping vs associative mapping
11. socket, kernel and monolithic kernel
12. main memory vs secondary memory

DBMS
13. ACID properties
14. Normalization vs denormalization
15. Conflict serializability
16. Concurrency control protocols
17. Shared lock vs exclusive lock
18. Indexing
19. Vertical vs horizontal scaling

Computer Networks
20. TCP vs UDP
21. Types of network delays
22. What happens when you type google . com
23. Three way handshaking
24. IPv4 vs IPv6
25. how DNS works
26. IP address vs MAC address
27. Subnetting
28. Firewalls
29. Server side load balancing
30. HTTP vs HTTPS

good sources for prep are Gate videos and TUF blogs.

[Swarnit @swarnitghule]

spend time on learning cache memory today

1. L1i, L1d,L2, L3 cache
2.SRAM VS DRAM
3.cache locality
4.cache miss/hit rate
5.cache lines
6.false sharing
7.cache line bouncing
8.cache thrashing , LRU, P-LRU, random, LFU etc

tomorrow i will start simd and os internals.

[AADITYANSHA @aadityansha_06]

https://t.co/oyaqtAOcgT

[Andrej Karpathy @karpathy]

Personal update: I've joined Anthropic. I think the next few years at the frontier of LLMs will be especially formative. I am very excited to join the team here and get back to R&D. I remain deeply passionate about education and plan to resume my work on it in time.

[OpenCode @opencode]

OpenCode x Qwen 3.6 Plus - free, again

Last time y’all treated our capacity like an all-you-can-eat buffet.

We found more GPUs. Round 2.

[Arpit Bhayani @arpit_bhayani]

pro tip: get good at sounding confident even when you know nothing. ask questions when essential, and trust yourself to figure the rest out.

because everything is figureoutable.

[FFmpeg @FFmpeg]

official donation link: https://t.co/Vz6mGGH0jg

[FFmpeg @FFmpeg]

A few "behind the scenes" shots from our @lexfridman podcast

1) Post interview photo
2) Post interview @TerryBlacks_BBQ
3) Fatigue after six hours of talking

Thank you to all the people who took the time to write nice messages.

[Tom Dörr @tom_doerr]

Emulates 45 AWS services locally with 13 MiB of memory

https://t.co/uIwqUrWPLg

[anthony @asusevski]

is this what dario sees when he keeps saying SWE will be automated

[Mrinal @Hi_Mrinal]

redis, valkey, aerospike, lmdb, postgres, clickhouse, duckdb, zookeeper, redis cluster, redis sentinel, redis bloom, redisearch, redisjson, redisgraph, redispubsub, redis streams, redis modules, valorant, slowlog, latency monitor, info all, client tracking, hot key, key eviction, lru, lfu, lru-clock, maxmemory, volatile-lru, allkeys-lfu, oom score, memory overcommit, jemalloc, tcmalloc, arena, slab allocator, ziplist, quicklist, listpack, skiplist, intset, hashtable, dict, rax (radix tree), hyperloglog, bloom filter, cuckoo filter, count-min sketch, redis cluster proxy, twemproxy, predixy, haproxy, envoy, connection pooling, pipelining, multiplexing, resp3, resp2, client-side caching, tracking invalidation, broadcast mode, nodelay, tcp backlog, jaeger, zipkin, flamegraph, perf, bcc, ebpf, redis prof, database profilers .....

And people still ask me "give me topics to write technical blogs on" 🙏😭

[FFmpeg @FFmpeg]

Sending patches to FFmpeg is the way out

[Mo @atmoio]

The greatest con of the decade was calling autocomplete “AI”.

The second greatest is calling autocomplete-in-a-loop an “agent”.