HNHacker News
TopNewBestAskShowJobs

rkwz

2,909 karma · joined October 8, 2010

sheshbabu.com
submissionscomments
rkwz··on Are Open-Source Large Language Models Catching Up?
Oh, can install Ollama in Linux/WSL as well
rkwz··on Autospam and Naive Bayes
> Naive Bayes is also surprisingly powerful for deduplication of large datasets

Thanks for this! I implemented Entity Resolution from scratch 2 years back and was mostly able to do deterministic linkages using identifiers or using TF-IDF on names as a suggestion for human intervention. Will be writing about my learnings in upcoming months [1].

Was mostly looking at research papers before but will take a look at your project - your documentation looks good!

[1] https://www.sheshbabu.com/posts/entity-resolution-challenges...

rkwz··on Autospam and Naive Bayes
Thanks for this, been meaning to read up on the Fellegi Sunter paper for a long time!
rkwz··on Entity Resolution Challenges
Thanks, TIL!
rkwz··on Entity Resolution Challenges
Thanks for sharing the tip, would definitely try this out in near future!

> figure out how to pull out entropy scores Can you elaborate on this? "entropy scores" == how unique a value is?

rkwz··on Entity Resolution Challenges
Curious, how would LLMs help here?
rkwz··on Man spends entire career mastering crappy codebase
Unrelated: this link opened almost instantly, what’s this magic? Even HN takes longer!
rkwz··on Deutsche Bank's “dysfunctional” IT division (2018)
> Now that Cryan's gone, however, it's becoming apparent that his authority over some senior managers at Deutsche Bank was weak. He and Hammonds' ability to challenge Deutsche's most powerful tribal leaders appears to have been correspondingly limited. "It's going to be a far more uphill task to cut systems in future," says one senior insider. "The people who control these systems don't want them cut. It's all about politics."

> "Hammonds was innovative, but she couldn't do what she set out to do," says one DB insider. "So we're just sitting here, waiting to see what happens next."

Yikes! Sounds like an impossible uphill task.

Is there any examples of such turnarounds in the recent years?

rkwz··on Meta’s Microservice Architecture [pdf]
> Your chief compliance officer is intimately familiar with the data protection benefits that microservices bring and is leaning heavily into it in their regulatory submissions as a way to compensate for some other gap in the business

Just curious, how does microservices help here?

rkwz··on Show HN: HN Trends
Looks great!

Just checking, do you use stop words (removing “the”, “an” etc)? Can also add “its”.

rkwz··on AI Assistant in JetBrains IDEs
The name suggestion is a good idea!
rkwz··on GitHub Global Navigation Beta
Off topic, how do they create these nice gifs of screen recording?
rkwz··on Ask HN: What Is Going on with Neo4j?
I was leaning towards Neo4j as it had support for graph algorithms such as WCC, louvain etc

What would be a good alternative for it?

rkwz··on Tell HN: My “Rust Web Development” book is released
Congrats!! 2 years is a very long time, but your persistence payed off!! Any tips that helped you stick to the same project for long period of time?
rkwz··on Ask HN: What are some examples of cloud lock-in?
We migrated from OCP to EKS, and IIRC, OpenShift router is Kubernetes Ingress? Also, Kubernetes Dashboard [1] is a similar tool to OpenShift dashboard.

[1] https://kubernetes.io/docs/tasks/access-application-cluster/...

rkwz··on Show HN: We have built a benchmark platform for graph databases
The numbers look impressive compared to Neo4j!

Do you know what is the reason for these results? Is it because of C++ (MemGraph) vs Java (Neo4j)?

Any other queries where MemGraph does poorly compared to Neo4j? What are the downsides to using MemGraph?

rkwz··on Who ranks better? Memgraph vs. NetworkX PageRank
Thanks for the context.

Offtopic, does Memgraph have something similar to NetworkX's connected components [1]? Wondering what's the performance difference between both for different sizes of graphs.

[1] https://networkx.org/documentation/stable/reference/algorith...

rkwz··on Who ranks better? Memgraph vs. NetworkX PageRank
NetworkX is widely understood to be slower compared to other graph libraries - you tradeoff performance for ease of use.

How does it compare to igraph, graph-tool etc - https://graph-tool.skewed.de/performance

rkwz··on Ask HN: Favourite Lesser-Known Alg?
Big fan of all the fuzzy matching algorithms:

* Levenshtein

* Jaccard

* Cosine

* Jaro–Winkler

* Soundex

and many many more!

rkwz··on How to Find and Stop Running Queries on PostgreSQL
There's a "statement timeout" feature you'd find helpful - https://www.crunchydata.com/blog/control-runaway-postgres-qu...
rkwz··on Entity Resolution: The most common data science challenge
This is my favourite paper on the different indexing (blocking) techniques - http://users.cecs.anu.edu.au/~Peter.Christen/publications/ch...

Very easy to read and explains the tradeoffs of the algorithms

rkwz··on Neon – Serverless Postgres
What are the intended usecases for "branching" a database? Currently, I use separate databases for different environments, are branches better?
rkwz··on Streamlit Launches 1.0
Hi Adrein, thanks for the amazing framework!

I used Streamlit to build a small prototype and without any prior knowledge, I was productive within a couple of minutes! Really impressed with how simple and easy to learn you've made it.

I think Streamlit might be a good platform to build internal apps, only thing missing is a way to provide authentication. Is this something on your roadmap?

rkwz··on Injecting environment variables into static websites using Nginx
We can also create a mapping between the hostname<>configs to have env specific configs.

I've written more about it here: http://www.sheshbabu.com/posts/frontend-config-management/

rkwz··on Do you really need Redis? How to get away with just PostgreSQL
> Should you do it? Probably not. Is it good enough for a theoretical ~70% of the startups out there who really don't shuffle around too much data or need to pretend to do any hyper scaling? Maybe.

It's also useful when you want to quickly build a "good enough" version of a feature like search so you can get it in front of your users fast and iterate on their feedback. Most of the time, they'd be quite happy with the results and you don't have to spend time on something like managing Elasticsearch.

I wrote a post on how you can use postgres to add search capabilities with support for queries like

jaguar speed -car

ipad OR iphone

"chocolate chip" recipe

http://www.sheshbabu.com/posts/minimal-viable-search-using-p...

rkwz··on The architecture behind a one-person tech startup
It's definitely possible but we need to spend sometime to make sure the changes are backwards compatible and gradually rollout the destructive change.

Here's a really good guide on how to do this - https://docs.gitlab.com/ee/development/avoiding_downtime_in_...

rkwz··on The architecture behind a one-person tech startup
You can use any normal DB migration tool. For k8s, I put the app's readiness probe to false, run the migrations and then toggle the probe back to true.

Here are some migration libraries:

Go - https://github.com/golang-migrate/migrate

Node - https://github.com/salsita/node-pg-migrate

rkwz··on Show HN: Plausible – Self-Hosted Google Analytics alternative
> hash(daily_salt + website_domain + ip_address + user_agent)

That's a pretty interesting technique!

I built a privacy friendly analytics project https://github.com/sheshbabu/freshlytics and I've been wondering how to correctly count unique visitors to a website. I don't store cookies or any PII data at all so by definition, it's hard to distinguish between two different visits - are they from same person or different people?

An alternative approach is used by Simple Analytics - https://docs.simpleanalytics.com/uniques where they use referrer header to derive unique visits. They mention that they don't use IP addresses as they're considered fingerprinting.

But it looks like a hash function (whose salt gets rotated daily) strikes a good balance between fingerprinting while maintaining user privacy. Any downsides to this approach?

rkwz··on Nakabonne/ali: Generate HTTP load and plot the results in real-time
> jmeter - classic, true and tested, works well but config format is a hassle to deal with

You can use Taurus [1] to write scenarios in yaml files instead of using JMeter's JMX.

[1] https://gettaurus.org/kb/Basic1/

rkwz··on Cloudflare Web Analytics
You can try https://www.goatcounter.com if price is an issue :)
← PreviousPage 4 of 12Next →