Hacker Newsnew | past | comments | ask | show | jobs | submit | shutty's commentslogin

TLDR: we didn't pick rust because we're a Java shop.


Java was the baseline for this project, yes. The post is really about how we looked at the trade-offs, not about locking in a language choice forever.

For what it’s worth, the team isn’t Java-only either: some of us have spent a lot of time in C or C++ before, and others were already very comfortable with Java. The question here was mostly whether bringing in a new production language made sense at that stage of the project.


I originally thought it was a joke, but it's not.


Surprisingly, 3GB is a real practical RAM limit for aws lambdas in 2025: you can only have more than that if you submit a support ticket. But it's not really mentioned anywhere in the docs.


https://docs.aws.amazon.com/lambda/latest/dg/gettingstarted-...

The default Lambda quota for all accounts is 10240 MB. I've never seen it below that (in recent memory, at least), even on fresh accounts not connected to a big org.

I know I routinely use 10gb of RAM for my account that's never talked to support for the related CPU allocation.


https://aws.amazon.com/about-aws/whats-new/2020/12/aws-lambd...

Is what you're talking about a new thing? Or respectfully, are you just wrong?


They probably have never requested service limit increase to unlock all of AWS.


If they had, they would know that it involves many weeks of arguing with support, of course


Justifications upon justifications, man, so glad I no longer run infra.


Building a search engine on top of Apache Lucene and S3 block storage. Think Elasticsearch done right: https://github.com/nixiesearch/nixiesearch


https://github.com/nixiesearch/nixiesearch

A Lucene-based search engine on top of S3 block storage.

Index schema is immutable (but supports migrations), so you cannot just screw up your index.

Separate indexer/searcher tiers, so heavy indexing load does not affect search latency.

And embedding/reranker local inference, so you can run the whole AI search within a single docker container.


Looks nice! What did you use to make the first diagram (one with documents on left and 3 searcher boxes on right)?


https://excalidraw.com/


Wow, I perhaps need a kubernetes cluster just for a demo:

    CONTAINER ID   IMAGE                                                    
    0f2f86615ea5   nvcr.io/ohlfw0olaadg/ea-participants/nv-ingest:24.10     
    de44122c6ddc   otel/opentelemetry-collector-contrib:0.91.0              
    02c9ab8c6901   nvcr.io/ohlfw0olaadg/ea-participants/cached:0.2.0        
    d49369334398   nvcr.io/nim/nvidia/nv-embedqa-e5-v5:1.1.0                
    508715a24998   nvcr.io/ohlfw0olaadg/ea-participants/nv-yolox-structured-images-v1:0.2.0
    5b7a174a0a85   nvcr.io/ohlfw0olaadg/ea-participants/deplot:1.0.0                                                                     
    430045f98c02   nvcr.io/ohlfw0olaadg/ea-participants/paddleocr:0.2.0                                                                  
    8e587b45821b   grafana/grafana                                                         
    aa2c0ec387e2   redis/redis-stack                                                       
    bda9a2a9c8b5   openzipkin/zipkin                                                       
    ac27e5297d57   prom/prometheus:latest


It may be least of your worries considering it requires 2x[A/H]100 80GB Ram.


You can just use k3s/rke2 and run everything on the same node.


You can run vanilla k8s on a single node too


Also, they're rolling the dice continuing to use Redis https://github.com/redis/redis/blob/21aee83abdbfe8878d8b870b...


You think there is a risk of them pivoting from this project to providing redis as a service?


It would be great to see this approach as a part of PEFT library, so you don’t need to rely on custom code


Looks like a wrapper on top of a couple of popular Huggingface libraries. All the heavy-lifting is done with TRL, Transformers and PEFT.


Xwayland is maintained by Xorg


https://youtrack.jetbrains.com/issue/JBR-3206/Native-Wayland...


Wait, I’m old enough to remember when IntelliJ used Swing. They wrote their own GUI toolkit since then?


Thanks


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: