HNHacker News
TopNewBestAskShowJobs

mattbillenstein

1,783 karma · joined April 5, 2011

matt (at) vazor (dot) com
submissionscomments
mattbillenstein··on Load Balancing
What a great post!
mattbillenstein··on Transpile Python to MongoDB Aggregations
Hmm, yeah, there is probably some C++ code in the mongodb source tree you could use to parse the aggregations. This whole app needs a refactor anyway - probably just live with it until then.
mattbillenstein··on Transpile Python to MongoDB Aggregations
Is it possible to reliably do this in reverse?

A friend of mine built an app using aggregations - they are hard to work with and understand imho, I'd rather convert them to simpler Python operations. Fetch some rows, join them to some rows from another table, etc.

mattbillenstein··on Show HN: Data Ingestion / Transformation SaaS
Hi!

I've done mostly consulting for 3rd parties under that Hexcloud name, I haven't released a product before using it.

Yeah, the space is crowded, I think with a lot of these products you end up in the enterprise sales pipeline in some sort of subscription. I'm aiming to keep it self-serve and simple, aimed at engineers, and api-driven.

And I'm still studying the market and surface area - I have a lot of work to do re positioning and what problems this solves. I can't say I know exactly where I'll end up with it yet; very much in the conception stage.

mattbillenstein··on Making Python faster with Rust
Kudos - this is a very nice blog post - "real" engineering ;)
mattbillenstein··on Ban 1+N in Django
Now you have like three problems instead of one - N+1 queries in the cold-cache case is slow, cache invalidation when something changes, and much more overall complexity...
mattbillenstein··on Ban 1+N in Django
I agree it depends, but my hot take is almost universally the time people like to say ORMs save them they end up paying back in spades debugging them. Learn SQL!
mattbillenstein··on Ban 1+N in Django
You're on the right side of the bell-curve meme my friend, but there are a lot more people in the thick-framework camp who spend their days getting lost in the complexity of ORMs and related tech...
mattbillenstein··on GitHub's User Content certificate has expired
I built a free monitoring service some years ago if anyone doesn't want to be the victim of this...

https://ismycertexpired.com/check?domain=objects.githubuserc...

mattbillenstein··on Expired SSL Certificate at GitHub
I built a free monitoring service some years ago if anyone doesn't want to be the victim of this...

https://ismycertexpired.com/check?domain=objects.githubuserc...

mattbillenstein··on The simplicity of single-file Golang deployments
So can I just start another process with SO_REUSEADDR and gracefully shutdown the old process?

The master/worker thing that nginx / gunicorn et al do is pretty neat, but relies on signals; so seems pretty messy and error prone to write yourself.

mattbillenstein··on Joint statement by the Department of the Treasury, Federal Reserve, and FDIC
What is the shortfall once all the assets of the bank are sold though? It's not like anyone is putting $150B into this to make sure those deposits are made whole.
mattbillenstein··on Python 3.11 is much faster, but Pyston and PyPy still show advantages
Great gains for Python 3.11 - I think there are supposed to be big things in 3.12 as well.
mattbillenstein··on Ask HN: What HTML template engines do you use?
Re Python, I like Mako. Jinja2 essentially has a python-like DSL, but I already know python, so it just seemed like a hassle to learn another thing.
mattbillenstein··on Reliability: It’s not great
Yeah, point taken, I wasn't thinking to cluster across the WAN - more like an api wrapping postgres in a single DC. But you pay the price of read latency I guess... it's a hard problem no doubt.
mattbillenstein··on Reliability: It’s not great
Do you have a guide in mind?

If it's sorting and sifting and clicking a bunch of stuff in the console, that's not painfully simple. If it's some easy cli commands, I think that's in the ballpark...

mattbillenstein··on Reliability: It’s not great
Yeah, distributed systems at the global scale are very very difficult - at least with the Heroku style problem, you'd be looking at scaling in a single datacenter I think - deployments to multiple datacenters wouldn't share dependencies.

I do wonder however if they'd be better off using less l33t tech - do almost everything on Postgres vs consul and vault, etc. Scaling, failover, consistency, etc is a more well-known problem and there are a lot of people who've ran other DBs at tremendous scale than the alternatives.

Simplicity is the key to reliability, but this isn't a simple product, so idk.

mattbillenstein··on LastPass says DevOps engineer’s hacked computer led to security breach in 2022
Why anyone ever trusted any of these products is beside me - it's a huge SPOF and you're literally betting your business they don't get compromised...
mattbillenstein··on Tell HN: macOS is degrading fast, and GNU/Linux is now better for most uses
Yeah, agreed - Thinkpad X1 Carbon and Ubuntu 22.04 LTS. The only thing I don't do is ever try to hibernate - I'm not even sure that doesn't work, I just turn it off when I need to pack it away.
mattbillenstein··on Sex, food or water? How mice decide
If you want a dose of odd - this talk at Odd Salon from a few years ago on the Mouse Utopia experiments is pretty entertaining: https://www.youtube.com/watch?v=CG2Xh2JCoMY
mattbillenstein··on Is America experiencing mass psychosis? (2021)
This is why the machines will eventually replace us - humans are too flawed - this need for "religion" is a serious defect imho.
mattbillenstein··on There Are Only Four Billion Floats–So Test Them All (2014)
Not often you can employ exhaustive testing, but it's useful in some cases.

I did a project in college for a VLSI minor where we implemented a 16bit multiplier - after we got the chips back we were supposed to do scan testing, but I just wrote up a little program in an fpga, wired up my chip to test, and swept all the possible inputs verifying the output. I think I could only run this at 10MHz, and you had to shift in the operands one bit at a time, but it still only took on the order of an hour or two to complete.

I also verified this setup by capturing some cycles on a logic analyzer - then pulled together a little awk script to turn a dump of that into the actual numbers I could check. Fun project.

mattbillenstein··on A Quick Guide to Mutt
Will give it a whirl, I also have a roundcube install on my mail server (via mail-in-a-box), so it's easy enough to just flip to a tab and read it there... Still only use mutt for composing.
mattbillenstein··on A Quick Guide to Mutt
I still use mutt mostly, but it's harder and harder as more emails come in as html - w3m only seems to get you so far...
mattbillenstein··on The Linux Upskill Challenge
I think the next level above this might be https://sadservers.com/
mattbillenstein··on SQLAlchemy 2.0 Released
This is neat - I'd never heard of it.

I've been using psycopg2 directly with a thin layer for managing connections and transactions using thread-locals for a long time - RealDictRow and some adapters for jsonb and a few other types make it really easy to work with.

mattbillenstein··on Thoughts on the Python packaging ecosystem
Maybe in the 2.4 -> 2.7 transitions - Python3 after about 3.6 is pretty good, you might see some minor package breakage if you're running the very latest version, but that's fairly rare.

Generally speaking, if you're on a modern OS - like recent Ubuntu LTS or MacOS you're getting 3.8 -> 3.10 and compatibility is very good in these releases.

mattbillenstein··on Production Twitter on one machine? 100Gbps NICs and NVMe are fast
Interesting abstract system design type problem - I think it becomes difficult if you have to shard the data though because all the assumptions about the hot set being in RAM break all the performance guarantees now I think... Which is I think basically what Twitter's existing backend has to deal with.
mattbillenstein··on Production Twitter on one machine? 100Gbps NICs and NVMe are fast
It does strike me as a feature you'd typically serve out of some sort of search index since if you had to build search, you'd essentially get indexing of hashtags "for free"
mattbillenstein··on Classifying Python virtual environment workflows
Separation of concerns...
← PreviousPage 7 of 31Next →