HNHacker News
TopNewBestAskShowJobs

mattbillenstein

1,783 karma · joined April 5, 2011

matt (at) vazor (dot) com
submissionscomments
mattbillenstein··on Ask HN: Spending Tracking Tools
Do not trust Plaid or services that use it - it stores your login credentials directly...

I use spreadsheets - download csv from your bank, import, label categories, then you can look at it any way you like. Doesn't take that much time, and with a few heuristics, the tagging can be mostly automated...

mattbillenstein··on The Antikythera mechanism – 254:19 ratio
Excellent videos - and the one thing that sticks with me was the speculation that watchmaking and the processes that it requires eventually leads to the type of technology we have today. I think the line goes if the greeks of that era had been allowed to progress another 300-400 years, they may have been able to land on the moon...
mattbillenstein··on From where I left
Well written and thoughtful piece - I don't have problems with the license - they control the project, they can do what they want, I can use it or not as I want.

The thing that's interesting to me is the idea that these things have to continue to "evolve." Like we need to keep adding features to keep people engaged to have an active community, yada yada.

In my mind, and maybe I'm wrong, some of these projects are just sorta finished - they've evolved to a completed state more or less and just need maintenance. Memcached has sorta went this route afaict - it's still tremendously useful for some things and it's not constantly trying to add data types to compete with other things.

mattbillenstein··on Advent of Code 2024
Aiming to get all the stars this year to round it out with 500 total - all the years, all the problems.

As of last week there were something around 1024 people who had all 450 stars.

Only started on like day 6 of 2022, but became hooked and had some time early in 2023 to go through the previous years. Once you have a few algorithms canned, it's not too difficult and some themes repeat across years.

It's fun to brush up on stuff you don't touch all the time - actual algorithms and stuff.

Hats off to the volunteers and Eric - I aim to donate every year now - it's a great event.

mattbillenstein··on My new POWER Indigo 2
Actually no, just 32-bit intel with Linux - we were a Sun shop and Linux with EDA tools hadn't matured enough although Linux had been around for awhile.

But yeah, it was only a couple more years until amd64 came around and really made a difference.

mattbillenstein··on My learning of prefix-dev tools
Also, from a hot pip cache, on Ubuntu installing lighting takes ~45s and the venv is 5.2G. On an m1 mac, this takes ~17s and is 400MB.

Torch, nvidia, and triton site pkgs are quite a bit bigger on Ubuntu since I guess nvidia gpus aren't available on mac.

mattbillenstein··on My learning of prefix-dev tools
Haven't the issues conda et al were trying to solve been solved with wheels and pip?

I've been supporting pytorch + cuda just using 'python3 -m venv foo' and pip (+ pip-compile) for the last year and it works - not super fast, but I prefer using the python.org blessed tooling.

mattbillenstein··on My new POWER Indigo 2
Yeah, LSI had their own standard cell libraries up to around 180nm, then they didn't want to invest in fabs and started working with TSMC - beginning of the end for them really. I signed off a couple chips on both sides of that transition.

I worked in a group that did physical design, timing closure, test insertion, etc. I did a lot of layout automation in Avanti's tools; sometime after I left, it all went to Cadence I believe.

mattbillenstein··on My new POWER Indigo 2
Nice, that was before my time - we were basically a Sun shop and actually did a lot of chips for them (and SGI) too. I never saw any of the HP, IBM, or other systems.
mattbillenstein··on My new POWER Indigo 2
And now flipping through the pictures of the internals - I ended up working for LSI Logic in the early 2000's, so it's fun to see all these LSI chips although they were before my time. Certainly some of my older colleagues worked on these.

You'll notice the numbers on these L1Axxxx - this is an internal LSI number, after they ran out of numbers for the xxxx part they bumped it to L2Axxxx and I worked on a few chips with those designations.

mattbillenstein··on My new POWER Indigo 2
I interviewed at Intel with the old Alpha group after all the acquisitions back in the day - around 2000 - and I remember talking with the guys there in that they couldn't get their EDA tools ported to the Alpha architecture; so they were designing this amazing cpu, but had to do it on "gasping and wheezing" Sparc systems. Good times.

We had the same issues where I ended up working; it was a year or two before 32-bit intel systems started to show up and they absolutely screamed compared to Sparc, but couldn't handle really big jobs. When the amd64 stuff started to come around, that's when you could just see the writing on the wall - Intel / AMD were gonna absolutely kill Sun...

mattbillenstein··on Today is Ubuntu's 20th Anniversary
I've been using LTS releases for just about everything since 10.04. It's been very stable; the transition through all the init systems wasn't that fun (sysv, upstart, systemd), but it's mostly just always worked. And it's sorta the default for 3rd party software to support first. Congrats to the team.
mattbillenstein··on Against service layers in Django (2020)
Yeah, I'd think it's probably a really bad fit for Django. I'm doing it with SQL + psycopg2, so tired of fighting the ORM.
mattbillenstein··on Against service layers in Django (2020)
I've been using this service pattern although not in Django - in Flask. The main benefit is a lot of code doesn't need to depend on a request context. We can happily have backend scripts calling services and we intentionally don't let the request context bleed into that layer.

Also, may just be me, but I've mostly seen just bad Django apps - God object patterns galore, the schema/orm/business logic all being very tightly coupled and essentially nothing is able to be refactored because of this. They often have to be simply thrown away and you start over.

But, I think having an explicit service interface solves a lot of these problems - as long as you maintain the interface and pass back the same type of object, you're free to talk to whatever backing store you like. And you get a lot less issues with circular imports and whatnot in my experience. ymmv

mattbillenstein··on Does your startup need complex cloud infrastructure?
Sure, but again, complexity - stuff people have to learn/maintain/upgrade, etc. ymmv

Running and configuring VMs isn't hard to do correctly, it just takes discipline to never "hack it in the moment" - or if you do, can that change in your config system.

mattbillenstein··on Does your startup need complex cloud infrastructure?
I have a custom deployment system which idempotently configures an Ubuntu LTS VM. All the config templates are checked into source control. I don't configure anything by hand - it's either handled in this thing or via a small user-data script run at provisioning time.
mattbillenstein··on Does your startup need complex cloud infrastructure?
Yeah, I've used the github ssh key thing before, but never heard of key leases - will take a look. Thx!
mattbillenstein··on Does your startup need complex cloud infrastructure?
True, I use it mainly for a few convenience things - holding ephemeral monitoring data, distributed locks, redis streams for some pub/sub stuff, sorted sets can be handy - things I could do in Postgres, but are a bit simpler in Redis.
mattbillenstein··on Does your startup need complex cloud infrastructure?
Basically doing this for a small startup - there are some complexities around autoscaling task queues with gpus and whatnot, but the heart of it is on a single VM (nginx, webapp, postgres, redis). We're b2b, so there's very little traffic anyway.

The additional benefit is devs can run all the same stuff on a Linux laptop (or Linux VM on some other platform) - and everyone can have their own VM in the cloud if they like to demo or test stuff using all the same setup. Bootstrapping a new system is checking in their ssh key and running a shell script.

Easy to debug, not complex or expensive, and we could vertically scale it all quite a ways before needing to scale horizontally. It's not for everyone, but seed stage and earlier - totally appropriate imo.

mattbillenstein··on Finding and optimizing N+1 queries on a relational database
You either spend your time learning SQL or debugging an ORM - I prefer the former...
mattbillenstein··on Better Dotfiles
Clever, but doesn't work for directories.

I just have a small shell script which symlinks everything into my home dir - also serves as a reference as to what actually gets put where.

mattbillenstein··on The poor man's uptime alerts
Wrote a similar curl -> email thing in python: https://github.com/mattbillenstein/pingthing

I've been putting an http endpoint in my sites that checks all the backing stores and returns a 204 on everything ok, and a 200 + error message on something broken. This is a remarkably simple and effective way to do entire site monitoring given the called endpoint can check server stats, task queue depth, backing store health, anything really.

mattbillenstein··on The 1 Hour per Year Bug, but only in Pacific time
Clocks roll back at 2am Pacific time just like everywhere else - so the author's explanation seems unsatisfying although I've encountered this type of bug before.
mattbillenstein··on 13ft – A site similar to 12ft.io but self-hosted
Counterpoint - if you like the content enough to go through this - just pay for it. Monetary support of journalism or content you like is a great way to encourage more of it.
mattbillenstein··on Ask HN: Should we bring software dev in-house?
It all hinges on hiring the right people for the job - you may want to find someone very senior who perhaps doesn't want to build it themselves, but can help you evaluate tech talent - help you hire the first few key senior people to get started.

And I'd say, if whoever you hire can get the database schema nailed down more or less correctly in the first iteration or two - you'll be good. It all hinges on finding someone who's done this sorta thing before, won't over complicate it, and can wring out the first level of mistakes before implementing an entire software system around it. Knowing the business in and out would definitely help here.

mattbillenstein··on Ask HN: Services to sync directory without central server?
Syncthing
mattbillenstein··on Ask HN: Hosting email server for our users?
I use mail in a box [1] for my personal email - I don't know your scale but might be a place to start - they use Roundcube Webmail and support imap / etc.

1. https://mailinabox.email/

mattbillenstein··on Trump injured but ‘fine’ after attempted assassination at rally
Believe I heard it in the CNN feed, I don't have a link.
mattbillenstein··on Ask HN: Which folder/directory do you store code in?
~/src/<org-or-project>/<repo>
mattbillenstein··on Ask HN: Laptop or Desktop for Development?
Desktop PC with big screens; generally better Linux support for the hardware, less thermal throttling, better GPUs, etc. And a Framework 13 with Linux for travel/work.

All depends on how you work and what you're doing really.

← PreviousPage 4 of 31Next →