philbo

9 ms·

91.
by philbo 3y ago
Thanks for the advice! We're not tied to any model, no. Naively I guess, at first we hoped to get by using a 3rd party API. We're hosted in GCP and tried using the Vertex AI `textembedding-gecko` model initially. But now we'r
92.
by philbo 3y ago
How do you generate your vectors? I'm working on something that needs a similar, small and fast, vector search implementation. Crucially we also need fast indexing speed for our usecase, but a bottlneck we're hitting is the time i
93.
by philbo 3y ago
Answering my own question, sorry, looks like I may have got the wrong end of the stick here. Bazel and buck seem like language-agnostic build tools, rather than rust-specific package managers?
94.
by philbo 3y ago
As someone who hasn't used Rust seriously since 2018, I'm tangentially curious what are the improvements (real or perceived) that buck and bazel offer over cargo? I didn't know there were alternative package managers, and car
95.
Lessons learned from integrating with GPT in production(philbooth.me)
3 points by philbo 3y ago|0 comments
96.
by philbo 3y ago
Complete tangent but this sent me down an interesting rabbit hole reading about the Celts. I had no idea they (probably) originated in central Europe, I'd assumed what we now call the Celtic nations were the origin. https://
97.
by philbo 3y ago
Ah, that's a good point. I explicitly didn't call Redis a "footgun" at any point, because of what that word means. However the post is tagged with `footguns` and perhaps that contributed to the misunderstanding. I'l
98.
by philbo 3y ago
Thanks, I'll update the post and link to your comment for attribution.
99.
by philbo 3y ago
No. As it explains at the beginning of the paragraph you're quoting: > If you're particularly naive, like I was on one occasion, you'll exacerbate these failures with some poorly-implemented application logic. Then a few p
100.
by philbo 3y ago
> This has nothing to do with the redis server. This is bad application code monopolizing a single connection waiting for an unrelated operation. Well, yes. That is why the preceding sentence, which you didn't quote, said "poor
101.
Ways to shoot yourself in the foot with Redis(philbooth.me)
179 points by philbo 3y ago|80 comments
102.
by philbo 3y ago
This sounds right to me. If I were building a Q&A site that genuinely wanted to encourage high-quality answers, I'd implement something like a 1-hour window where all answers are invisible immediately after the question is posted.
103.
by philbo 3y ago
https://philbooth.me/blog
104.
by philbo 3y ago
Realistically, how much data is it practical to stash inside these continuations? It seems like it could get out of hand quite quickly for non-toy examples. If you were propagating e.g. game state, you'd still want to stick that in a d
105.
by philbo 3y ago
From the first blog post linked above: > We discovered evidence of fraud in papers spanning over a decade > We believe that many more Gino-authored papers contain fake data. Perhaps dozens.
106.
The Speed Project: A Race Between Los Angeles and Las Vegas(bbc.co.uk)
43 points by philbo 3y ago|40 comments
107.
by philbo 3y ago
Producing television shows also costs money, but no-one forces me to watch the ads there. I can change channel, or skip past them if recorded, or just leave the room while they're on.
108.
by philbo 3y ago
Yep, a few people have made the same comment. I'll make a correction to the post soon, sorry that it upset you. As obvious as it seemed, I had no idea. Thanks for pointing it out though, feedback from experts is the big payoff to posti
109.
by philbo 3y ago
This is a great tip, I had no idea there was `set local work_mem`. Thanks!
110.
by philbo 3y ago
Yeah, a lot of people have said this. I was wary of asserting "do's" because I'm not a Postgres expert. It felt more honest to phrase stuff in terms of my own mistakes and "don't do what I did", but of cou
111.
by philbo 3y ago
Yep, a few people have mentioned this to me here and on Reddit. I didn't know about the issues with the approach I proposed, so was pleased to read the comments. Will add a correction to the post as soon as I have a sec, thanks.
112.
by philbo 3y ago
> Also, if you're doing string concats manually for your Json, this might cause some overhead for larger objects. ?? Good point, I hadn't considered that part of it. It wasn't string concats, we were building it with `json
113.
by philbo 3y ago
> So, would the better advice not have been to use simpler SQL instead of complex recursive statements, instead of taking a drastic approach to abandon ship (move logic to a completely new layer)? Probably, yep! But I didn't know th
114.
by philbo 3y ago
Fwiw the specific case which motivated that section in the post was a set of recursive functions we used to denormalise an irregular graph structure (so not suitable for CTE) into a single blob of JSON to be sent to another data store. 99%
115.
Ways to shoot yourself in the foot with Postgres(philbooth.me)
743 points by philbo 3y ago|268 comments
116.
An example of LLM prompting for programming(martinfowler.com)
2 points by philbo 3y ago|0 comments
117.
by philbo 4y ago
> How do you evaluate 3rd party dependencies? I actually blogged my answer to that exact question recently (shameless plug): https://philbooth.me/blog/how-to-evaluate-dependencies
118.
by philbo 4y ago
Yep, this one is the headline item for me. Look at the code and, if it has further dependencies of its own, look at the code for those too. The main question I'm asking myself while looking at the code is: if I had to fork this thing a
119.
by philbo 4y ago
One of my stock interview questions asks people how they evaluate 3rd-party dependencies for use in a production environment. So many interviewees respond with GitHub stars as their main or only criterion. It depresses me every time.
120.
Technical Regress(stumblingandmumbling.typepad.com)
12 points by philbo 4y ago|0 comments

More ›