To be fair, the most interesting bugs that can be caught with tests always feel like “I would’ve never guessed that part of system actually depends on this one”.
My friend and I had a similar idea a few years ago, so we’ve built a prototype of a tool that converts personal data exports to a single SQLite database: https://github.com/bionic/bionic (repo includes “popular Spotify songs when I’m in transit according to Google Maps” query). Unfortunately, we haven’t found ourselves actually using the aggregated data: we’ve looked on it a few times, but it didn’t end up solving some real pain. It was fun to build though!
I've had a few uses for PDPL so far and I'm trying to add functionality when I need it rather than taking a shortcut and just writing the script I need in $LANGUAGE. But, like you said, it was fun to build and I learned a lot so if it's not useful to me or anyone else, that's OK too!
I’ve recently listened to a talk from Jane Street about their review system: they use this method exactly with a helper tool that keeps track of what you’ve already reviewed in the PR, minus any changes that were committed after review. PR author also “reviews” the branch, so the code review comments and changes would show up on self-re-review (because they were not reviewed).
It's very hard to understand what Foundry actually is from the provided landing page. Would you be able to describe it? Is it like Retool, but with more data sources and blocks? Can you actually drag-n-drop new applications in it or it's customized by Palantir only?
If you are interested in tinkering with personal data exports using SQL, my friend and I made Bionic: https://github.com/bionic/bionic.
The README includes an example of calculating songs you often listen to while walking/driving/using public transit (by combining Google Maps and Spotify data).
The problem with these monetary estimates is that they involve current valuations, which are probabilistically based on future outcomes, not enterprise value.
PG once formalized YC value prop as average outcome multiplier [1]. Whatever the current valuation, would YC improve your expected outcome over lifetime of the company on 7.5%? If yes, then you should do it.
interesting framing. (and awfully convenient for yc/pg, haha, but fine)
this feels like a second order analogue to the "ideas vs execution" debate - we know that idea is worth ~0, execution worth ~100 - but the harder question is - what is YC-fueled 3 months (where lets say you spend max 24 hours with YC partners/peers/events) + say 1-2 years worth of residual relevant connections worth vs a ~10 year hard grind on your own?
not gonna get the answer here but was fun to contemplate :)
very kind. the silicon valleyite answer is annie duke's thinking in bets. get the book or crank up any of the 3000 podcast interviews she's done.
the realer answer is have about ~16 years of being in love with/thinking about microeconomics. maybe do some math/linear algebra (where you get very used to projecting spaces onto different dimensions), and then make a few high stakes decisions with regards to job negotiations optimizing for cash, salary, learning value, career path, etc.
but dont think for a second i have any of this figured out haha. i just write well.
Given that enough people use this heuristic, there will be companies focusing on earning karma on HN, writing comments and voting for products they are getting paid for.
While true, it is also not a given that you'd need to trust _all_ of HN. I visit sufficiently regularly that I see some people post and recognize their username. I often think -- I wish I could _follow_ them. Not so much of a stretch to think of rings of trust built around particular users. Bringing people in and kicking them out of these trust circles could play a role. PageRank -> TrustRank? Of course it would also be only one metric, among many possible trust rankings and many possible other signals and settings.
I bet the (niche) product as such wouldn't be as hard to build as it would be to scale. Imagine every user constantly tweaking (directly or indirectly) their search result settings, and having that impact millions (or more) indexed items, for every user.
Wouldn't the world become a little bit better if they did? Earning karma on HN is not the easiest thing. I would hate to speak for all of us but collectively, don't you think we have a pretty good marketspeak alert system here?
I think you are kind of ignoring the issue that parent brought up.
The hn machine is not all that smart and can easily be gamed, and would, if the stakes were to become high enough. Farming hn karma by making pointed statements on crowd favorites (parent named two, oss licensing or privacy also spring to mind) gets you your 10k, no originality or honesty required, in no time.
What's protecting hn is a lot of moderation + relative irrelevance. If those 10k were to systematically bring you enough eyes (by driving search results), you are in effect printing money. There is no reason to assume the number of people doing it would not scale with the return attached to doing it.
Anything (points/karma/coins) which is free & unlimited will find it's way to get exploited. "What if" there's a barter system for karma/points, you do a +1 and get a -1.
Though I do not know how the initial allotment of karma/points could be distributed for the pioneers and for the new growing community, maybe allot 'n' points for new user after a year...
Trust can work both ways. They'd have to be really careful to not lose the earned trust. On the way, they would have to write a lot of high quality HN posts. I'd say this model is a win over current spamming and fake reviewing practices.
The other side of them having to write a lot of high quality HN posts is me having to read and evaluate more HN posts trying to game me. That is work, and if I sense a lot of it (like I did on Reddit), I will leave, and I suspect others would too.
Playing defense is exhausting when playing offense is extremely cheap.
I believe Hacker News’s backend is written on a Lisp (or, more precisely, Arc). In this context, JavaScript, Python and C are being debated for a role of client-side language handling DOM events.
Also, C and Lisp are different breeds. It’s wrong to put them in one bucket of “greybeard” languages. Some Lisps gained traction in the front end space (see ClojureScript).
Hacker News does have a bit of frontend, written in JavaScript. You can see some here https://news.ycombinator.com/hn.js, I'm unaware if it's everything or just a part.
reply