I find building out forums exceedingly fun too (which is why I've been at it for a decade). Like you, we realized that federation between forums is quite important from a communication POV, though I'm not sure if you went that direction or just used PDSes as your user store.

We ended up integrating ActivityPub and its really reinvigorated my passion for building forums again :)

Usually when someone on HN talks about building a forum out, I tell them it took me a year (3 devs) before we reached rough feature parity. Perhaps it's possible for AI assisted clones to reach this point in weeks or months rather than years.

Good luck! When you get tired of it, just tell your agent to migrate all your data to NodeBB.

Good balance of intelligence and speed.

Obviously bigger != better but I don't know what the differences are.

I'm not talking about greenfield dev, I'm talking about interfacing with an existing decade old codebase.

snort

Full stop.

What is with this obsessive need for "growth".

If you believe what you read here, the gap is closing fast.

Still not an excuse to not read every line of course...

Unit tests give me the confidence that at least those tested logic paths are unaffected.

Sometimes with older codebases one cannot assume the paths have adequate test coverage.

Yes, let's unilaterally decide that query strings are bad because one website (ab)uses query strings to load different fonts.

It's the query strings that are the problem, not the website!

jfc.

Look, I'm against utm fragments as much as the next guy, but let's not throw away a perfectly good thing because tracking is evil.

For example, if there is a code block that needs to be wrapped within another function call, it'll rewrite the entire function call and you'll just have to pray that the re-written code block wasn't subtly changed.

I _think_ so far it hasn't introduced any changes....

Sorry, you thought a prompt was a suitable replacement for a testing suite?

Interesting, must try tomorrow.

Gemma 4 26B-A4B is much quicker on my setup vs Qwen3.6-35B-A3B (by about 3x), so the thought of a 1.5 speedup is tantalizing.

Have tried draft models to limited success (the smaller 3B draft model in addition to a dense 14B Ministral model introduced too much overhead already)

Others will have a better idea since they've been messing around with local inference longer than I, but I am quite impressed with the models I have been loading on my laptop with only iGPU. As of this week I no longer feel like I am playing second fiddle with slow inference and small models. Gemma 4 (and maybe Qwen3.5, haven't tried it yet) seem to have changed the game this month!

Even with trying some absolutely shiiiiite models (I only had 16GB unified RAM at the start), I was suitably impressed that I splashed the $300 to double my RAM. I am happy that this one time cost was enough to break through to smarter models and faster inference. No ongoing cloud costs!

If one likes a model then it's capable of one-shotting entire apps.

Otherwise it's "only suitable for the most trivial tasks".

Never in between.