What if they simplified it so commas were ignored and thus optional. Then they are purely for readability, allowing for whichever approach a project preferred as their standard (no commas, traditional SQL, with leading commas, with trailing conmas).
I could have missed something but I don't see anything that would be impossible to parse without commas so long as there's at least one character of whitespace separation.
Very impressive but I don't follow how the author can license it under MIT terms when they've trained it using material only licensed for research purposes.
Wouldn't research purposes be non-commercial and therefore be incompatible with MIT terms?
In terms of copyright law, it's not clear that the model weights are derivative works from the training material. At this time, I think courts would lean towards the weights not being a derivative work. And even if it is a derivative work, it could be considered transformative use and therefore not a violation of copyright. Therefore, IMHO the creator of the model can license the model weights however they want to, regardless of the material that was used to train it.
For now, anyways.
If model weights are ruled derivative works of copyrighted material, it would wipe trillions of dollars off the S&P500 overnight. Currently, copyrighted works are needed to train all the big foundational models - and there's no practical or cost-effective way to get a license for all of it.
Presuming they also combine this with foreign post abuse too? ie where one country's post office takes the mail at a knock down discount and then rely on the deal that post offices take on foreign sourced mail delivery for free, meaning the sender gets a massive discount vs sending if it were charged at anything remotely sustainable.
I would prefer something between this and current HN, simply because the number of mis-taps on HN is still annoyingly high with everything being so close on a mobile, but I agree the design here is a touch OTT for space
Yeah, I've been there too, I'm working on an approach for more readability.
I went through other HN clients and found that they were expecting multi-tap upvotes which makes things, again, same as difficult as og hn interface, it's a little bit of historic but less has less usability at the sane time.
If you'd like to include features from current HN what would they be?
I like much of it currently but I would ideally put "hide" and "flag" somewhere less accessible.
In the main view pretty much all the details are useful but I'd be fine to drop the name of the poster of an article (this never guides my attention to an article in the slightest, unlike the other details).
It would be fine to play around with layouts for points, time since post and source site, possibly even using abbreviations to save space, especially if they were positioned differently.
For instance:
19p | 7hrs | 20c
Instead of:
19 points by hadat 7 hours ago flag | hide | 20 comments
Except this is a pretty good case for using an SBC:
1. they're fairly low power and
2. the LLM speed is less of an issue because SMSs like this don't warrant an instant reply anyway so if it takes 2 or 3 minutes to generate that's fine
I wish they would build features to defend against those near-spam type comments where someone is lazy and appends their clearly unrelated issue onto an existing one that maybe shares a few keywords but nothing substantive.
It's not quite spam: there's often a real person behind it with a real issue but they need a (metaphorical) slap before they muddy the waters and disturb countless people already in the conversation.
This would be worthwhile for the individuals too - you often find a recurring pattern with all / most of their raised issues and with basic guidance they might up their game (or give up)
I could have missed something but I don't see anything that would be impossible to parse without commas so long as there's at least one character of whitespace separation.
reply