Hacker News .hnnew | past | comments | ask | show | jobs | submit | Calms's commentslogin

I own a few GreenPan products and started to wonder so I did some digging - seems they were subject to a class action lawsuit [1] which was voluntarily dismissed since it reached settlement[2].

Hard to know if it had much merit since it was unsuccessful though.

[1] https://www.classaction.org/news/class-action-seeks-to-debun... [2] https://truthinadvertising.org/class-action/greenpan-cookwar...


"Breaking Ground" by Philip and Freya Mulvey.

A soil scientist's take on climate change which ties together some pretty compelling points.


I'm not with Amazon and as much as I'd like to do the same, I suspect not.

The whole idea of the competition would be to use the same hardware so it's a level playing field from a hardware perspective which turns it into a software competition, right?


My understanding is that Alphastar is trained on reinforcement learning. I suspect this is unsupervised so it would have learned this behaviour independently without pro replays.


Per DeepMind's blog post[1], an agent was initially trained via supervised learning on pro matches. Then, the agent was forked repeatedly, as the population of agents learned via tournament-style self-play. So, while initial strategies could have been seeded by pro play styles, the final models were the result of models learning from games with other models.

[1] https://deepmind.com/blog/alphastar-mastering-real-time-stra...


Meaning that the final models evolved by playing against each other and they all started by using pro strategies, so, again, to me it seems kind of obvious that they would end up using pro strategies and in the best case just try to make them better


Other entities burning energy isn't a good argument for "wasting power" mining cryptocurrency as it boils down to whataboutism [0].

The proof of work is a potentially valid argument but wouldn't it be better if the work was useful in some way?

[0] https://en.wikipedia.org/wiki/Whataboutism


No, there is no impact to Synology's Hybrid RAID running on btrfs. Synology uses mdadm/lvm for RAID.


Great stuff.

Do you feel explicitly stating any licensing (or lack of) be a good idea for the samples?


I know you're asking the site creator, but I'd like to explore this matter for their benefit.

The code is hosted on GitHub which encourages authors to choose an open source license. I say "encourages" and not "requires" intentionally. See [1] for GitHub's take on it.

As an employee of a commercial software company, I'm willing to take others' advice with regard to best practices, security implications, code efficiency because I can't possibly be an expert in everything. Generally, I'll prefer someone else's solution over my own if it's well designed, documented, and simple. I'd love for the best solutions to each problem to propagate into everyone's code. But I must know the status of my ability to use your code - I must have a license. If there's no license, I simply can't use it. If there's a license, I can at least consider the code, and I can consider whether and how it affects the rest of my company's codebase.

1 - https://help.github.com/articles/open-source-licensing/#what...


While the question was aimed at the repo creator, I did intend on raising the question to incite further discussion since licensing concerns can be frequently forgotten - thank you for your insights!


I need to add a license to the samples repository (license suggestions welcome). The samples are all so short and generic that I'm not sure what legal significance it would actually have, but it's worth doing anyway if only to promote sharing and contributions.


You say "the samples are all so short" but there have been cases [+] where a very few lines were found by a court to infringe a copyright. Keep in mind that the heirs or purchasers of your copyrights might not be as benevolent as you.

As for suggestions, it depends entirely on your motives. The GPL would force users of your code to license their code under GPL also (if distributed to others.) An MIT or BSD license would only require attribution (a copyright notice) in the distribution. Personally? I'd prefer you went MIT or BSD.

+ - I've worked with former employees of companies where this kind of thing happened. I know this is vague and anecdotal, but I'll see what public information I can find just in case someone Really Must Know these cases


> The samples are all so short and generic that I'm not sure what legal significance it would actually have

People reading the page also will be "not sure", and that is a problem for them, because they need to be sure they can use the code.


A public domain dedication, like that of SQLite, is the least restrictive and simplest option.


but has no legal value in France for example as you can't reject all your rights.


It has disputed legal effect in the US, too. But people (individuals and businesses) in practice don't seem to be concerned by it -- in either jurisdiction, or most others -- enough to not use software that is dedicated to the public domain like SQLite, so as a creator if your concern is signalling to people that it is okay to use it, a PD dedication seems likely to be clean.

And any license text is unlikely to have been vetted by lawyers familiar with the legal system of every foreign jurisdiction where it might be applied, so you probably have the same problem with them (and the more complex the terms are, the more likely that they have some unforeseen problem in an unconsidered jurisdiction.)


Is it possible to declare something public domain, and license it as a "fallback"?


That's the effect of the CC0 license: dedicate the work to the public domain, if that isn't possible give away as many rights as possible

https://creativecommons.org/publicdomain/zero/1.0/deed


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: