bshipp

It amazes me that reading PC vs Mac vs Linux arguments sound the exact same today as they did 20 years ago. PC is too bloated and intrusive, Mac is too expensive with limited software options, Linux is user-unfriendly but great for servers. Reading this thread feels like reading an IRC or Usenet argument from the late 90s.

My personal server is a 9900K with 64GB ram and normally around 35 various docker containers running on it. It's very rare to see it ever above 10% CPU utilization.

For example, I host between 5-10 minecraft servers for my kids, but they sleep while the kids aren't playing on them and only spool up when a link is requested.

A properly configured Linux box would barely need a trickle of energy when nothing is demanding its resources, so I'd guess that the estimates are likely a bit on the high side.

I'm not certain, but looking at that site it appears that the 2nd edition is the R version, whereas there is only one edition of the Python version.

It has nothing to do with me supporting those actions two centuries ago. I'm just describing history. Were the same tribes in charge of the same land for the past 10000 years, or did borders fluctuate depending on intertribal wars for resources?

Acting like the Iroquois and Algonquins (or substitute any other aboriginal tribes with adjacent borders, for that matter) didn't spend hundreds of years slaughtering each other for territory before they were both deposed by the European invaders is a little naive. So who, precisely, is supposed to get the "stolen" land back?

EDIT: just to be clear, I'm not advocating that first nation's are not due compensation or allowances for the loss of their lands, nor that slavery must be swept under a rug. On the contrary, we must look for the root causes of social unrest/upheaval and meet them head on. But you're not providing any practical courses of action that should be undertaken.

Go back far enough and you'll discover that, at some point in the past, all land was one taken through force without compensation by one conqueror or another.

Excellent post. I'd also argue that the Excel team would do the data analysis world a huge favor by eliminating the "General" datatype interpolation guesswork Excel does when pasting data or opening a CSV file. I'm convinced more data files have been ruined by uninformed users opening a CSV in Excel and then saving it, than any other method.

I enjoy using SQL but the idea of using it instead of Excel to visually explore and manipulate data makes me wince. I can do without Word or PowerPoint but Excel, despite all its data handling warts, remains an exceptionally powerful tool for quick one-off data explorations.

I prefer combining SQL with Excel by generating a query that gets most of what I want inside the database and then importing it into excel's powerquery to further explore the individual vectors.

This rural/urban divide is probably a dynamic worth exploring regardless of region. The relative anonymity of large urban areas provides very different incentives toward maintaining personal integrity than in a small rural town where reputation damage can result in real and permanent losses.

I'm not arguing rural is better than urban; on the contrary, damage caused by idle gossip can permanently ostracize a person for reasons beyond their control and result in long term economic and social damage. The dynamic of smaller groups vs large anonymous groups is one worth exploring, in my humble opinion.

Medicine, biology, and even quantum physics could be argued to not be science by your standard because they often aren't explicitly deterministic. The authors are using the scientific method to evaluate population differences. How else would one undertake this analysis?

The emoji is certainly problematic, but it wasn't the sole piece of evidence regarding the contract.

"The buyer, Kent Mickleborough, later spoke with Swift Current farmer Chris Achter on the phone and texted a picture of a contract to deliver the flax in November, adding 'please confirm flax contract.'"

If I've talked to you on the phone about the contract and then send the image with a request to confirm the flax contract, it's harder to argue that a thumbs-up only reflects receipt of the message instead of a confirmation, especially if no further response is sent by the recipient.

I've worked in agriculture all my life and a case like this doesn't surprise me in the least, nor does the inevitable responses in social media that infantalize farmers as dumb hicks who have to be protected from themselves. If you're responding to a buyer offering an $57,000 contract with an emoji you're not acting very professionally. This is nothing more than seller's remorse. You'd better believe that this farmer would be threatening lawsuits if the price for flax had fallen in the meantime and the buyer was the one arguing about the thumbs-up signal. Regardless, this shouldn't have been an issue at all if the farmer had adequately hedged his position with an offsetting buy option at the sale price that would have captured any significant price movement (outside of any basis shift).

This farmer was playing loose with risk management and wanted someone to eat his $25,000 mistake (higher $82K later crop value minus original approx $57K contract value at $17/bu for 86 tonnes).

The more I think about this, the more I'm convinced that this farmer completely forgot about the sale agreement and this argument about the emoji was a post-hoc invention by his lawyer to try and weasel out of paying the terminal $82,000. This amount of flax is likely grown on about a quarter section in Saskatchewan, or 160 acres. The average grain farm size is almost 1,800 acres, so this size would be less than 10% of the total harvested area, and many farms are double and triple that size. Easy to forget about six months after the text exchange.

If this is the first time you've interacted with a buyer/seller, absolutely you'd need clarification. But if they've been dealing together for years and the farmer has previously responded--and delivered--on a text message contract, it's a different story.

I standby my assertion that the farmer simply forgot about this exchange and was shocked to get a bill for the replacement value of the contract when he didn't show up with it. This amount of flax is likely grown on about a quarter section in Saskatchewan, or 160 acres. The average grain farm size is almost 1,800 acres, so this size would be less than 10% of the total harvested area, and many farms are double and triple that size. Easy to forget about six months after the text exchange.

Precisely. The use of the emoji was tangential to the evidence that contracts between these two parties had previously been negotiated via text messages.

> "The Court of King's Bench decision said a grain buyer with South West Terminal sent a text to farmers in March 2021 saying that the company was looking to buy 86 tonnes of flax for $17 per bushel to be delivered in the fall."

> "The buyer, Kent Mickleborough, later spoke with Swift Current farmer Chris Achter on the phone and texted a picture of a contract to deliver the flax in November, adding 'please confirm flax contract.'"

It's not like the court determined a thumbs-up emoji is now equivalent to a witnessed signature or anything.

I've worked in agriculture all my life and a case like this doesn't surprise me in the least, nor does the inevitable responses in social media that infantalize farmers as dumb hicks who have to be protected from themselves. If you're responding to a buyer offering an $57,000 contract with an emoji you're not acting very professionally. This is nothing more than seller's remorse. You'd better believe that this farmer would be threatening lawsuits if the price for flax had fallen in the meantime and the buyer was the one arguing about the thumbs-up signal.

Regardless, this shouldn't have been an issue at all if the farmer had adequately hedged his position with an offsetting buy option at the sale price that would have captured any significant price movement (outside of any basis shift).

This farmer was playing loose with risk management and wanted someone to eat his $25,000 mistake (higher $82K later crop value minus original approx $57K contract value at $17/bu for 86 tonnes).

EDIT: now that I think more about it, I'd guess this farmer completely forgot about the sale agreement and this argument about the emoji was a post-hoc invention by his lawyer to try and weasel out of paying the terminal $82,000.

SQLite can be awesome for huge data sets as I've rarely found anything that can ingest data as rapidly but, as with any database, it requires some specific tweaking to get the most out of it.

The biggest headache is the single write limitation, but that's no different than any other database which is merely hidden behind various abstractions. The solution to 90% of complaints against SQLite is to have a dedicated worker thread dealing with all writes by itself.

I usually code a pool of workers (i.e. scrapers, analysis threads) to prepare data for inserts and then hand it off for rapid bulk inserts to a single write process. SQLite can be set up for concurrent reads so it's only the writes that require this isolation.

It'd be cathartic if our kids referred to a gigantic financial blunder in the billions of dollars as pulling an "Elon".

> Why is this true? If someone is making $200k and leaving to make $350k, an employer may well be able to afford the $100k for 6 months to prevent them from immediately handing over IP, but not be able to match the $350k their new employer is offering.

Sorry, I worded that poorly. I meant companies could keep the ability to impose non-competes if they originally compensated employees who were let go.

I know it gets muddled when it comes to employees moving around for increased pay, but non-competes aren't supposed to be a mechanism for holding down salaries. They're only intended to protect vital corporate secrets that, if they were released, would be so costly to the company that they'd risk bankruptcy. The value of those secrets would likely be enough that many competitors would be hesitant to hire those employees because they might fear opening themselves up to future litigation for patent infringement or IP or whatnot.

The fact that many HR departments apply them to every single rank and file employee and don't explicitly define what constitutes a direct competitor or what explicit activities/information are protected by the NC is why they're a pretty useless legal tool in most circumstances.

If companies actually paid employees not to compete they would be able to keep them. Correct me if I'm wrong, but--at least where I live--non-competes lack legal authority (even before they were outlawed) because most companies don't provide compensation for limiting an ex-employee's ability to make a living in their field in the future.

So, it's conceivable that an ex-employee could receive 6 months severance in trade for a non-compete of that length, but I doubt it's very popular.

I agree, but if we are incorporating the opportunity cost of free-riding 3rd-party users than Reddit also needs to incorporate the opportunity cost of losing thousands of unpaid moderators who keep the site running. I can't imagine the math will be favorable.

> i can fork half of github and it wont mean much without maintenance

Well....yeah? open source doesn't mean you don't have to do anything to make it work. Many times I've forked a project, cleaned up maintenance issues so it ran on my instance, and then did a pull request so the original author could incorporate the update.

While we are blessed with many turn-key projects that require little more installation effort than following detailed instructions, FOSS is supposed to be a community effort, not something falling on the shoulders of a tiny maintenance team.

For sure, and I haven't moderated in a few years so perhaps the internal reddit tools have improved since I last used them. They used to be pretty basic and cruddy. The moderation team I was a part of relied pretty heavily on the moderator toolbox:

https://github.com/toolbox-team/reddit-moderator-toolbox

I don't know if the proposed API changes will affect this specific tool, but I doubt moderators would be nearly as up in arms if their tools were going to keep working the way they had in the past.

> the free API never enabled a moderation tool that automatically detected and removed gore or child porn

I have no idea what subreddits are using these days, but I'm familiar with spambot, that compared any submitted image against blacklists. For anyone managing a NSFW subreddit, such a tool is an absolute necessity to avoid getting sent to the penalty box by Reddit admins.

https://github.com/4pr0n/spambot

I don't know what's being used now, but when I was familiar with it, spambot was monitoring dozens of NSFW subreddits that had tens of thousands of subscribers. This is only one of thousands of such customized moderation bots that use the free API.

I'm not an aggrieved moderator, but I am someone who has moderated in the past and used the third-party tools and bots that allowed a subreddit to remain focused on a specific topic. I don't think you recognize how much random spam and troll submissions are caught by filters and cleaned up automatically in even small subreddits.

Having given up moderation ages ago, I have no dog in this fight. I'm sitting on the sidelines watching the drama with a bowl of popcorn, just like everyone else.

Apologies, I likely clicked the wrong comment to reply. I agree with you, it's impossible for Reddit to provide that missing functionality. The moderation teams on some popular subreddits are almost the size of the entire paid Reddit admin staff. There is simply no way Reddit can take over the free moderation duties currently undertaken by volunteers.

Pushshift.io used to host a complete repository for Reddit. I think there were archives on the internet archive as well. There are numerous torrents with terabytes of text content for training AIs. Perhaps they might lock it down going forward, but language training horse fled and was eaten by coyotes a decade ago.

I've only moderated small subreddits and even I know the chaos that results from not having bot-driven subreddit management tools. You may hate Automoderator when it deletes your carefully crafted post for whatever reason, but I guarantee you will hate Reddit itself when your favorite subs become overrun with illegal and off-topic spam.

The race will be between Reddit admins recognizing the degradation of the user experience and the outflow of users who don't want to visit their favorite subreddit because they keep seeing decapitated heads and underage pornography and every conversation is off topic.

If they're broke than they should have first stopped self-hosting videos and images. I can't believe their API call cost is more excessive than their video and image hosting costs.

Reddit management already did this when they began self-hosting all images and videos. It's pretty disingenuous to complain about excessive API call costs when you voluntarily decided to pay for hosting and streaming Petabytes of video and images.

Reddit was not a clear alternative to Digg, but it only took a short while for the transition to occur. it's not the consuming user base that needs to transition, but the moderators who are willing to manage subreddits and users who spend the time find and share content. If they move somewhere else because Reddit becomes too difficult to use, the rest will eventually follow.

Reddit itself isn't a terribly complex site to implement as is shown with the relatively small staff they have for a very large user base. Until relatively recently you could download the source code from Github and spin up an instance yourself. The actual "work" is being done by the unpaid moderators and contributors.

The bots enabled by the free api was what allowed Reddit to transition from a small hangout into "the front page of the internet". Without the bots and tools enabled by the free api the site will rapidly devolve into a cesspool like Facebook.