noodlesUK

I think what would help here is to understand what you need to run on windows. If you're averse to using Windows as a daily driver, you might consider limiting its use to just the one class that requires that particular software, e.g., with a VM or dual booting.

What software do you need to run?

I find that the experience doing this is essentially unbearable due to graphics problems on retina displays, input lag and the like. I also figure that the kind of person who wants an editor which is designed to paint as fast as possible probably wouldn’t want to have a whole VM and spice/similar client sitting between them and the editor.

I tried out the editor because of this post: it looks very promising. Unfortunately I can't use it because it doesn't have support for remote hosts/devcontiners. That feature of VScode is critical to my workflow, as I don't actually want to program on a Mac host, but rather use my Mac as a portal to the VMs and containers I actually code on. It massively helps with segmentation of my projects and improves my security posture (by not having a development environment or dependencies on my actual host machine).

Unfortunately I couldn’t file my UK self assessment for the same reason. Only a few of the self assessment pages are supported, and anything which requires the residency or foreign pages requires commercial software. The software is dramatically cheaper than it is in the U.S. but I still take issue with it on principle.

Unfortunately it does not cover form 5471 which is required if you own an interest in a non US company (even if the company is dormant), or form 8858 which many more internationally self employed people now need to fill out as of a few years ago, as most people self employed abroad are now seen by the IRS as a “foreign branch” of a U.S. person.

I am of the opinion that each time the IRS adds a new form, they should by law provide a way to e-file it for free. Essentially no end-user tax software supports my personal tax situation, because the support for international situations is usually pretty terrible. Alternatively they could dramatically reduce the reporting burden for international people, perhaps with a cap on income or net worth, but I suspect pigs will fly before the US does that.

I don’t see how AI cameras are a solution to any crime issues in places like tube stations.

The staff at these stations know when people are jumping the turnstiles and evading fares. They know if there is a fight. It’s already recorded on classic CCTV (though with shockingly bad cameras if we are to believe recent news reports).

The issue is not in my mind the detection of crime (especially petty crime). This may or may not improve the detection of crime, with a cost to civil liberties and loads of false positives.

I don’t know if increasing detection of more minor crime is useful if there is not a clear way of being able to respond to the volumes of crime we already know about.

You can’t jail every fare jumper or shoplifter in London. It simply isn’t practical. Even if you did jail them, you’d eventually have to let them go, whereafter they very well might reoffend, as they would be more desperate than before their incarceration.

Where are you based where derisking is illegal? Here in the UK it makes the news fairly frequently when someone notable becomes unbanked because of CIFAS or SARs or similar.

Also, I wasn’t able to get a debit card until I was 12, but perhaps that has changed over the past few decades.

I strongly suspect that the reason cash is seen to be cheaper for these businesses is that they don’t report all of their cash income to their tax authority, rather than because cash is actually that much cheaper.

Alternatively, they were counting the time they spent cash handling as a business owner as zero cost, which isn’t really true.

The issue is, whilst I’m sure that cards are cheaper overall, they present a serious burden for people who can’t use them: unbanked people for various reasons, older people, children etc.

There were already costs. There were always costs and risks associated with cash handling: armored trucks doing cash collection, risks of money in tills being lost. If it really were a lot cheaper you wouldn’t see businesses clamouring to go cashless.

That’s not to say I think cash should go the way of the dodo, but I don’t think there’s that much difference paying g4s to collect your money vs Mastercard.

IMO the benefits of cash are on the consumer, not the business.

I think this makes a lot of sense - this would be sufficient for an island to have its own ISP. It probably doesn't make a ton of sense for places that are on the mainland of a significant country, but if you're an island community somewhere in the pacific or off the coast of some country, it's this or bust. It also makes sense for remote research or military outposts.

Having a big pipe that then fans out with conventional technology is almost certainly a better way of getting connectivity to a town or similar. That way the bucket of service is much more efficiently allocated. They can also do things like put Netflix caching boxes or other CDN boxes in a site like this.

I think that to most software people kernel means something like the Linux kernel, but to an ML engineer it is a specific mathematical term-of-art.

Like the other commenter I’m amused that they included rocks in the shredded notes to make up weight.

I think something like this is unfortunately doomed to fail, as whilst there might ideally be approximately 138 banknotes worth of mass in there, it seems likely that the shreds are from thousands of different banknotes, and the container holds only a fraction of each whole note. Therefore no matter how good your reconstruction is, you simply won’t have the right pieces to work with.

That is a fair counterpoint. I think then the issue is potentially misleading marketing, where a good is being marketed as robust and high-quality when it is not in fact either of those things.

How and why is a non-profit being sold to a VC firm? Why did the non-profit agree to this purchase?

I get that buying used products might be a partial solution at a personal level, but that doesn't change the fact that we're producing ever more cheap, low-quality stuff. What the market is ideally supposed to do is optimise for the tuple (cheap, quality). It doesn't seem to be doing that, as it seems like things are being optimised for (cheap, perceived quality), and just straight up lying about their quality. When people are buying stuff online, it is even easier to just focus on the price. How do we fix this?

Serious question: enshittification seems to be plaguing all parts of our consumer economy at the moment. How can we actually address this? Are there economic or regulatory steps that we can make to ensure that we aren't forced into an eternal race to the bottom where product quality just trends to zero permanently?

In terms of a privacy policy, the UK ICO has a good starting point for UK based organisations: https://ico.org.uk/for-organisations/advice-for-small-organi...

It covers the basics of GDPR compliance.

Do any of these resources target multiple different countries? The requirements for these kinds of documents tend to vary wildly, even between countries with similar legal traditions in the Anglosphere.

The M&S tills are funny. I've identified what the issue with them at my store is: the scanner doesn't beep when it sees an item, it beeps when it can no longer see an item. I'm pretty sure someone configured the scanners wrong so that it's set to trailing edge rather than leading edge.

I think the error of self-checkout was installing self-checkout terminals as a replacement for traditional checkouts as opposed to a speedy option. Particularly in grocery stores, I really enjoy scan-and-shop style self checkouts, as I am able to pack my bags as I go through the store. I can then present myself at a till, pay, and leave. That actually saves me time.

What I don't enjoy are paranoid self-checkout machines with scales that are constantly complaining about unexpected items, or self-checkout machines in locations where pretty much every customer is making an age-restricted purchase or a large portion of the items have security tags. There needs to be a sensible attempt to understand the purchase habits of the people using a particular store and tailor the checkout experience accordingly.

On a related note: I was in a Boots (similar to Walgreens) earlier today and I witnessed a truly crazy amount of shoplifting: 3 times whilst I was there. Partially that was down to the layout of this particular store, which had a number of different exits and an insufficient amount of staff. The self checkouts don't help at all, because they cause far more false positives with the exit alarm, where a security tag doesn't get removed.

It would be straightforward for an inventory system to hook up directly to the security tags, rather than having them be totally separate from the SKUs which are actually scanned into the POS. You'd then have item-level tracking of stock, which I'm sure would help from more than just a shrinkage perspective.

This doesn’t even begin to surprise me. Even on a tiny scale, myself and a couple of the other top students at our university got an unsolicited phone call from palantir about 10 min after we finished our last exams congratulating us on finishing university and encouraging us to apply to Palantir. They had clearly gotten access to our exam timetable so they could call us right after we had finished.

It was deeply creepy, and left a very unpleasant taste in my mouth.

Not really because then projects started needing multiple containers and orchestration, and might be fussy about the order the different services come up.

I can tell you that this article was written by someone who has never attempted to use Facebook’s developer tooling.

Their dev portal is constantly broken in different ways, it blocks you from making changes at random because it needs to run some kind of heuristics to see if you’re doing something malicious (even with webauthn 2fa).

We recently tried to get a Facebook app approved to use the pages_messaging permission. You need to have test users to pass verification. Unfortunately, you can no longer create test users on their platform since October of this year. If you try to use a real dummy user, it will get banned immediately when the application testers log in from another country.

Aside from that, you end up with corporate assets being owned by people’s personal Facebook accounts, because there’s no reasonable way to use SSO or similar. It’s just downright terrible.

I’ve used the offerings from all of the big tech companies, and Facebook/Metas is far and away the most hostile. You get what you pay for.

I will eat my hat if Meta successfully builds a cloud services platform for other tech companies given the state of their developer tooling in all their other business lines.

My understanding is that all sellers, including Amazon itself have co-mingled inventory. Therefore you can’t actually guarantee that what you’re getting is from Amazon’s stock, as opposed to some other random seller who gave the FBA warehouse a truck full of fake products.

I think the issue is multifold: a lot of the fake products are not dangerous, but they are either useless, or otherwise inferior. There’s essentially no recourse for this as a consumer, and you probably won’t even notice that you don’t have an original product.

In some cases the fakes are downright dangerous. This is much more the case in supplements, cosmetics, food, and occasionally electrical appliances.

People will only sue them when they get actually dangerous products. Even then it’s a difficult process.

Funnily enough AmazonBasics is probably the safest set of products on Amazon, as nobody but Amazon is allowed to sell those products. A brand name item might be a fake, but AmazonBasics is always the original (potentially crappy anyway) product.

I think with the state of things, Amazon should be completely banned from selling anything that might be construed a supplement. They have such lax controls that your protein powder or vitamin pills might contain pretty much anything. Even reputable brands bought from their brand storefront might be counterfeit [1]. You’d be shocked at the number of things that are counterfeit. A friend of mine recently bought a niche optical device, and it turned out it was a fake, despite being allegedly sold by the manufacturer. Amazon needs to stop co-mingling inventory, and it also needs to stop selling things that have no safety testing whatsoever, especially in the food and supplement space.

https://www.inc.com/sonya-mann/amazon-counterfeits-no-starch...

There are a few things that I wish I could change which would improve my personal security posture massively:

1. When advanced data protection is turned on, it should not be possible to change the Apple ID password without the associated hardware tokens.

2. Again when advanced data protection is on, it should require the hardware tokens to add a Face ID appearance or change the device passcode.