Fedora did that. There is no Perl in the base image. Surprised the hell out of me a few months ago.

Absolutely it's a thing. I made sure my daughter got her own bank account and debit card shortly after she turned 7. That makes it so easier to manage allowance and everything else. Nobody wants to deal with cash.

Money is abstract these days, and the sooner you learn to keep a mental handle on how much money you have in the bank vs. how much you're spending, the better you're set up for reality. And learning about managing money in the abstract sense is equally useful for the rare cases where you get handed hard currency.

In the EU, e-bikes are limited to 25 km/h (15.5 mph) and 250w motors, which is already a very safe limit. There are very few non-e bicyclists who go faster than 25 km/h. And you have to actually pedal for the motor to help you, which is a big mental aid that this is a bicycle and not a scooter/moped, so people behave like bicyclists.

An e-bike that can go 30 mph (48 km/h)? That's absurd. EU scooters/mopeds are limited to 45 km/h (28 mph).

Ha! Apple is actively hostile to developers and power users. Things that are trivial on Linux and BSD, are completely unsupported on macOS and/or actively prevented in the name of security.

E.g., want a minimal copy-on-write chroot or macOS container for isolated builds? Sorry, can't have that, because that might compromise system integrity - even though it wouldn't.

Apple could enable developer and power user workflows that rival Linux/BSD and keep their security, but they don't want to.

As an idea, this is great. Technically, this is not really impressive. Steam has cloud saves and automatic screenshot sharing. Valve could implement a similar feature in a week, which games could then take advantage of.

But some games don't allow saving at any time, for various "balance" reasons. And some games have Iron Man modes where you're not supposed to be able to revert to an earlier save, which this would defeat. Both are pretty bogus reasons, but they tie in to achievements - which themselves are pretty bogus, admittedly.

Point is, technically this is not hard. It's more whether games want you to. I'd say they should all want to, because artificial scarcity is silly.

What? Generics can be fully instantiated at compile-time. There's no reason to defer this to runtime. I have no clue if Go will do it that compile-time safe way, but I know C++ does and it works great.

It is very rare that I need to edit snapshots, but I still need that ability. Usually it's because some idiot, often myself, left a huge file in a folder that gets backed up nightly. With ZFS, this file is now part of the immutable backup storage until it rotates out in a year. With btrfs, I can just delete it and free up that space.

Even if one used finer-grained snapshots with ZFS, it would never be fine-grained enough unless it was on a per-file basis - and that's just not feasible.

btrfs does support immutable snapshots, but I have not so far found a use for them. I generally loathe baked-in immutability, whether in storage or programming.

I want mutable snapshots, transparent compression, and deduplication. I used to use ZFS because it support those features, but ZFS gobbles RAM and is not usable for external USB HDDs - it would just die in an unrecoverable way every ~3 months. And ZFS snapshots are not writable - there is no way to delete a file from all snapshots and actually free up the space, because the underlying snapshot is immutable.

Switched to btrfs around 7 years ago, and it's great. Writable snapshots, transparent tuneable compression, on-demand deduplication. And with compress-force, the performance quirks with large files such as databases are mostly mitigated, because changes only need to COW the 128 KiB block it is modifying.

There are certainly still features missing from btrfs. Recovering RAID1 is abysmal - you get 1 and only 1 chance to replace a failed device, and if you do it wrong you need to recreate the array. And parts of btrfs is not aware of its own COW - e.g. defrag will unshare blocks.

But even so, I use btrfs on workstations, production servers, development servers, backup servers, etc, and it's been excellent. Just remember what the workload is and set the compress-force algorithm accordingly - though these days zstd is a really good default for everything.

Machine translation disproportionately works for languages with lots of resources, with things like news sites and dictionaries that can be used as training data.

Well that's just false. If you use Rule-Based Machine Translation, then you don't need any training data. You just need a person who knows the languages and grammars well enough to encode it as rules.

You can get very reasonable general-purpose machine translation results in 3 months of work - empirically proven by the many Google Summer of Code projects Apertium has mentored.

I rather dislike the media always saying you need big data for functional machine translation, because it's just plain wrong. And we have proof.

...better mouse experience, an integrated clipboard, windows resizing, and shared folders...

Features that VirtualBox has had for many years.

So just as StumbleUpon dies, a possible replacement is brought to the fore. Interesting.

(no, mix.com is not useful)

I asked them about that 2 months ago https://blogs.msdn.microsoft.c... to which the reply was "You are right but we needed to keep it simple.".

So they know, and they forge on ahead regardless. I predict many future files will be sent to Mac users with CR EOL.

If anything, that's yet another damning indictment of the US education system.

Here in Denmark, your standardized scores coming out of secondary education (high school, et al) mean everything, and can be relied upon to do so. There are no entrance tests for universities, no essays to write, no customized applications. Your test scores represent you - and it works, because the whole (free!) public education system is good enough, from the ground up.

(Universities here do have non-standard application options for people who want to go that route, or don't qualify for first priority for any reason.)

The first step to any regulation is transparency. Who has our data? Is it accurate? What are they doing with it? Who are they selling it to? How are they securing it? Can we delete it...?

That and much more is what the EU General Data Protection Regulation mandates. Now we just need to push it to the rest of the world via trade treaties.

These are very cool advances, but they don't solve the major problem of machine learning (ML): Having lots of data.

While these approaches don't need bilingual corpora, they still need big monolingual corpora. Very few languages have those, and those that do usually also have bilingual corpora to one or more of the major world languages.

This does lower the barrier to entry significantly for those doing ML machine translation. But, if one took the resources spent on gathering and curating corpora and instead invested in rule-based systems, you could get much further in less time.