The biggest symbol of a statically linked swift binary that I have, is icudt_swift65_dat with 27.98MB, so I think that's not so easy to remove (nm v3.1.2 --size-sort --radix=d|swift demangle) And I think if you strip debuginfo it will be smaller (For a statically linked program of mine: 98MB -> 56MB)
But I think for a distribution it makes more sense to link swift programs dynamically against the runtime libraries, like it's the case for e.g. the C standard library, OpenSSL etc., as you can assume they all work with the same version and are ABI-compatible.
I tested it with a nearly static build (Still links against glibc and friends): 55MB get stripped to 44MB, so not that much. 27MB of that is icudt_swift65_dat, so I guess you would have to optimise that first
With the new Foundation work going on, when you migrate to that, if you don't import `FoundationInternationalization` then you won't pull in all of ICU and it won't be bundled in
If I’m not mistaken, the embedded swift mode aims to make ICU (the 27mb file for Unicode support) optional (and thus easily removed where it isn’t needed)
The company control their private property.
The labour can only control their own labour, so they choose to use that control.
How is that not in their right?
I totally agree, they all have the right to not work there, but they don't really have a right to prevent others from delivering packages to a building.
Yes, but people still have the right to say "Pay us $X and ensure safe working conditions or we strike" Tesla can either withdraw or negotiate. Nobody's right is infringed here, it's just free market. Especially as usually the workers have it worse than companies.
You can do that with m3u8 files. This is done for example by Pluto TV. Sure it is easily circumvented by simply stripping out these URLs, especially if they are marked as ads (And removing a m3u8 tag)
But in the end it would be again an arms race of adblockers and Google, so it would change basically nothing.
Wouldn't even be checking Microsoft's server be an unnecessary connection? You could argue, that VSCode would still work, as updates are basically optional and could be triggered manually, too
It's easy to install on ArchLinux and Fedora. The build process seems to be just standard CMake, so calling it "borderline impossible" is really a stretch.
It's just a liability and risk. Imagine you have some mission-critical system written in such language, the vendor goes bankrupt or creates absolutely unacceptable terms. Then you have a problem.
Sure if something similar happened to an open-source project, you would still have to either hire engineers to work on the compiler/tooling/language or to rewrite it in a more supported language, but I would consider it a little less riskier as you aren't dependent on one vendor
No, you just have to follow the regulations. It's no "war on US companies" or anything like that. They wrote some good law (Albeit it has a few weaknesses), and enforce it now. And if you want to do business in the EU, you have to follow them.
I would say it's more a problem of the US companies, if they can't do business without violating EU regulations.
If we're being a bit generous it's easier for the EU to make these rules, because it has very little impact on it's own companies. The majority of the large companies being impacted by these laws are US based.
Similarly you could argue that the companies are US based, simply because the US have pretty terrible privacy rules which makes it easy for them to get started with a fair amount of users/targets.
The TL:DR of GDPR is "just don't drack your users without their consent". Why is that difficult for US corporations to follow? Even '90s web was by default GDPR complaint.
"Won't someone please think of the big-tech ad profits?" - You
Why should the web's profitability or lack thereof, be my problem as a user? That the SV elite can't buy platinum plating on their yachts? Are they sharing that wealth with me? Then, good riddance! My privacy is more important than your wealth.
There's been profitable SW companies and careers before user tracking became the norm. Remember when Windows came without any ads and blogs and forums had generic non-targeted ads?
Same how the big tobacco industry got kneecapped for our own health and the greater good, a similar tech industry correction is long overdue. Is this the world you want for your kids?
You must list all kinds of data processing you perform, find the appropriate legal basis (and data retention duration, etc.), make sure you only gather data you need (data minimization), know to who you transfer data, make your services secure by default, monitor for unauthorized access, and tell affected people when there is a breach. Perhaps make a risk assessment, but it depends on the processing you do.
Yes, it's work. But quite frankly, I'm cool with a law that expects anyone who processes personal data to secure their service, to properly inform people, and holds them accountable.
Personal data has a very wide definition under GDPR:
>‘personal data’ means any information relating to an identified or identifiable natural person (‘data subject’); an identifiable natural person is one who can be identified, directly or indirectly, in particular by reference to an identifier such as a name, an identification number, location data, an online identifier or to one or more factors specific to the physical, physiological, genetic, mental, economic, cultural or social identity of that natural person;
An IP address, or an email address is personal data. Even a pseudonym or a session ID is personal data. Yes, having a log for security purposes (GDPR recital 49) captures personal data (even just access dates and requested URLs may be considered to be personal data). Yes, a comment section on a blog may capture personal data.
Once again, I'm fine with all of this. But ignoring GDPR by not capturing personal data is more complex that it might seem.
Spoken like a person who never even touched it even on a small project.
No cookies... so no language preferences. That is a profile cookie.
Right to erasure. What about invoices and orders? When can it be anonymous and when is it old enough to anonymize.
Do I get away with replacing personal data random data? Do I replace references to real people with 'anonymous person'? Will my sql constraints still work?
When I restore data from backup and someone has been anonymized in the meantime, what mechanism will be used to anonymize the user after restore?
Right to data portability. How much of the database and in what format?
> Right to erasure. What about invoices and orders? When can it be anonymous and when is it old enough to anonymize.
Invoices should typically fall under the “legal obligation” legal basis (article 6(1)c). See for how long the law requires you to keep them. In my country, it's 10 years.
>Do I get away with replacing personal data random data?
Yes, see WP216.
>Do I replace references to real people with 'anonymous person'? Will my sql constraints still work?
How do you do when someone deletes their account?
>When I restore data from backup and someone has been anonymized in the meantime, what mechanism will be used to anonymize the user after restore?
It's up to you to decide.
>Right to data portability. How much of the database and in what format?
The same as for a DSAR. As for the format, it's up to you to decide, provided it is a commonly used format.
I've also implemented GDPR. Of course it's complex, it's a continent wide law, how could it not be complicated?
But I still think that "don't track your users without their informed consent" is a good summary of the intention of the law.
And I would also say that's it's only really complex to implement if you were already tracking your users and now you need to change everything. If you weren't doing that, you'd probably find it remarkably easy to implement.
For €2300 I assume you're looking at an M2 Pro model? Note that neither the M2 or M2 Pro Mac Mini currently have working display outputs[1] so no you should not. Apple changed the way the display outputs work in M2 so they're now dependant on Thunderbolt/DP alt mode support which is not implemented for any Apple silicon machine yet.
On the other hand a cheap M1 Mac Mini would make a great machine to try it out. The M1 Mac Mini is the best supported machine currently.
I don’t run Asahi on anything currently, but I do have two desktop Linux machines, an M1 Macbook, and have previously run Linux on an Intel Mac… I can see the argument for laptops based on battery life/heat/build quality, but for a desktop machine I’d need a lot of convincing to justify the price premium and risk of compatibility issues in choosing a Mac Mini over a SFF/USFF/Tiny desktop with fully supported hardware.
I’ve owned both windows and Apple computers, quite many of them, over the last 20 years. On average, the Apple machines last at least twice as long as the windows machines while still being fully usable. One could argue just based on that basic math that they are worth twice the price.
See, here's where the undefined nature of things comes in. "Comparably spec'd" needs to be conditioned on what task you're aiming for.
A "pure gold hammer" is a terrible idea and would also be terribly expensive. But asking for a "comparably spec'd" hammer presumes the absurd premise that the material of the hammer must be kept consistent regardless of its intended use just for the purpose of being comparable.
To preface, I totally understand the value proposition of Apple devices for some use cases, but it is important to recognize that they are aiming for certain workloads.
As examples:
I have one friend that runs virtualization workloads that require a lot of memory, a lot of storage, a lot of cores, but they don't really care about memory bandwidth, "having a display", or even the noise of the device. An older server with 192G of RAM, 24 cores and >8TB of storage can easily be had and upgraded within $1k, whereas a "comparable" Mac Pro costs upwards of $10k! (Of course nobody would use a Mac Pro for this workload, so the comparison is moot)
I also have friends that are digital artists. They care about having a high brightness and color accuracy display but otherwise don't do anything that taxes the computer. They also appreciate having high quality speakers and long battery life. Some of them run M1 Macbook Airs at the lowest 8G memory configuration for ~$800 (discounted new from other retailers) + a digitizer for ~$100, while the closest comparable non-Apple laptops are all premium devices upwards of $1.5k and even then they are still worse in the battery department!
As for myself, I do light dev work, virtualization, gaming, but also travel a lot and present at conferences. I use a GPD Win Max 2 for a little over $1k (early Indiegogo pricing). The closest Apple offering would be a 14" MBP, and configured as needed (32GB/2T) would be about $3800 and still be short a 4G modem and a couple of extra devices like a digitizer, game controller, and dongle for USB-A. ¯\_(ツ)_/¯ Can't win 'em all.
A micro ryzen 5600U build with really bad quality components can be half the price of a mac mini with equal CPU performance. If you upgrade all of the mac specs you can probably get a larger divide, but IMO maxed out macs don't make much sense for most people.
It's that focusing on one specific aspect of the system and compromising on everything else that produces the really big discrepancies. I tried to use PCPartPicker to spec out a rough equivalent of a maxed-out Mac Pro in terms of CPU cores, GPU performance, and RAM and SSD capacity, but still ended up at with at most a 3.5x disparity, and that's ignoring the GPU VRAM capacity limitation and features like Thunderbolt and 10GbE and assembly and warranty and support. If you want to assign $0 value to a large portion of a Mac's features then you can make it look wildly overpriced, but that's mostly an admission that it's the wrong product for your needs.
The years of the keyboard issues left a bad taste in my mouth, but I switched to a non mac laptop for my previous laptop and now I'm back again. The coupling of the OS and hardware really do make for a great user experience. I don't want to play games on my laptop, which is the only real use case where I hear valid complaints. I just need my dev environment and snappy research and communication.
A valid complaint from me is linux based container resource utilization. The only really good fix for that IMO is if apple did something like WSL2 or FreeBSD's linux ABI and had an efficient compatibility layer. For now I just run dev containers on my (linux) desktop.
Yeah, the M1 MBA was really so out of the norm in terms of value that it's pretty much impossible to find anything like it at the price point.
It literally turned me into an Apple person, essentially overnight. I no longer even switch on my windows machines.
There do not exist "equivalent"s to any Apple devices (I don't see them licensing the M1/2 chips to anyone else anytime soon). But depending on what you care about, a "comparable" Apple device could be 10x more expensive. Of course, for other tasks a "comparable" Apple device can also be _cheaper_ than any non-Apple device available!
Only looking at aiming for similar "longevity" (since the parent is using that as a benchmark), there are plenty of devices that have a useful life comparable to Apple devices at 1/4 - 1/10 the price.
There is one above in the server/homelab space. Items like memory and storage are charged huge markups* so if you need a lot of capacity of those you are going to quickly get into the 10x range!
As for longevity, if you consider software support ending as EoL, software/OS support for a huge swath of Intel iMacs (especially those with DGPUs) was dropped by Apple quite a few OS releases ago and you have to run community patches to keep them working. Whereas similar decade old hardware still run Win 10 and Linux out of the box.
*: Don't get me wrong though, the markups are for good reason. x86 platforms don't offer anything close to Apple's ARM chip memory bandwidth (which are closer to DGPU levels). Similarly, for flash/SSDs.
My M1 MacBook Air (8GB RAM) was cheaper than the 16GB Dell Latitude 13 i7 I bought a year and a half previously. I am much happier with my MBA, even at 8GB, than I ever was with my Dell.
You can get a used M1 mini for more or less 400€.
Get a glimpse of whats going on in your local facebook marketplace, most likely you'll come up with nice offers.
In my experience the experience for those is quite bad, as you're sharing that 8G across both the CPU and GPU...
Judging from the OP's post of 2.3k€, they're probably considering a maxed out version, which has a completely different experience since you can fully take advantage of the high memory bandwidth for hybrid tasks unlike the low-memory models where you're sharing the limited capacity.
Not the OP but I got a 13" and a larger model Mac of the x86 variety when they were still reasonably young and even though I eventually got all of the bits and pieces to work it usually pays off to wait until a somewhat larger distro supports the hardware as well. That way you benefit from a much larger crowd of testers and once they have no more issues you should be good to go.
Moneywise it was definitely worth it, both machines are still working many years later and have been pretty much trouble free after the initial bugs were ironed out.
If I was in the market for a new laptop right now I'd wait for a bit and then pull the trigger on the latest model with broad support.
I'd say getting a macbook or a macbook air would be worth it, but rather than spending that much on a mac mini I'd probably get one of those new ryzen mini-PCs like from Beelink or Minisforum. You could get something with a 7735HS 8-core chip, terabytes of diskspace and a shitload of LPDDR5 RAM for 500€ and it's as small as the mac mini.
New M2 mac minis start at $600 (8 core CPU; 8GB ram; 256GB SSD). You can probably find similarly specced x86 PCs for a comparable price, but this doesn't seem unreasonable. https://www.apple.com/shop/buy-mac/mac-mini
But I think for a distribution it makes more sense to link swift programs dynamically against the runtime libraries, like it's the case for e.g. the C standard library, OpenSSL etc., as you can assume they all work with the same version and are ABI-compatible.
I tested it with a nearly static build (Still links against glibc and friends): 55MB get stripped to 44MB, so not that much. 27MB of that is icudt_swift65_dat, so I guess you would have to optimise that first