Julio Merino

I'll add Raymond Chen's "The Old New Thing": https://devblogs.microsoft.com/oldnewthing/ . Helped me change my views towards Windows (for the better) many years back and it's a good read anyway. ... and while here, I may also plug my own: https://jmmv.dev/

No, but if you are doing something like rebuilding a distro's worth of packages from source from scratch, the configure step starts to dominate. I build around 550, and it takes around 6 hours on a single node. Most checks are common, so what can help is having a shared cache for all configure scripts so if you have 400 packages to rebuild, it doesn't check 400 times if you should use flock or fcntl. This approach is described here: https://jmmv.dev/2022/06/autoconf-caching.html It doesn't help that autoconf is basically abandonware, with one forlorn maintainer trying to resuscitate it, but creating major regressions with new releases: https://lwn.net/Articles/834682/

The second article has videos of more comparisons with different software on a greater variety of hardware: https://jmmv.dev/2023/06/fast-machines-slow-machines.html The point is that the good old days aren't just pure nostalgia, some parts were genuinely good compared to modern bloated software.

> I remember seeing a video on here where someone actually booted up a computer from the 2000's and showed how snappy everything was... Was this what you were referring to?: https://jmmv.dev/2023/06/fast-machines-slow-machines.html

It depends. Can Windows 3.11 be faster than Windows 11? Sure, maybe even in most cases: https://jmmv.dev/2023/06/fast-machines-slow-machines.html

For those pesky random executables there's a couple of escape hatches -- buildFHSenv and nix-ld. This is also predicated on good provenance of the executables in question. One should probably not even ldd sketchy binaries: https://jmmv.dev/2023/07/ldd-untrusted-binaries.html

> There’s also Void [...] Started[0] by NetBSDs (and Hacker News') own jmmv[1]. So if you're getting BSD vibes, now you know why. [0] https://en.wikipedia.org/wiki/Void_Linux?#History [1] https://jmmv.dev/software/netbsd.html

> Also, WSL 2 always felt like a "cheat" to me... is anyone else disappointed they went full-on VM and abandoned the original approach? Oh yes, some of us are. This was a decent look: https://jmmv.dev/2020/11/wsl-lost-potential.html HN discussion: https://news.ycombinator.com/item?id=25154300

Reminds me of this blog post comparing modern Windows on a modern system to Windows NT 3.51: https://jmmv.dev/2023/06/fast-machines-slow-machines.html And the HN thread: https://news.ycombinator.com/item?id=36503983

> https://jmmv.dev/2021/08/useless-use-of-gnu.html Tangential, but your blog software seems to have mangled the headings; it's serving eg: make $(shell …) expansion instead of make $(shell ...) expansion (note that (as is correctly written in the body) that's three pediods, not one elipsis, so mldr wouldn't be correct even on it's own, meaning there's two probably-unrelated bugs affecting it).

A lot of the GNUisms described in https://jmmv.dev/2021/08/useless-use-of-gnu.html are very useful in interactive use. Search the current directory without an explcit . - useful. Append an option to a command you just typed - heck yeah, always annoyed when commands don't support that. For scripts though, sticking to POSIX sh often makes sense, yeah. You should at least be aware if your use of Bash-isms.

AKA "config directory". See https://jmmv.dev/2020/08/config-files-vs-directories.html

> more, faster Not in the world de live in: https://jmmv.dev/2023/06/fast-machines-slow-machines.html

The article is about this: https://jmmv.dev/2023/06/fast-machines-slow-machines.html

Right. I found <https://www.ise.io/wp-content/uploads/2017/07/apple-sandbox.pdf> which agrees with you (on page 9). Guess it's down to defining "at runtime" now; the TinyScheme code (and GC) and Scheme code for the sandbox definitions are shipped to a user of macOS, and the user can invoke sandbox-exec with new sandboxing definitions (e.g. <https://jmmv.dev/2019/11/macos-sandbox-exec.html>), running TinyScheme and Scheme code. So I'd think Scheme is part of the entire sandbox system, even if not in the kernel?

Always be Quitting. This was on HackerNews some time back -- https://jmmv.dev/2021/04/always-be-quitting.html

BASIC can be quite powerful. The version on the Sinclair QL was ahead of its time, for example. For modern BASIC, you have some interesting choices such as EndBASIC, which is written in Rust and which has some very cool features: https://jmmv.dev/software/endbasic.html

The closest you're probably going to get is sandbox-exec https://jmmv.dev/2019/11/macos-sandbox-exec.html

Ah my mistake Anyways there are some more subtle caveats discussed in this blogpost I was trying to remember: https://jmmv.dev/2018/03/shell-readability-local.html

> Bash is "real language" if you treat it as such. Yes, there are plenty of spaghetti scripts out there, riddled with global variables and full of side-effects, but if you write shell in a principled manner, you can solve some pretty big problems with ease, and you can write maintainable code. Even carefully written bash code tends to be much less maintainable than code in better languages. It's just badly designed on multiple levels, like the famous post about PHP. Yes, if you're really careful you can write decent code in Malbolge - but why would you? > I wouldn't start new projects in shell, of course, but one area in which I think the shell shines is in writing integration tests for tools. More on this in a recent post I wrote: https://jmmv.dev/2023/10/unit-testing-with-shtk.html The fact that you've written something that compiles to shell scripts rather than writing shell scripts rather undermines your claim that shell is a decent language. Integration tests of tools are important, but I'd still find e.g. TCL a much better way to write them.

Aside from a centralized, trusted packge archives, with signed packages and vetting, and secure downloads and verification, one difference is that with individual install.exe's - that installer can do whatever it wants. Windows package managers are just a facade over this and just run the installer for you. An actual Package Manager _is_ the installer - they are part of the OS and know how to correctly install and fully uninstall things. Someone else in this thread wrote a nice summary of this aspect here: https://jmmv.dev/2022/03/a-year-on-windows-winget.html#uninstallation This part is mostly true with MSI's, they're just a package format like .deb or .rpm - they just don't have any of the other infrastructure around them. > It does in fact centralize the update flow, but that's only beneficial when you update everything at once, which I don't have any reason to be doing: either I want it on latest, with the only interruption being not quite yet, or I don't want it updating at all Having this all centralized is actually really nice, so I wouldn't discount that. If you want to pin a package at a particular version, you can and the package manager will leave it alone. If you don't want things updated yet, just don't run the update yet.