Ah yes, I’ve spent decades cringing when I meet a self-proclaimed or even peer-proclaimed “rockstar”, “ninja”, “guru”, “jedi”, or probably a half dozen other “cool” designations for a tech worker.
Ah yes, I’ve spent decades cringing when I meet a self-proclaimed or even peer-proclaimed “rockstar”, “ninja”, “guru”, “jedi”, or probably a half dozen other “cool” designations for a tech worker.
No, right field
Also, that election was 17 years ago…
I’ll second this. Every time they mention the Israeli casualty toll, they mention the Palestinian right at the same time.
While interviewing someone about Israeli response, they took the time to raise the question about plans for Palestinian civilians (unfortunately the answer was “we must destroy Hamas”, clearly indicating the civilians dying is perfectly fine by him).
But there is a similarity, the hype pulls in all sorts of companies to blindly add buzzwords without even knowing how it might possibly apply to their product, even if it were the perfect realization of the ideal.
Yes AI techniques obviously have utility. 90% of the spend is by companies that don’t even know what that utility might be. With that much noise, it’s hard to keep track of the value.
True, though I’m mostly invested in the kernel networking behaviors, rather than having a nicely standardized place for proxy settings so that applications have a logical place to go.
It’s a fair criticism that in userspace, proxy settings have been not standardized and also TLS certificates are similarly a bit messy.
Of course the problem is that wingetui isn’t there by default, isn’t integrated to Windows Update, no matter what, WinGetUI basically becomes yet another tray icon, alongside a half dozen other auto-updater tray icons that various vendors added since there’s no integrated facility to rely upon.
So sure, it’s a bandaid on winget, but it’s still awkward and the ecosystem is a mess. Compared to Linux where a distribution will have, in the box, an extensible central update facility maybe serving two different types of repositories (e.g. apt and snap, or dnf and flatpak).
How does that pan out when the company discontinues that product model or product line?
A family member got a fancy device with zero buttons on it, and an app that basically provides ‘+’ and ‘-’ buttons that has to go through that companies cloud service. When their internet went down, they had to unplug it because it was set wrong and without internet, it couldn’t be set.
This wasn’t some advanced capability. It didn’t require massive data or computational power. It literally could have been handled with a 7-segment LCD panel and three buttons (+/-/Power). If you buy that device now, they require you pay a monthly subscription for the privilege of being able to do that (under the excuse that they use ‘AI’ to know the right values without being told, but conveniently even the ‘+/-’ functionality is now locked to the subscription plan.
True, for some uses.
If you only need command line use, it’s fine. I personally strongly prefer the environment in, say, Linux distribution running Plasma, but if you are fine with Windows applications, then fine.
If you need GUI Linux… WSLG can kind of sort of get you there, but it sucks. So if you live with any Linux GUI application for significant periods of time, then you’ll want to strangle WSLg and it’s weird behaviors. VcXsrv can help on this front.
If you are like me and find dnf+flathub an appealing strategy for installation and update of software, you like Plasma desktop management, then Linux ‘for real’ is the way to go.
Well, I don’t think it’s anti-monopoly evidence, but instead a way to intercept a popular search phrase and control the narrative.
You search for “how to download and install linux” in google, and the very top link is the Microsoft page. And the narrative is:
-I just want to get started: Oh, use WSL, that way you are using Windows really, and just a touch of Linux
-I need to use it for real: Oh, then use Azure, you can have us set up those scary Linux instances for you and Microsoft Terminal will hook you right up to those instances
-I really really want to use it: Ok, but remember, you’ll lose access to Windows applications, so there are downsides, and also, we are going to make this hands down the scariest looking procedure of the three…
WSL may be fine for a Windows user to get some access to Linux, however for me it misses the vast majority of what I value in a desktop distribution -Better Window managers. This is subjective, but with Windows you are stuck with Microsoft implementation, and if you might like a tiling window manager, or Plasma workspaces better, well you need to run something other than Windows or OSX.
-Better networking. I can do all kinds of stuff with networking. Niche relative to most folks, but the Windows networking stack is awfully inflexible and frustrating after doing a lot of complex networking tasks in Linux
-More understanding and control over the “background” pieces. With Windows doing nothing a lot is happening and it’s not really clear what is happening where. With Linux, it can be daunting like Windows, but the pieces can be inspected more easily and things are more obvious.
-Easier “repair”. If Windows can’t fix itself, then it’s really hard to recover from a lot of scenarios. Generally speaking a Linux system has to be pretty far gone
-Easier license wrangling. Am I allowed to run another copy of Windows? Can I run a VM of it or does it have to be baremetal? Is it tied to the system I bought with it preloaded, or is it bound to my microsoft account? With most Linux distributions, this is a lot easier, the answer is “sure you can run it”.
-Better package management. If I use flatpak, dnf, apt, zypper, or snap, I can pretty much find any software I want to run and by virtue of installing in that way, it also gets updated. Microsoft has added winget, which is a step in the right direction, but the default ‘update’ flow for a lazy user still ignores all winget content, and many applications ignore all that and push their own self-updater, which is maddening.
The biggest concern, like this thread has, is that WSL sets the tone for “ok, you have enough Linux to do what you need from the comfort of the ‘obviously’ better Microsoft ecosystem” and causes people to not consider actually trying it for real.
Indeed, it’s to contain the “Linuxification” of the developer community.
Before WSL, any developer dealing with backend development almost had to install Linux to have a vaguely decent development environment to align with what they get to use on the servers. While they were dragged into that world by their requirements, they may find that the packaging and window management is actually pretty cool. There reluctance to venture out of the Windows world transforms into acceptance and perhaps even liking it.
Now with WSL, those Windows desktop users say “I just need to click a distribution in the Microsoft Store and I’m golden and don’t have to deal with that scary Linux world I don’t know yet.”.
I’ve repeatedly have people notice I’m running a Linux desktop when I’m presenting and off hand say “you know you can just run Linux under Windows, you don’t have to endure Linux anymore”. They seem to think I’m absurd for actually preferring Linux when I can get away with it.
And even when you do make such a product, there’s overwhelming marketing spend against you to make sure no one knows about your product.
It’s massively frustrating to just want to make good products but knowing that the wider business context demands that you get recurring revenue or otherwise it is imperative that you fail. Your success would fundamentally undermine rent seeking, and that’s a bigger existential threat than any other mere competitor.
Unfortunately they often force round trip to go through their Internet services, so the local app won’t work without Internet and their device will be a paperweight when they retire their support
He didn’t imply that Israel were the goodies. It’s more like both sides have people being baddies.
Also, you have a lot of innocent on both sides. That’s why both sides get called out for being baddies as they are hurting innocents. There’s a good chance that Hamas even killed some folks who have never done anything but be sympathetic to the Palestinian plight.
Terroristic is the right description, and can also be applied to some of Israel’s behavior towards Palestinians.
He didn’t tell any of the non participants any such thing… His statement obviously refers to those commiting the acts, not generalized to everyone.
That perspective does not excuse Israel’s behavior or blame any victim of Israel’s injustice, it simply points out the attacks are terroristic (meaning targeting civilians). Terrorists often have sympathetic reasons, but go about it in a way that is wrong.
And to be explicit about it, zypper, dnf, apt, flatpak all have a specific mechanism to declare repositories and one ‘update’ check will walk them all.
snap does not, and manually doing a one off is useless. AppImage also has no ‘update’ concept, but it’s a more limited use case in general, it’s a worse habit than any repository based approach.
I’ll give flatpak and snap one thing: they did make some concession to avoiding duplication, unlike docker which utterly duplicates everything.
With flatpak and snap, applications can depend on/pull in a maintained external layer. So you might want ‘gnome environment, version 43’ and other applications that want that can all share. That layer can be updated independently of dependencies. You might have two instances of the gnome layers (say 43 and 44) due to different applications declaring different versions, but it’s not too bad.
Now there is some duplication, some libraries that an app says “oh, no particular layer for this one, fine I’ll just bundle it”. But at least there’s a mechanism to not necessarily do that for everything.
I’ll add that the ‘pseudo-sandbox’ is of some dubious value, as far as I can tell the app declares how much or little sandboxing it wants and the user doesn’t really get the opportunity to consent or even know that a snap has full access versus limited access.
I’ll also say that some functionality is broken in snaps (and flatpak). For example I used KeepPassXC browser integration, and then when snap was used instead of native, it broke. A number of extensions broke and the development attitude was everyone pointing fingers everywhere else and ultimately saying “just find a ppa with a browser ok?”
I’ll second the “screw it, I give up on packaging, my app is now a monolithic flatpak, snap, appimage, or docker container” mindset of a lot of developers.
“web development” casts a wide net.
The classic imagery of someone playing with frontpage back in the day, or screwing around with html in a text editor, sure. But those folks wouldn’t call themselves web developers (there was a phase over 20 years ago where anyone that cobbled together a geocities would declare ‘web developer’ on their resume, but I haven’t seen someone do that in ages).
However, you can get in pretty deep with code running in the browser as javascript and/or wasm. Backend gives them some nested dictionary in json or protobuf and they parse, manipulate, iterate over it, sometimes making some pretty complex visualizations. Basically a ‘web developer’ is nowadays on par with any Game or GUI application developer in terms of what they might be writing. There are a few things left out of direct reach by a browser runtime, but you have access to plenty and the backend abstractions to get something in reach of HTTP are often no easier than the thing being abstracted, it’s just reframed as ‘http’.