its no training wheel language - rust simply makes it harder to write memory correct programs by exchanging runtime bugs for compile time errors.
its way better to whinge at compile time than to crash at runtime
primarily to operate nicely with other things. rust can only guarantee memory safety if you write idiomatic rust which is great and all but incompatible with some things. hardware initialization is one, and calling into functions written in other libraries is another, including making syscalls
The idea is to delegate this stuff into tiny areas of unsafe code clearly marked as unsafe: you have to clearly indicate which code is unsafe
you’re confusing unsafe with garbage, those two aren’t always the same, and its far better to have a small fraction marked as unsafe than the whole program
yeah and it has, the community has lost its mind over a few things regarding unsafe
all languages do this. C was made for general purpose portability
you have to understand what you’re doing to do anything. C programmers have to know how to malloc and free memory because they need to know to do anything in it
maintaining an understanding of the lifetime of memory in a large C project is tedious and difficult to reason about
why? why make people care about memory semantics when we can just enforce one and only one owner of a piece of memory? That’s the philosophical question rust poses
it’s a new language who’s entire ideology is memory safety at all costs and it has bread a gross militant ideologue community who can’t behave themselves when it comes to unsafe leniency
people really want to attach themselves to an identity
Yeah I said this 3 years ago. Get your identity politics out of my software. Seems the meaning was lost on people. Unsafe leniency should not even be a question
Anyone use GLPI with or without OCS? Looking for something FOSS/self-hosted to help with general IT ops. Currently using disjointed services which are all ok on their own, but it would be great to integrate inventory into tickets and that sort of thing without having to manually patch APIs together.
I have a Synology that is sending DHCP requests on it’s default gateway interface every second. It has another DHCP interface on a different subnet that does not do this. The interface works fine otherwise, just annoying in the logs. I set it to static so nbd, but wtf?
I wonder if Synology is just insanely aggressive about its default gateway to reduce customer complaints related to stale leases that they don’t understand. Every second though seems overboard. I looked to see if there was an obscure “minimum lease” option in dhcpd.conf but no dice.
This is mostly for a drop in compatibility reason. There are also times you need to do this with a real time OS. Like breaking out of the application and setting system resources using the array of pointers function of C, which is otherwise strongly not advised.
But HUMANS are going to abuse things because it is there. You can set your environment to never run these things though.
But that was the beauty of C. I could debug anything I wanted because of that freedom of choice. I just hope with this push for better code that it doesn’t hurt us engineers on the tin
Same here, but after having gone through people’s spaghetti code, I think that I welcome our code safety enforcement overlords. Just because you can write code does not mean that you should. Most of the things that I have had to fix looks like carpentry fail posts on Reddit.
I work for a company which allocates my work to another company. Company B uses the dreadful Citrix Workspace crapware and has recently implemented something named App Protection which is kinda like an even more soulless version of anticheat you find on games. It hooks into the OS API and prevents the user from taking screenshots, recording screens and also looks into stuff like keyloggers and so on.
The problem? This breaks lots of shit. Just imagine how anti cheat companies take a lot of criticism from being intrusive and slowing system and take into account that those companies are gaming focused making products for high performance systems. Now imagine Citrix, a crappy corporate dinosaur from the 90s trying to do a similar product. Have a 4090 and game is running at 2FPS. Yeap that’s Citrix.
But hey just use it from a VM. You can pretty much do whatever you want including the stuff that shit is meant to prevent you from like screenshot. It will run blissfully without a care. It’s just a stupid feel good piece of tech meant for managers to feel fuzzy while users suffer through its bullshit. Its not enough that people who work through Citrix develop some time abstraction layer to take into account the latency of that crap.
Seriously Citrix… just die already and take your shitty early 2000s crapware tech with you.
Now let’s talk about company A. They are bread and butter MS stuff. So I needed to use lovely Microsoft Teams to speak with HR. Since I have a work VM I’ll use it. Wait… company B App Protection blocks Teams from running on the client side.
Ah well let’s install a VM specifically to communicate with company A. Simple right? So I was installing the new stuff I’ve noticed Teams Desktop mentioned something about a requirement where “company B needs to be able to manage the device”. Alright whatever let’s go with it.
It forced me to enable Hello, and restarted my computer. So after reboot it seems teams applied a bunch of GPOs on my personal computer (VM). How the hell a terrible excuse for a chat program has so much control over the OS?
This is all ridiculous. I’m not letting these creeps near any of my personal devices. I even isolated those VMs on a separate network so they don’t get some fun ideas of scanning the security of my home network.
Citrix workspace has been a thing for quite a while, I’ve noticed that its been used for years here in Canada in our health sector.
BYOD being dead is likely due to conflicts with security as companies are realizing that humans have terrible security practices and it is far better to have a slow inefficient process than to have a costly security breach.
IMO if that’s the case then just give people locked down laptops instead of applying GPOs to they personal devices.
Company A distributes surface laptops for their employees. Company B gave me some bulky Lenovo office machines which I don’t use because I live in a tinny soviet block apartment.
Connection log doesn’t matter. What they don’t want is the risk of leaking customer data or other business related information hence the protection against screenshot. But again it’s shit because it happily runs on a VM.
It’s corporate crapware of the highest magnitude and I heard the company is going down in flames.
Well more like a dedicated server that creates windows desktop instances and serves it to you over the network. Once again, it keeps all the data located on their servers to prevent leakages. If my workplace wasn’t so compute heavy, using Citrix workspace would be a dream, no more dealing with setting up your settings on your laptop.
