Today's (Aug 18 2021) update wiped my WIFI dongle's files

I accepted the update sent out today and rebooted when asked. When my computer came back up I had no internet. I'm using a dongle from CUDY.com model RTL88x2BU My OS couldn't see it nor use it. I had to go find the original files I downloaded and recompile and install. Today's update really messed me up. I saw no way to send a report to the engineers at Zorin which is a bad omission. Very bad.

If your wifi adapter initially required a dkms build to get it working, an update can cause the need to redo the dkms build - which is something I warn users of on the forum often when they use that method.
This is noted here (The instructions within should help you get up and running again):

1 Like

Thank you much for your reply and info! , Considering that there is no other method available from the manufacturer to use this dongle this bug in updates wiping out wifi is a killer for replacing PCs on my n/w with Zorin. Sad indeed. I am glad you are warning people but as this OS is aimed at newbies the warning and solution should come before the updates do this. And from the "manufacturer".

Thanks again for your help. Much appreciated.

The problem is that the manufacturer does not respect open source. The end result is Linux Techs working on the kernel must backwards write the code for the drivers.
The linux Kernel guys are usually quite good at getting the Realtek drivers sorted out. But nothing is ever perfect- so the dkms builds act as a "patch" if you will.

What many users do is save the following in their documents as a Realtek-install.txt to reference:

git clone "https://github.com/RinCat/RTL88x2BU-Linux-Driver.git" /usr/src/rtl88x2bu-git

sed -i 's/PACKAGE_VERSION="@PKGVER@"/PACKAGE_VERSION="git"/g' /usr/src/rtl88x2bu-git/dkms.conf

dkms add -m rtl88x2bu -v git

dkms autoinstall

Thanks. I have saved the instructions and package in a folder to quickly redo if an update kills it. Unfortunately linux pointing the finger at H/W manufacturers won't solve this critical business problem that hinders desktop acceptance so another direction needs to be explored for solutions.

Well, if the Developer will not share the needed information on the drivers... I'm not sure what other direction there could be.

Impressive,
How does "/g" of your, stream-edit,
help validate dkms --version?

/g/ is for Global Placement.

1 Like

This: Linus Torvalds on why desktop Linux sucks - YouTube Sorry won't be ready for prime time until Linux solves it's own problems

By all means - you have every bit the choice at your disposal.

I could as easily post a youtube of Steve Jobs lamenting problems within Apple or Bill Gates lamenting the things wrong with Windows.
Coding and evelopment is not a one entity or a one person job. Linux and Mac, for example both use Unix-like systems. Does this mean that Mac or Linux is obligated to solve ALL possible problems within Unix-like systems? Who is responsible for doing so? If it is Mac, does this mean Linux should never use Unix-Like coding then?

If you do not want to use Linux:
Don't.
No one is forcing you.
You can use Mac. And go on their forums complaining that Mac is not ready for Prime Time.
Or you can use Windows. And join their forums and complain that Windows is not ready for Prime Time.

You are both right and wrong about this, Newby.

Beyond the desktop, Linux dominates prime time. Up, down and around, excepting the desktop, Linux powers just about everything we do in our current computing environment, from the Cloud to IoT to SaaS, and does so with remarkable stability and efficiency. The companies that recognize this (Canonical, Redhat, Microsoft, Google, Amazon and so on and so on) have changed the face of business and of computing, including the model for software development.

But the Linux desktop, I agree, is not yet ready for prime time. Although users like me (using business computers built with mainstream components, limiting ourselves to a few stable and well-designed/developed/maintained distros, and using those distros out-of-the-box with a minimum of tinkering) can go for years without CLI and without any issues, that simply isn't the case for many Linux desktop users.

There are many reasons for this. The Linux desktop environment is too fragmented, too small a user/developer base to attract high levels of support from outside, too often unsupported by hardware manufacturers outside mainstream manufacturers (e.g. Intel, AMD, Dell, and a few others) and too often dominated by developers without the rigorous professionalism that formal training and years of experience in a professional IT environment foster, and users insist on installing Linux on outdated equipment. All of us involved with Linux on the desktop know that, and I for one am not sure that the situation is going to change.

Your initial post caught my eye because I had your exact experience about a year ago, with a Cudy WU1300S AC 1300Mbps WiFi USB 3.0 adapter. The adapter worked very well, but the Realtek driver isn't in the Linux kernel because the manufacturer can't be bothered to build a Linux driver, so I had to install/reinstall a community-developed driver every time the kernel updated. Because Solus Budgie is my daily driver, and Solus is a rolling release, that became a frequent issue. That got old quick, so I cut over to a Panda dongle that used OEM drivers supplied to Linux by the manufacturer and embedded in the kernel as a result. Not a burp since.

That's life in Linux. I'm not trying to be cavalier about the issues you raised, but (as Aravisian indirectly pointed out) Linux does not support everything, because it cannot support everything, and (in my opinion anyway) shouldn't try.

Apple doesn't bother with backwards/sideways compatibility and refuses to support any hardware beyond Apple's products. While Microsoft spent huge amounts of money our the last decade or so trying to ensure backwards and sideways hardware compatibility for Windows, Microsoft, too, seems to have come to the realization that the effort to support everything is a waste of resources.

The Linux community, which is volunteer based, seems to have reached that conclusion organically, and I think that's a good thing.

I do not want to get into a long, drawn out discussion of the issues involved. That is beyond the scope of this forum. I post this for a simple, single purpose: to acknowledge that Linux on the desktop has issues, and that (again as Aravisian pointed out) Linux on the desktop is not necessarily the best solution for everyone.

1 Like