Search results

You can scale up with HAST/CARP/LAGG instead of scaling out with something like Gluster/Ceph. The former is much simpler and requires less leg work. If you're not Google/Amazon, you probably don't need Gluster/Ceph.

Your objective is server administration right? Instead of fiddling around with X11 on a server; get acquainted with SSH and Sudo first. Don't use %wheel ALL=(ALL) ALL for your users either. Really learn Sudo. Learn about TCP/IP then go from there. Don't worry about network services yet. You...

I'm surprised we don't have our own performant, modern wifi/bluetooth stack. Considering our general network stack is second to none; this puzzles me. FreeBSD made access points and extenders would be nice. For graphics, I think the harder part would be convincing vendors to port their drivers...

Interesting. So is GDM being used in this situation? Or elogind?

The same benefits you'd get from a web GUI, except I don't have to install a plethora of extra stuff just to navigate and interact with the same information. I'd rather not install X.org, etc. on a server just to be able to better view subtrees. Akin to navigating filesystem hierarchies with...

Good work! Doesn't GNOME depend on logind? How were you able to get it working without satisfying systemd dependencies?

A great tool; but that's not an ncurses utility.

why not write a ncurses menu utility for all of this? Why require a web browser?

I've got my popcorn ready for when they break their filesystem/network code with all of this pointless rewriting. An exodus of engineering talent coming to FreeBSD in the near future?

You've endured slow hardware for so long, you've accepted how much time you're wasting. IPC core performance has improved drastically over the last decade. I'd say you can build a reasonable system for under $500 using AMD chips. 30 minutes tops would've been my limit personally for electron...

You're expecting reasonable compilation times with an electron/nodejs abomination, on a CPU that is literally 10 years old with only two cores? Wat?

I just don't see it. Only time will tell. Upstream has drank the GNOME/Systemd koolaid a bit too much. I still think Wayland is a chance for BSD to write a clean, grassroots implementation of a display server. X.org is a mess.

My point is that most users are using SwiftUI/Quartz based alternatives to major gtk2 based applications on the mac. Those individuals can only maintain gtk2/Quartz bindings for so long before RedHat fully migrates to Wayland. Who knows what they'll do then. X11/Wayland and Quartz are two...

I don't see that. A lot of various open source projects actually ditched gtk and ported their UIs to Qt. Win32/Cocoa exist in completely different worlds so they're irrelevant to the discussion of the open source desktop. A lot of engineering hours were put in to DWM and Quartz. That's fine...

It really is though. The whole GTK/GNOME/systemd/Wayland sandwich is practically control by our IBM/RedHat over lords. They have their tentacles on the entire userspace ecosystem outside the Linux kernel. That was their entire plan. So we can just forget about GTK. Now if Qt and Firefox/Chrome...

pkg on macOS would be nice. Both macports and homebrew are cumbersome to use in my experience. homebrews CLI is badly designed and it's made in Ruby.. yuck.

Maybe some brave soul could spin off Weston into an API based compositor as a drop replacement in for the swaths of existing X11 WMs on FreeBSD. Who wouldn’t want a stable API/ABI to develop on? That’d be interesting. Probably the only thing keeping Linux relevant in the enterprise. IMO It has...

WPF/SwiftUI are vastly superior to anything that exists in open source; so that's hardly a problem. Mind you, I'm referring to the consumer desktop here. RHEL thus -> Ubuntu, Oracle Linux, and OpenSUSE are progressing towards Wayland by default. Even the OpenJDK folks are inching towards...

The entire desktop driver and user space ecosystem is controlled by Linux upstream. If the GTK and Qt Projects ever decide to drop their X11 backends; we have no choice but to follow suit. This will get worse should Firefox/Chrome do the same. IBM, thus Oracle/Canonical/Suse are stewarding this...

Agreed. At some point this will be a serious issue we'll need to address. For the topic though; i'd like in-kernel SMB server (with ZFS integration and full support for NFSv4 ACLs) better zfs/vm integration (ZFS ARC+mmap/page cache) SMF (init/rc.d is old and archaic) bhyve on aarch64