In general, browser benchmarks seem to often favor Firefox in terms of startup and first interaction timings, and often favor Chrome when it comes to crunching large amounts of data through JavaScript.
I.e. for pages which use small amounts of JavaScript, but call into it quickly after loading, Firefox tends to come out on top. But for pages which load lots of JavaScript and then run it constantly, Chrome tends to come out on top.

We’re usually talking milliseconds-level of difference here though. So if you’re using a mobile browser or a low-power laptop, then the difference is often not measurable at all, unless the page is specifically optimized for one or the other.

There’s a bunch of extensions that allow you to switch user-agent easily, I personally use this one, it includes a list of known strings to choose between as well.

They used to also use the unreleased version 0 of shadow DOM for building the Polymer UI, which - being a Chrome-only prototype - understandably didn’t work on Firefox, and therefore instead used a really slow Javascript polyfill to render its UI.

I haven’t checked on it lately, but I imagine they must’ve changed at least that by now.

One thing you can test is to apply a Chrome user-agent on Firefox when visiting YouTube. In my personal experience that actually noticeably improves the situation.

If you’re going to post release notes for random selfhostable projects on GitHub, could you at least add the GitHub About text for the project - or the synopsis from the readme - into the post.

I’ve been looking at the rewrite of Owncloud, but unfortunately I really do need either SMB or SFTP for one of the most critical storage mounts in my setup.
I don’t particularly feel like giving Owncloud a win either, they’ve not been behaving in a particularly friendly manner for the community, and their track record with open core isn’t particularly good, so I really don’t want to end up with a decent product that then steadily mutilates itself to try and squeeze money out of me.

The Owncloud team actually had a stand at FOSDEM a couple of years back, right across from the Nextcloud team, and they really didn’t give me much confidence in the project after chatting with them. I’ve since heard that they’re apparently not going to be allowed to return again either, due to how poorly they handled it.

I’ve been hoping to find a non-PHP alternative to Nextcloud for a while, but unfortunately I’ve yet to find one which supports my base requirements for the file storage.

Due to some quirks with my setup, my backing storage consists of a mix of local folders, S3 buckets, SMB/SFTP mounts (with user credential login), and even an external WebDav server.
Nextcloud does manage such a thing phenomenally, while all the alternatives I’ve tested (including a Radicale backed by rclone mounts) tend to fall completely to pieces as soon as more than one storage backend ends up getting involved, especially when some of said backends need to be accessed with user-specific credentials.

I feel like this could go really well together with Piet.
Just imagine; an album consisting of a bunch of Velato programs with Piet code as the artwork.

If you build a linked list in C, and put the pointer to the next entry as the first element in your struct, then you only need a single variable (and two comparisons) to do sorted insertion into the list.

People really have no love for JPEG-XL - though to be fair that’s mainly Google’s fault at the moment.

Haven’t really used any proper JMAP clients - since the setup is broken anyway, so mainly just curl.

You could also just run IMAP/JMAP/SMTP as separate components, I can’t see any place in the Stalwart documentation - or in the Docker image itself - where monolith is the only option.

I haven’t tested the setup myself yet, but me and another root are planning on testing a setup of Stalwart to replace a semi-broken IMAP/JMAP setup for a computer club, keeping the SMTP as is.

Probably not what you’re looking for, but I’m going to note that Turris make some great OpenWRT routers.
Currently running theTurris Omnia, and using both Wireguard and Yggdrasil through it.

I’ve been personally using KDEs Itinerary app, but it might not be what you’re looking for

It’s great to hear that they’re not just giving up. And it’s also definitely good to hear that they’re not sticking with PHP either, that language is a true bane to modern hosting - and especially Kubernetes.

I’ll remain cautiously optimistic that they’ll be able to stay relevant, and not go hard in again on cutting away core functionality in the name of enterprise offerings - what caused the NextCloud split in the first place.

Has anything actually happened in ownClouds development?

The last I saw of them was FOSDEM a few years back, where NextCloud were handing out whitepapers and showing off their new Hub, chat, VoIP stack, group sharing system, and more. And ownCloud were sat somewhat opposite with two people and a screen showing a screenshot of a default ownCloud install, along with a big sign hanging from the ceiling saying “Join the winning team.”

https://piped.video/watch?v=SyimUCBIo6c

Lots of people instantly think of security when they look at WiFi-connected IoT devices, but oftentimes they never think of the WiFi signal itself - what with all the added communication noise and send time limitations of having lots of small devices.
Especially with regular consumer equipment, it doesn’t actually require that many devices to fully saturate a regular home router or AP.

Took this a few days ago. I’d been reading, and put my pad down to go grab a cup of tea, returned to this.
The image is downscaled quite a bit, was originally posted elsewhere and had to fit in the size limit.

I think the file upload size limit could become a problem in my case, at least in terms of posting the complete ACLs.

We’ve recently managed to come down to only ~1.4k VLANs though, and the network firewall pair for our server networks now only handles ~600 SPB services.