My POV (having not used a Jolla phone) is that it’s likely to be less convenient than GrapheneOS (I do use Graphene). I understand SailfishOS has an android compatibility layer but I assume there will be random apps that don’t work well with it (Graphene isn’t free of this either). That said, I want consumers to have more options, which means there needs to be early adopters willing to accept limitations or convenience to build up the ecosystem of alternatives, so if you’re willing to do that, I think you should. I realize this is selfish in some ways, but it’s also good in the bigger picture.

I wouldn’t be surprises if the Motorola phone gets heavily delayed and eventually cancelled (I have no inside knowledge, but have been hopeful about too many things that get cancelled to be optimistic until it’s widely available). As such, I think another benefit of Jolla is it’s more likely that you’ll get one (in a reasonable time).

I’m not disagreeing with anything you’re saying and I think the articles about this are sensationalized in the impact in some ways, but I think you’re focusing too much on the type of traffic that is typically encrypted with HTTPS/TLS.

I think the bigger issue is internal networks where it is still common to run non encrypted and/or unauthenticated services. This is particularly an issue when SSID segreagation (lile guest networks) was used to mitigate this kind of issue. The AirSniff paper shows that SSID isolation in many APs can be bypassed.

Years ago i was looking for EV kits and found several people out there selling them. Idk what the current availability is, how much tech they have, or how open they are. The ones I saw were pretty low tech (lacking regen braking and such). Think accelerator pedal controls motor speed and a battery pack is about all they were. Again this was a while ago when I was looking (like 2010ish).

You could offer kits for older vehicles, but considering the cost of the kid and installation cost/effort, does it make sense to start with an older car that may have other issues coming soon?

So what’s the alternative? Start with a new car and throw out the ICE? Sure, but a bit wasteful and even more expensive than an older car or you could find an existing manufacturer (idk like Lotus) who will basically provide you the car without the ICE related components (aka a glider). ;)

I wasn’t able to do so and ended up switching to Mint in scenarios where I’d have used Ubuntu.

As long as you do pass through of the USB device (or USB host controller), it should be fine. The VM acesses it directlty without passing through a virtualized version of the device (like what normally happens with sound, network, graphics) and the VM can even DMA to it. Down side is that the hardware isn’t visible to the host anymore, so if you pass through a GPU, it’s used exclusively by the VM, not the host. If you connect a monitor to the GPU, you see the VM, not the host. So you can only do this with hardware that is intended specifically for use within the VM. Zune management sounds like an ideal use case. See IOMMU if you’re interested in some if the tech side if it.

This is my first time doing pass through and idk what other optimizations I could do or if I’ve done anything wrong/sub optimal. I’d like your input. I’ll send you a message later

For anyone curious if this is an option (running windows in a VM for audio), I’ve had good luck so far. Feels native, but I have more tests to do before I say it’s viable. Without passing through (iommu) the audio interface, it’s unusable.

I installed Linux and set up a Win 10 pro VM yesterday. I did GPU pass through (although I doubt that was necessary). I also did pass through of my audio interface (MOTU 8pre-es). Windows saw it and it showed up in the proprietary MOTU software after installing the drivers in the VM, however I couldn’t get audio to play. I then tried passing through the USB root hub (built in to mother board) that the MOTU was connected to and then it worked. It worked just as it would running on bare metal. I tried playing a couple projects in Cubase and had no audio dropouts. Cubase has a meter that shows you if you missed audio buffer deadlines and why (CPU, disk) and it didn’t, to my surprise.

Things I still want to try:

1. How low can I get my buffer size / ASIO latency?
2. Can it handle 192KHz sample rate and at what buffer size? The tests I did yesterday were 48k and 44.1k projects.
3. How does it feel (in terms of latency) when using a MIDI controller keyboard.
4. Can I do multi channel recording without dropouts?
5. Does the VM break Cubase’s audio latency compensation when recording (this determines recording latency and automatically aligns the recording to where yiud expect it tobel). I have a feeling that the VM may introduce a latency that Cubase doesn’t account for.
6. Does iLok or another license that I need fail in VM? I only used Steinberg licenser based software yesterday.
7. Probably some other stuff I’ve not thought of yet.

What I’ve not figured out is a way to sort of boot my existing windows install. I’m sure there’s a way, but idk. I know it’s possible to pass an entire disk to the VM, but my host Linux install is on the same nvme. I guess what I’d want is a way to create a virtual block device that maps the other partitions from the nvme and then pass that virtual block device to the VM. Alternatively, install Linux to a different drive, but I’d rather not buy another nvme at this time.

They also dropped support for older plugins of which I have a lot. This is a big issue for audio stuff IMO. Apple breaks backwards compatibility frequently, which has some benefits, but commercial audio plugins are expensive and updates generally aren’t free. I actually have a bunch of very old plugins that were free, but no longer supported. Many were windows only and I can still run them roughly 15-20 years later, but even the ones that were released for Mac, I have no hope of running.

If you’re doing audio work professionally, you probably keep buying updates for your plugins, so Mac is probably a good choice. I don’t even release music (I just make noise). I’m just a hobbiest (with some higher end equipment and software). There’s a lot of hobbiests who wouldn’t be able to afford the software update costs (ignoring the Apple hardware costs). Depending on the plugin libraries, it’s bigger than the Apple hardware costs. Granted there are some really good free plugins, but some of the really popular stuff isn’t.

I had some bad g.skill DDR4 last year. I assumed it was out of warranty. Thanks for the tip!

What DAW do you use? I built a new PC and was thinking of trying to run DAW (Cubase in my case) via windows VM with pass through of my audio interface. Given the timing sensitivity of audio, I’m not optimistic. Have you tried this?

I’ve used WxWidgets and Win32 API in C. I suspect OP will quickly learn why electron is popular even though it’s so bloated. That said, sounds like OP wants a light weight and cross platform option, so WxWidgets gets my vote. Granted it’s been over 10 years since I’ve used it.

I agree unless the backend server is including it in the response/response headers for some reason, which wouldn’t make a tool like this work in the general case. I thought maybe there was a Cloudflare API that would inadvertently leak the origin IP in an error response in some special case or something of that nature, but I’d assume they would have patched that rather quickly. I’m very curious if this tool ever worked and if so, how.

If you had a single specific host you were trying to find the origin server for, you could basically scan their ASN and well known data center, particularly the big cloud provider, IPs by sending requests to them with the desired host header to try to find an entry point (load balancer, reverse proxy, web server), but I don’t think that’s practical, particularly with a free API that (presumably) responded in a reasonable amount of time. The underlying API used by the linked script is no longer available, so I don’t know if it worked or response times.

Furthermore, a well configured system should ignore requests not originating from Cloudflare’s IPs (or use a tunnel) to prevent bypassing Cloudflare, although I’ve seen plenty not do this. Cloudflare even publishes the subnets you should allow. Easy to integrate that in to a cron type job, terraform, or other way to keep rules updated even though they’ve very rarely changed.

I was curious as to how it’s done unfortunately that repo won’t answer. All it’s doing is calling a separate http api that returns the IP. I looked quickly and didn’t find a repo for that other API.