Thx! And seems that `use-application-dns.net` works for me. But I will still consider replacing Firefox, because this behavior, which is inconsistent with the settings, is unacceptable to me. However, I'm glad I learned a new thing, thx again :)
No problem. I've never see Firefox exhibit this behavior. Out of curiosity have you started in safe mode disabling all the addons and does it still do this?
Interesting. If network.trr.mode is set to 5 this should not be occurring. If there is no plan to use FF again then I suppose submitting a bug would be a waste of time. Perhaps others will run into whatever condition is causing this behavior or perhaps it would happen on mine and my blackhole routes are breaking it.
Can you confirm the full technical method you were using to see DoH traffic? i.e. Destination IP/port/protocol
Suggested capture methods:
tcpdump -p --dont-verify-checksums -i any -NNnntt -B32768 -c2000 -s0 proto 6 and 'tcp[13] == 2' and not host ${Your_Router_IP} &
tcpdump -p --dont-verify-checksums -i any -NNnntt -B32768 -c4000 -s0 proto 17 and not host ${Your_Router_IP} and 'length <256' &
Don't paste the output, just suggestions for capturing HTTPS SYN and QUIC over UDP.
For what it's worth they have added a lot more outbound crap. Even using user.js [1] there is still a lot of leaky outbound noise. Loads of connections to fastly, cloudfront and others. It has gotten worse with time. I would probably also consider using a different browser but I can not give up the addons I use in Firefox. I will just blackhole route those CDN's and see what breaks.
I think the main thing was that we were doing a migration. We operate a very large-scale infrastructure, and our engineering team is super talented. However, you can imagine this is not a very common case.
I appreciate you taking a look. Let me know if you have any feedback for us. I am the DevRel, so I hang out in the community. If you see any issues where you think you need immediate support, post that in the community, and I will respond.
Good one, and I remember that somewhere (maybe Hacker News) I saw an article said that China is spending so much money and people to do these things, build some websites and make articles to change some "views".
But reviewing code I didn't recently write is one of those boring things. genAI seems to increase the amount of unpleasant things I have to do and decrease the amount of enjoyable things.
Thank you for sharing your perspective! You have prompted me to rethink this from the user's point of view. I believe we indeed need to first address the question of "how can we know if this project is still viable."
There is a solution that's already in use in the Python ecosystem (PyPI). Python projects have a classifiers field. Development status is one of the main types of classifiers. It has 7 stages starting from 'Planning' and ending in 'Inactive'. Honestly, this solution is elegant and simple to implement.
We could an issue for code forges including github to implement it such that it is shown as project metadata along side license, language, etc. Perhaps the project can be automatically marked as inactive if the maintainer(s) don't respond every 3 months or so (of course with provision to easily revert it).
Well you might also need a mission accomplished banner but it needs to say something a little different than "mission accomplished".
It could even be something like a badge of honor for developers to strive for which needs to be earned. Like they have in industrial environments where they proudly post how many months or years they have gone without a serious injury.
Maybe people would respond positively to a prominent "safety notice", if I saw "No further defects found since July 2023" that would be more encouraging than not for a lot of projects.
Yes, but this is actually a different use case. Using cloudflared means that your domain will resolve to Cloudflare's CDN instead of the real IP address, and all connections will also go through Cloudflare's CDN rather than directly to the actual device itself.
> The lack of support on LTSC is the least baffling thing going on here but I'm open to the possibility that I'm misunderstanding something....
And yea, you're right, but Indeed, many people need to use the store on LTSC, especially after Microsoft migrated many ecosystem attempts to the store, for example Microsoft Photos and some extensions like HEIC, and now not only UWP applications can enter the store; regular applications can also do so. It actually poses a very big problem that we cannot use the store anymore, at least that's what I think.
Furthermore, it is not just LTSC 2019 that cannot be used; this means that older versions of Windows (at least 1809 or older) are also no longer able to use it. In other words, we can no longer use the store on older versions of Windows. You might say that Microsoft itself didn't intend to provide support for older versions, and yea, I agree, that's true. However, the fact is that many people use Windows largely because of its compatibility advantages. I believe everyone should at least be aware that Microsoft is not as compatible with older programs, especially its own, which is what I want to express.
As for the license, I would like to clarify that it is only to prevent the packaging scripts from being used for commercial purposes and promotion. As you can see, this repository is not specifically intended for hosting store programs, so it does NOT apply to the store programs themselves, but only to the deployment scripts :)
