Yes, AppImage can run on more distro.

Still AppImage has disadvantages over DEB: No auto-update, No/less system integration, Bigger install packages.

They can’t possibly provide a package for every distro.

Signal’s model, ie keep tight control over development and distribution of the client, and the absence of federation, it well suited for Apple/Google’s stores, but not at all for open-source and Linux’ ecosystem.

Some projects of Signal-compatible clients and forks received a message from a Signal representrive requesting they stop distributing unofficial clients that connect to their servers.

That probably has on shilling effect on Linux distribution that may be considering building and distributing Signal in their repository.

I’m going to assume lack of volonteers.

Wrong choices happen when there’s deletion of useful historical data, motivated by short-term cost saving.

Wrong choices also happen when there’s unnecessary creation on data, such as logging and storing everything, just in case, with a verbose level.

Storage can be cheap in some cases, but high-availablility high-performance cloud storage is very expensive. Anyway, it’s not infinite.

The way to keep useful data is to be strategic and only store relevant logs. Fine tune retention policy especially for fastest growing data. Storing everything on high-cost storage, without smart retention policy, could lead to deleting git data to make place for a mix of debug logs and random shit.

Nice list. Note MicroG doesn’t appear in F-Droid’s default repository. It probably cannot be installed as a regular app, or requires adding a different repository.

NewPipe supports PeerTube and others, give it a try to lower your dependency on YouTube!

Quoting from another comment:

Nope, just install 1.19 and A-Apps that are installed should start working.

That’s good to hear! Did you had to tweak any setting to get this working?

There’s another comment where someone wonders if there’s any setting to enable this.

Great to see Taler continue to move forward. I hope a bank and/or fintech company finally adopt this. It shouldn’t stay an experiment and deserve adoption for day-to-day purchases.

Rewrite the application to be less greedy in the number of requests it submit to the server, make (better) use of caching. That’ll probably lower the number of concurrent request that have to be handled.

Automatic delay and time adjustment should be standard for consumer devices and electric car charging that consume lot of energy, and don’t need to be activated immediately.

For years and even decades, it’s been possible to setup water heaters to activate during low-rate electricity hours, and store the hot water for later.

Seconded, AGPL may be better especially in cases where software is network related.

That’s why I like the GPL. I’m fine with people using my software for personal use, or for commercial use if that’s integrated in another open source software.

If a company want to use my software with an unrestrictive opensource license, we’re going to have to discuss license price.

Even a Real Time Operating System cannot guarantee serial/network input will arrive in time.

Is this for an opensource software project, and if so can you tell more about the project?

If that’s for a work or university project, you should share salary and/or credit with whoever is going to give you a solution.

NVD state they task an analyst to review each CVE and assign a score, then do QC to review the analysis before publication.

No one’s perfect, but since NVD claim to do QC they should fix their mistakes. So now let’s see how they answer to Daniel Stenberg’s objection. The publication and objections are recent, it’s fair to give them a few days to react.

But if they’re giving up on doing proper analysis or QC, and are are just acting as a vulnerability number registry, then they shouldn’t publish CVSS values.

NVD analysts use the reference information provided with the CVE and any publicly available information at the time of analysis to associate Reference Tags, Common Vulnerability Scoring System (CVSS) v3.1, CWE, and CPE Applicability statements

CVSS V3.1 exploitability and impact metrics are assigned based on publicly available information and the guidelines of the specification.

Analysis results are given a quality assurance check by another more senior analyst prior to being published to the website and data feeds.

Source: CVEs and the NVD Process

Good point, that’s another difference between the two. Although you can probably achieve the same result with both.

Not depending on the cloud processing your data is more important in my opinion.

… and Python that actually gets executed on your machine, not someone else’s machine (ie the cloud).

The difference would probably be even more dramatic if factoring out download times.

A good DM need to strike a balance between configurability and ergonomy/ease of use.

As a novice user I loved to tweak the many configuration options, but it’s time consuming and often lead to something worse than the default, leading to further tweaking.

Now I appreciate good defaults because that means there’s few settings that needs tweaking.

Some DMs like early Gnome 3 releases went a bit too far removing configuration options, and have been slowly adding them back over several years.