javitury

I agree with you, uber is a good 1st option. In my previous comment I wanted to remark that it's not flawless, but I think we are on the same page about this as well. It pays off to keep alternatives in mind when things seem to go sideways (request another driver, use a taxi, be more vocal ...). Also travelling makes me tired and then I just let issues slide

Regarding Uber, I agree that their price transparency is very much appreciated.

However it's not rare to find bad drivers on Uber. On Christmas this year I took an uber from the airport, the driver had supposely arrived but he was nowhere to be seen. We called each other and I could hardly hear anything. After wasting about 30 minutes (and battery almost depleted) we finally found each other. It turns out he didn't know how to speak English or the local language. He had two phones, one he used to call a colleage who could (barely) translate english for him, the other phone he used to talk to clients, and both phones were placed mic-to-speaker to bridge the calls. What about the extra time that the driver wasted? I was billed for it and I had no way to dispute it. I could only report this behavior in a review to a driver that didn't seem to be him (was the main driver subletting his account?).

Yes, you need to spin a forgejo runner

I disagree. Mercedes used to be well known for reliable and easy to repair cars among the general public, or at least that was true among the people I know. E.g. there are stories about old mercedes with millions of kilometers, some of them in remote places and owned by workers (e.g. taxi drivers). Of course, this is no longer the case and newer models are very disappointing in this regard.

Also the right of repair is orthogonal to the concept of luxury. What is relevant here is whether mercedes is leveraging their market power during the car's design phase to hinder competition in after-sales services and profit from that. Those decisions could have negative externalities for consumers and the environment.

Immutable distros typically use a declarative configuration that is easier to manage with terraform

Yes, system upgrade is the main maintenance task. With some monitoring, security updates can be automated but after system upgrades I must check manually that everything is working. E.g. incompatible configuration files, changes in 3rd party repos, errors that surface one week after the upgrade, ...

There are also smaller maintenance tasks that are tipically ad-hoc solutions to unsolved problems or responses to monitoring alerts. One of this ad-hoc routines was checking that logs do not grow too large, which used to be a problem in my first systemd centos, although not anymore.

PD: thanks for the bluefin read, it made me discover devpod/devcontainer as an interesting alternative to compose files

I totally see the advantages of immutable distros, particularly in a professional or cloud environment. Even as a hobbist, I feel tempted to use immutable distros if it were not because of:

- Learning. Figuring out how to migrate a setup even to the most mainstream-like immutable distro (fedora silverblue) can take a while, and to niche distros like talos even longer. However, a k8s-friendly setup with low customization requirements would help to speed up the migration (but it requires more powerful machines).

- Long term support. Regular distros like Debian and AlmaLinux offer free 5 and 10 year support cycles which means maintenance can be done every 1 or 2 years. On the other hand, immutable distros would require much more frequent maintenance, once every 6 months. A weekend every 6 months is a sizeable part of my time budget for hobbies.

One aspect in which immutables distros have improved a lot is in resource usage. They used to require significantly more disk space and have slightly higher minimum requirements than regular distros, but that doesn't seem to be the case anymore.

I am surprised this policy makes sense, but somehow it seems to be the case, since they have been doing since the 80s.

My concern is that, to produce a single antivenom dose, they need about 150 spiders, which are collected as grown up spiders or eggs. Telling regular citizens to try catch some spiders on their own puts them at risk. In case of something going south, the worst case is death and the best case is an antivenom dose consumed. So by promoting this campaign, the ARP organization assumes that the average person is a highly skilled spider catchers and benefits outweight risks.

Instead I would have expected that antivenom producers would breed the spiders in captivity. This would be a safer alternative for regular citizens. If they really need to catch live spiders in the wild, I would expect some kind of public service in which individuals place a call and then a professional comes to safely catch the spider. In absence of such a public service, my instinct tells me to kill the spiders in the fastest and most reliable way.

Great article and very intuitive explanation.

I also wanted to point out a (minor) typo. On equation 3, dZt is multiplied by sigma squared, but it should be multiplied just by sigma instead.

What about managing sysmoEUICC1-C2G with EasyEUICC/OpenEUICC?

Your argument boils down to Apple being the first one to use "App", which is an abbreviation for "Application".

But as fas as I know, that's exactly the same case for OpenAI GPT, they were first to use that abbreviation.

I hope that javascript filter runs on the server and not on the web browser...

"But simpler not-plug-in hybrids bucked this trend, with 26 percent fewer reliability problems than conventionally powered vehicles."

I think it makes a clear distinction between EV, plug-in hybrids and non-plug-in hybrids

This is nuts, isn't Samsung aware that tourism is a big thing nowadays?

Imagine going on holidays to Mexico and your phone stops working because some smartass at Samsung is pushing for price discrimination.

Thanks for pointing it out, it was not obvious on my phone.

I think that this website needs to improve the experience of mobile users. First, the book index on the landing page didn't contain hyperlinks to the free book chapters and these chapters don't appear on the preview, so users may assume it is a shady marketing tactic (trial, requires retweet, subscription to book platform, ...). Second, towards the end of the page, there could be a one-sentence-description of the next section with an hyperlink. Right now there is only a "next" button that is easy to miss, and it can be confused with generic "next post" buttons used by many blog platforms that users learn to ignore because they tend to be useless.

There is no explanation about how Kalman filters work in the article, it is just an advertisement for a book. And the book preview doen't even cover how to apply Kalman filters to conventional state space models.

For those into board games, financial misconduct of early american railroad companies inspired the renown game "1830: The Game of Railroads and Robber Barons" and the 18xx game series.

I don't understand why modern thinkpads are so hyped.

Recently I bought a L14 AMD and it doesn't even come with a decent keyboard or touchpad. The touchpad feels laggy and clunky, there are many of such reports for L14 and T14 models. The built-in keyboard ocassionally gets stuck pressing a key, and the I need to suspend-and-resume to unstuck it.

As you said, usb-c dock problems are a nightmare. Better get a good usb 4 cable with 40 gbps and 100 watts, or else you will spend countless hours doubting if problems are because of the cable. Spoiler: it's not any particular cable, it's the laptop.

For comparison, we use dell latitutes (intel) at work and they just work fine. I wonder if problems are common to all modern thinkpads, or if it only affects AMD models.

Gmail does not support base64 images according to caniemail

https://www.caniemail.com/features/image-base64/

The most reliable method is to attach a png image to the email and reference it from the body using `<img src="cid:insert_cid_ref_here" `

For the particular case of creating a wireguard mesh network in kubernetes, I've been quite happy with Kilo[0]. Does anyone with experience in both kilo and netmaker know how they compare?

[0]: https://kilo.squat.ai/

Does a student J1 visa qualify? Or are you required to have a resident permit?

There is also mason for installing LSP servers, DAP adapters, formatters and linters.

https://github.com/williamboman/mason.nvim

Well, AlmaLinux has a support cycle of 10 years while CentOS Stream is 5 years

Turns out you are right, things have changed in the SUSE world. And it does indeed seem to be worth considering. However, in terms of long term support (LTS), Leap still can't match RHEL clones.

https://www.theregister.com/2021/06/04/opensuse_leaps_to_153...

Why do you think SUSE is better? As far as I know SUSE is even less accessible...

Those are different Red Hat products and belong to different layers of the cloud stack.

Previously third party developers targeted RHEL directly, enhancing the RHEL ecosystem. Red Hat could play long term by (indirectly) giving away the base product to the community, and profit from the increase in support subscriptions.

New projects don't target RHEL as often as their primary platform, they prefer containers now. There is less ecosystem enhancement and it's more profitable to milk existing RHEL users. Centos Stream makes it possible to smooth the transition to the new paradigm.

Openshift, Coreos, etc have different purposes and follow different market dynamics. Mainstream developers don't target them directly.

RHEL market share won't last forever, so they have decided to milk the Red Hat cow as fast as they can.

Long term, people will switch to containers and cloud apps (Kubernetes, AWS). In this new environment, operating systems are commoditized and RHEL's moat will slowly shrink. Instead, developers will focus on alpine, debian or similar.

Thanks for the suggestions but unfortunately the poisson-binomial distribution [1] is not a simple combination of those two distributions. It requires of specialized packages like PoissonBinomial[2] in R.

[1]https://en.wikipedia.org/wiki/Poisson_binomial_distribution

[2]https://cran.r-project.org/web/packages/PoissonBinomial/inde...

My bad, I didn't express myself properly. The Julia language requires structs to be annotated and only those, you are right. Then if you want to reach a primitive level of type safety, function input types have to be annotated.

As a typescript developer who is now starting his PhD in finance, I asked myself recently which of these languages should I pick, taking into account how easy is to code in each language. To this end I tried to code a quick probability calculator based on the poisson-binomial distribution.

I found that R has the best libraries. It's the easiest language to use but it's also easy to write spaghetti code. Static analysis is a joke.

Writing code in Julia is slower/more tiresome. I have to annotate all structs and function input types. In languages like typescript, I would see this as an investment because tsserver will later use that information for linting and autocompletion, so the net return is clearly positive. In Julia however static analysis is very primitive and limited, you feel like having to write code twice: once for annotations and then again for the implementation, with very limited cross-checking between the two. When writing functions, output types are so hard to annotate that I just skip them completely and only annotate input types. Type errors are usually caught at runtime. The runtime messages are helpful though.

Unfortunately I couldn't even find a reputable Poisson-binomial package for Python so I stopped here. From my limited experience with python in other projects, I think that static analysis using pyright is mature and worth using (nets a positive return, although inferior to tsserver). This experience varies greatly between frameworks and packages, it's only worth to annotate the code if the underlying libraries also provide type hints.