starshipwinepineapple @ starshipwinepineapple @programming.dev

Posts

2

Comments

138

Joined

2 yr. ago

Is Arch Linux the right fit for a newbie to Linux? The right answer is "it depends", not "never". Would I recommend Arch to my mom? No. Would I recommend it to my programmer colleague who already lives in the Powershell? Sure, why not.

Yup, i had a lot of people tell me that arch wasn't a good beginner distribution, and had some friends try to talk me out of it. But i was planning to move to Linux for over a year and had set up Linux servers in the past. Just hadn't used one for my main PC. I've been on arch for over a month and it's been fine. I still wouldn't recommend it to every beginner but I'm not going to say it's never appropriate.

1. User devices (main workstations, phone photos, etc)
2. Local NAS (sync w/ #1, backup to #3)
3. Cloud backup w/ commercial provider

They added 2 weeks to it.

Here's a txt summary of the books

And a shorter txt summary

I have had the same experience. Have used all three at some point but mostly use nginx for new servers

I assume Yale isn't broke but idk. Universities are just like any other business where they will cut products that aren't making money or performing as well as others. The article talks about the course needing many teacher assistants to field student questions and hold labs, and that originally these costs were covered by a donation which has now run out.

It also could just be some internal politics and blaming it on financials is the public reason.

But you're not wrong that student tuition costs should theoretically go to the courses they sign up for

CS50 is produced by Harvard and is opencourseware (free) that isn't going away.

What is changing is that Yale won't be offering CS50 courses going forward, seemingly due to funding issues.

The archinstaller script is pretty good if you're just needing a basic setup. Ive been really happy with a btrfs partion from the recommended disk layout, then using btrfs snapshots + grub bootloader to load from snapshots. You can also create a hook on pacman so that you create a snapshot when you upgrade packages.

Since you didn't mention your experience, id recommend looking at the various desktop environments so you know which one to pick during install. You can ofc change later.

And read the arch docs. They are very good and have a lot of time invested into them. If you find you don't have the patience to read them then you're probably going to want to look at a different OS. Good luck!

Edit: i see now they're talking about private IP, but in case you want to learn about getting a static IP for other things...

Many ISPs will give you a dynamic (changing) IP rather than a static (unchanging) IP. Just check your IP once a week for a few weeks to see if it changes.

There are some services that get around this by checking your ip regularly and updating their records automatically. This is called a dynamic DNS provider (DDNS). I used to use "noip" but since then there are quite a few like cloudflare DDNS.

Beyond that you just would want to make sure your router or whatever device is assigning IPs on your network to give a static assignment to the server. Assigning IPs is handled by a DHCP server and it would usually be your router, but if you have a pihole you might be using that as a DHCP server instead.

Between DDNS and DHCP you can make sure both your external IP and internal IP are static.

Yes. If you're using lets encrypt then note that they do not support wildcard certs with the HTTP-01 challenge type. You will need to use the DNS-01 challenge type. To utilize it you would need a domain registrar that supports api dns updates like cloudflare and then you can use the acme.sh package. Here is an example guide i found.

Note that you could still request multiple explicit subdomains in the same issue/renew commands so it's not a huge deal either way but the wildcard will be more seamless in the future if you don't know what other services you might want to selfhost.

Agreed, though i do think it's a privacy thing. Many people use privacy and anonymity interchangeably but they are different things.

The options are:

• use a single email. If it is leaked you need to update hundreds of accounts or risk falling for a malicious email
• use a catch-all email and each service gets a separate email, but you can't turn off receiving mail at a specific address unless you use a sieve filter. This doesn't stop people from just guessing random addresses.
• use specific aliases for each service. Idk about this specific project but usually you can turn off receiving mail at an alias. So if a company gets a data breach i just change my email (or close the acct), then i turn off the old alias.

I did the catchall for a few years but have been doing aliases for 5+ now. In the end, the only people/ companies who have my email are the ones I want.

Even has kitty timestamps further down 😻

I do all my coding in the cloud, via Google docs 😎

Just some general advice:

• get regular users. Contributors are going to be a subset of users as another commentor mentioned.
• make sure to have a CONTRIBUTING.md and that it is clear/ easy to follow. Some projects will link to a separate wiki from the .md which is fine. But make sure your "first time contributor" instructions are easy to follow to set up whatever dev environment needed. The less clear the documentation then the more motivated the contributors will need to be.
• if you haven't already, make issues with feature requests that you are wanting to add. Include enough details that someone other than you will understand your requirements.
• consider a label you use to signify "great issue for a first time contribution". These should be relatively simple fixes or simple features but give time for someone else to try them instead of completing it right away. Make sure to reference this label in your contribution documentation as a great starting point. If you're able to get someone to do a simple fix then they will have set up the dev environment and may do other future issues.
• advertise that you're looking for contributors. Point out your docs, first time contributor label, and any specific features you want/need help with.

So you dislike external sync options but also don't want to pay for internal sync options? Additionally you are in a self hosted community so you're looking for a presumably open source project (some you listed are not), and given internally supported sync services would be one way fund development i think this narrows what your are looking for by quite a bit. You basically would be looking for an open source project that meets all your other criteria and happens to let you sync the files to your own server for free. Why would such a project not just let you take things into your own hands with whatever flavor of sync/backup you prefer? Otherwise if they're building a sync system it would probably be a monetized cloud service which brings us back to the beginning.

Maybe such a thing exists, but I haven't seen such a thing since that is extra development for little to no gain. Most people are happy to either pay for the cloud service to fund development or sync on their own.

Logseq: Same issues as with obsidian: Paid sync. Didnt look much beyond

Logseq is open source. Obsidian is not. So yes, both have paid sync but you can also just sync or backup the files on your own. Just be careful of sync services that sync while files/db are in use to avoid conflicts.

This is how i initially got started and i always like to recommend it. CS50x (introduction to computer science) is their college curriculum made available for free as opencourseware. Their lectures are very engaging imo, and you get problemsets to practice and check your answers. The problems are done in an online environment which i like so you don't get bogged down in setting up your computer before you've even learned how to code. And then at the end you pick a project of your own and when you finish you get a free certificate (don't bother paying for the "verified" one)

One other thing i think cs50 does pretty well is help teach you how to solve problems and how to read documentation. The reality is that learning how to code isn't just learning a coding language. Knowing how to solve different types of problems and how to read documentation are core skills that let you get away from "tutorial hell" and start working on a project that excites you.

I deleted my 12 year old account over the latest privacy policy update which auto opted-in to using your data for unspecified AI purposes. There was some discussion here, and while a strava rep did give specific examples in their response, the privacy policy was not updated and continues to be both broad and auto opted-in with no way to opt out.

Regardless of the current use of AI, the broad privacy policy creates the potential to allow them to do many things with your data without telling you about it in the future. And that thread discusses some potential problematic uses that you could be opted into without ever knowing it.

The privacy policy needs to be more specific, and allow opting out (or better yet, make opt out the default).

And yeah, API change is pretty crap too.

Second the Automate The Boring Stuff recommendation, especially if you're looking for a physical gift (or free online as mentioned)

Id also just in general recommend CS50-python as a free course for python. Engaging lectures, problem sets you can check your solutions, and you finish with a project of your own choosing. No programming background is needed. Don't buy a verified certificate, the whole course is free along with a free certificate

Thanks, yeah looks like they are wanting to build on their own reader app.

https://elevenlabs.io/blog/omnivore-joins-elevenlabs

Everyone has a hobby 🙃