lemmyvore

joined 1 year ago
[–] [email protected] 10 points 1 day ago (1 children)

It will fall through much faster than that. I'm thinking two years, tops.

[–] [email protected] 30 points 2 days ago (2 children)

I use whatever online storage service I want because you can add your own encryption layer so you only sync encrypted files. rclone supports lots of services and will also encrypt files for you.

[–] [email protected] 1 points 2 days ago

They're doing IP location checks, and they're doing them badly (there's not really a way to do them well). It's not working for me with people in the same town, and other people are reporting it's randomly working or not working with locations in the same neighborhood.

[–] [email protected] 1 points 2 days ago (1 children)

Unfortunately over here it seems to be doing IP-based location as I'm not able to add my brother who lives in a different part of the same town.

[–] [email protected] 2 points 3 days ago (1 children)

The repo delay is not the main cause of AUR warnings. While it can in theory cause mismatched dependencies for some AUR packages, in practice it doesn't really happen that often.

The main issue with AUR is that it's completely unregulated. Anybody can put anything in it, there's no quality criteria, AUR scripts run as root and can do anything on your system, 75% of AUR packages were not updated during the last year, 15% were released once and never updated, and 10% are completely abandoned.

Arch itself doesn't support AUR for those reasons. You should be wary of using AUR packages in general, on any system that can use them, always assume they can break at any moment, and never use them for anything critical.

[–] [email protected] 2 points 3 days ago (3 children)

Manjaro uses the binary packages prepared by Arch but a distro is more than just a set of packages. (In fact a distro should be more than just copying packages, otherwise it wouldn't be worth being called a distinct distro.)

Arch's goal is to be an ultra-customizable distro. To this end it starts out extremely minimalistic and requires the user to "assemble" it during the install from basic components, just so it doesn't end up with anything that's not wanted.

If a user can do this then they're above average in experience and knowledge; and since Arch can reliably assume this about its users it doesn't coddle them. The maintainers can afford to issue breaking changes that may even go as far as render your install non-operational, because they know their users can deal with it.

Another big Arch feature is being a rolling-release distro and bleeding-edge. This means that packages are released as fast as their developers can make them. This means they often have new bugs. This is the price users pay for the privilege of having very fresh software all the time.

Manjaro prioritizes a safe environment for the user and a more stable experience, where the install doesn't break (at all, if possible), and can be very easily be restored if it should break. And as a consequence it attracts users with less experience and Linux knowledge.

However, in order to achieve this Manjaro does some things very differently from Arch:

  • It holds back new packages and releases them late(r), when the Manjaro curators deem them usable.
  • It offers an alternate package manager with a more user-friendly interface.
  • It recommends the use of long term stable kernel (LTS) releases and mandates installing crucial drivers (graphical drivers in particular) through its own custom tools.

These differences mean that if a Manjaro user were to ask for help from an Arch crowd, the Arch users can't reliably help because they have no idea what's going on on the Manjaro side. They may use older packages and the issue being described was fixed in a very fresh version. They use tools (the kernel manager, the package manager, the driver manager) that Arch doesn't have.

Also there's very little overlap between the average Manjaro and Arch userbase. If an Arch user is more experienced and the Manjaro user isn't they're going to have trouble relating to each other. The Arch user doesn't see an issue in some occasional breakage, whereas a Manjaro user might consider that unacceptable and so on.

Last but not least there's a purely technical reason – Manjaro not only delays packages but hosts them in their own repositories, and sometimes goes as far as changing them. This makes it literally "not Arch" – using distinct repos is a step too far in terms of distro heritage.

[–] [email protected] 7 points 4 days ago

The Hoffman recipe is 12g of coffee, 250ml of water, 2 minutes steep time, give a small swirl to the recipient, steep another 30 seconds, then press down slowly over at least another 30 seconds. You can find the video on youtube.

There are many other factors involved such as the size of the grind, the uniformity of the grind, the temperature of the water, the steeping time, and the quantities of coffee and water – so really the recipe is just meant as a starting point. You will need to dial it in for each different batch of coffee.

Most of these factors have to do with caffeine extraction aka "yield". More time steeping, hotter water, more water & coffee and finer grind all increase extraction but in different ways, and over-extraction usually ends up tasting bitter. The opposites decrease extraction and under-extraction ends up tasting sour. The Hoffman recipe is a balanced start.

With the Aeropress you have easy access to all these factors and can customize the brew extensively but you have to do some trial and error.

[–] [email protected] 28 points 4 days ago

Well that's the nice part about the Aeropress, the process is so customizable that you can find a good recipe for just about any coffee.

The Hoffman recipe is not meant to be perfect, just a safe starting point. It can't possibly fit every single coffee batch out there.

[–] [email protected] 1 points 4 days ago (6 children)

How long have you been using each of them? In my years-long experience it's been the exact opposite. Manjaro goes out of its way to not break anything and offers safety measures out of the box to recover if something should break. Arch doesn't care, it introduces breaking changes all the time and expects its users to be able to cope with them.

They target very different types of users and have very different goals. Manjaro explicitly tries to be stable and user-friendly whereas Arch exclusively caters to advanced users and aims to be customizable above all.

You can achieve the same with Arch that you get out of the box with Manjaro but it's not there by default – because that's not something a lot of Arch users are seeking.

For a normal user, you probably won't notice that technically manjaro is not arch and EOS is.

What's a "normal" user? On Linux you get all sorts. But you will most definitely notice a difference between daily driving Manjaro vs driving Arch.

[–] [email protected] 2 points 4 days ago

All distributions make mistakes. It's a complex job. Debian stable had a local root elevation exploit on for a while a couple of months ago and nobody batted an eye. People would have a field day if that happened to Manjaro.

It's a double standard borne out of the resentment of a vocal minority and that sucks. The Linux community wastes so much energy on these pointless feuds. (And then they wonder why there's never the year of the Linux desktop...) Linux and FOSS are not about treating user share as a zero sum game but unfortunately there are people who can only think in terms of "if you use another distro you're dumb and I must ridicule you".

It's an especially narrow-minded take with distros like Manjaro, which is different enough from Arch that its users were never going to use Arch anyway.

[–] [email protected] 1 points 4 days ago (8 children)

First of all would be the fact that Endeavour is basically just an installer. It should have been an alternative offered by Arch alongside archinstall. I know it also offers some desktop setup but IMO that's too little to qualify as a distro. You can replicate looks and themes fairly easily. Might as well install Arch.

...but I don't want Arch because I'm at a point where I want my desktop distro to be boring and predictable, so it enables me to focus on other things. Arch needs more maintenance than I'm willing to put in. But I also want a rolling distro and having recent-enough packages.

Manjaro is a unique combination of rolling and stability. It's that combo that's the main factor but I'd be lying if I didn't say I enjoy not having to ever think about the graphics drivers, or about the kernel, and it's nice to have a graphical package manager.

As a sidenote, Garuda goes the extra mile and adds similar quality-of-life tools, while staying true to Arch repos. I think Garuda should get the publicity as an actual alternative in-between Arch and Manjaro, rather than Endeavour.

 

I took some photos at an event and I need to go through them and get rid of the bad ones (eyes closed, things in the shot, out of focus, blurred etc.) I'm not a pro photographer so no idea where to begin with photo apps. I've used RawTherapee and Gimp a bit.

What app will let me quickly browse the photos and handle (delete/tag) photo formats together (both the RAW and the JPG)?

 

I wanted to run my VPN/Tailscale setup past you, see if anybody has suggestions on how I could do things better.

  • Setup: home LAN (10.0.0.0/24), router+DNS on 10.0.0.1, server running docker containers on 10.0.0.2.
  • LAN DNS points *.local.dom.tld to the server, public DNS points *.dom.tld to my dynamic public IP.
  • Containers run in bridge mode with host, expose ports on host IPs via "ports:" mapping.
  • NPM with LE certs also in container, exposes 10.0.0.2:443, forwards to various other services.

Goals for Tailscale:

  • Accessing HTTP services via NPM from my phone when away from home.
  • Exposing select UDP and TCP non-HTTP services such as syncthing (:22000) or deluge RCP admin (:58846) to other tailnet devices or to phone on the go.

Goals in general:

  • Some containers need to expose ports on the LAN.
  • Some containers need to expose ports via Tailscale.
  • Some containers need to broadcast on the LAN (DLNA stuff) – but I don't want them broadcasting to Tailscale.
  • Generally speaking I'd like to explicitly control what's exposed from each container on either LAN or Tailscale.
  • I'd like to avoid hacking images with Dockerfile. I can make my own images to do stuff, just don't want to keep up with hacking other images.

How I progresed with Tailscale:

  1. First tried running it directly on the host. Good: tailnet IP (let's call it 100.64.0.2) available on the host's default network stack. Containers can use "ports:" to map to 100.64.0.2 (tailscale) and/or 10.0.0.2 (LAN). Bad: tailscale would mess with /etc/resolv.conf on host. Also bad: tailscale0 on host picked up stuff that binds to 0.0.0.0.
  2. Moved tailscale to a container running on the host network stack (network_mode: host). Made it leave /etc/resolv.conf alone. tailscale0 on host stack still picks up everything on 0.0.0.0.

This is kinda where I'm stuck. I can make the tailscale container bridged which would put the tailscale0 interface inside the container. It wouldn't pick up 0.0.0.0 from host but how would I publish ports to it?

  • The tailscale recommended way of doing it is by putting other containers in the tailscale's container network stack (network_mode: container:tailscale). This would prevent said containers from using "ports:" to map to host anymore. Also, everything they publish locally would end up on tailscale0 whether I like it or not.
  • Tailscale has an env var TS_DEST_IP that can mirror another IP. I could allocate an IP on host eth0 like 10.1.1.1, mirror that from the tailscale container, and target it from other containers explicitly with "ports:" when I want to publish a port to tailscale. Downside: 10.1.1.1 would be in the host's network stack so still picks up 0.0.0.0.
  • I could bridge the tailscale container with other containers on a private subnet, say 192.168.1.0/24 and use tailscale serve to forward specific ports to other containers over that subnet. Unfortunately serve is fairly limited; it can't do UDP and technically it refuses to forward TCP either to non-localhost (but you can dump the serve config to JSON, and hack that config, and use it with TS_SERVE_CONFIG= 🤮).
  • I could bridge tailscale with other containers and create a special container with a fixed IP on that subnet, mirror the IP from tailscale, and use iptables on that container to forward specific ports to other containers. This would actually solve everything I want except...
  • If I ever want to use another VPN which doesn't have the mirror feature. I don't know how I'd deal with that.
 

I'm posting this in selfhosted because Gandi increasing prices actually helped me a lot with being more serious about selfhosting, made me look into things like DNS and reverse proxies and VPN and docker and also ended up saving me money by re-evaluating my service needs.

For background, Gandi.net is a large and old (25 years) domain registrar and hosting provider in the EU, who after two successive rounds of being acquired by investment funds have hiked up prices across the board for all their services.

In July 2023 when they announced the changes for November I was using their services for pretty much everything because I manage domains for friends and family. That means a wide selection of domains registered with them (both TLDs and European ccTLDs), LAMP hosting, and was taking advantage of their free email hosting for multiple domains.

For the record I don't hold the price hike against them, it was just unsustainable for us. Their email prices (~5€/mailbox/mo) are in line with market prices and so are hosting prices. Their domain prices are however exaggerated (€25-30/yr is their lower price now). I also think they could've been smarter about email, they could've offered lower prices if you keep domains registered with them. [These prices include the VAT for my country btw. They will appear lower in USD.]

What I did:

Domains: looked into alternative registrars with decent prices, support for all the ccTLDs I needed, DNSSEC, enforced whois privacy, and representative services (some ccTLDs require a local contact). Went with INWX.com (Germany) and Netim.com (France). Saved about €70/yr. Could have saved more for .org/.net/.com domains with an American registrar but didn't want to spread too thin.

DNS: learned to use a dedicated DNS service, especially now that I was using multiple registrars since I didn't want to manage DNS in multiple places. Wanted something with support for DNSSEC and API. Went with deSEC.io (Germany) as main service and Bunny.net (Slovenia) as backup. deSEC is free, more on Bunny pricing below. Learned a lot about DNS in the process.

Email: having multiple low-volume mailboxes forced me to look into volume-based providers who charge for storage and emails sent/received not mailboxes. I've found Migadu (Swiss with servers in France at OVH), MXRoute (self-hosted in Texas) and PurelyMail (don't know). Fair warning, they're all 1-2 man operations. But their prices are amazing because you pay a flat fee per year and can have any number of domains and mailboxes instead of monthly fees for one mailbox at one domain. Saved €130/yr. Learned a lot about MX records and SPF/DKIM/DMARC.

Hosting: had a revelation that none of the webpages I was hosting actually needed live dynamic services (like PHP and MySQL). Those that were using a CMS like WordPress or PHP photo galleries could be self-hosted in docker containers because only one person was using each, and the static output hosted on a CDN. Enter Bunny.net, who also offer CDN and static storage services. For Europe and North America it costs 1 cent per GB with a $1 minimum/mo, so basically $12/yr since all websites are low traffic personal websites. Saved another €130/yr. Learned a lot about Docker, reverse proxies and self-hosting in general.

Keep in mind that I already had a decent PC for self-hosting, but at €330 saved per year I could've afforded buying a decent machine and some storage either way.

I think separating registrars, DNS, email and hosting was a good decision because it allows a lot of flexibility should any of them have any issues, price hikes etc.

It does complicate things if I should kick the bucket – compared to having everything in one place – which is something I'll have to consider. I've put together written details for now.

Any comments or questions are welcome. If there are others that have gone through similar migrations I'd be curious what you chose.

 

I'm thinking of putting all my email archive (55k messages, about 6 GB) on a private IMAP server but I'm wondering how to access it remotely when needed.

Obviously I'd need a webmail client but is there any that can deal with that amount of data and also be able to search through To, From, Subject and body efficiently?

I can also set up a standalone search engine of some sort (the messages are stored one per file in regular folders) but then how do I view the message once I locate it?

I can also expose the IMAP server itself and see if I can find a mobile app that fits the bill but I'd rather not do that. A webmail client would be much easier to reverse proxy and protect.

 

I've repurposed a 32 GB M.2 SATA SSD as a bootable "USB stick" and I'm putting useful tools on it. So far I've got memtest, seatools, gparted live, system rescue, clonezilla, and a live install iso of the distro installed on my PC. What other great bootable tools am I sleeping on?

 

I use multiple workspaces and I open text files all the time.

Once upon a time Mousepad used to behave sanely and would open them in a new tab if there was already an instance on the current workspace, or open a new window (on the current workspace) if there wasn't.

They broke that at some point. Now it's anybody's guess where the file will open. Maybe it opens in a tab in an existing window on this workspace. Maybe in a tab in a window on a random workspace. Maybe a new window on this workspace even though there's one open. I've given up trying to figure it out.

As a last resort I can use wmctrl to figure out how to open the files and can script a sane launcher myself – provided that the editor has --tab and --window options AND lets you specify the window instance. Mousepad has the former but not the latter.

So, do you know any editor that can do it by itself or has those options so I can do it myself? TIA

 

Hi, I'm trying to find the subtitles for Harmy's "Despecialized" Star Wars remakes and I was wondering if anybody has any ideas. The original website for Project Threepio points at a blog that seems abandoned and an old private tracker (MySpleen) that never opens to public anymore. Even just the English subs would be great (the original pack contained extensive language coverage in DVD format so I was given to understand it was quite large). TIA for any hints.

2
submitted 1 year ago* (last edited 1 year ago) by [email protected] to c/[email protected]
 

I need a very simple method for non-advanced users to share each other's screen explicitly when they need help. They're running XFCE on Manjaro and the machines involved are using Tailscale. Edit: SSH access is also available, with key authentication.

I need something super simple because they are remote from me and from each other and any graphical setup will have to be assisted sight-unseen over phone. So ideally just (1) install something (which I can do for them over SSH), (2) pick something from the Applications menu and maybe (3) press a big "START" button.

It's also ok-ish if the remote capability is present all the time and I can connect without their explicit permission, but you can see why it would be best if they did something to enable it...

I've been looking for a solution but all I find is stuff that's way too complicated OR starts a new desktop session instead of showing the current one.

Edited: to clarify I'm not the one who will be remoting-in and to mention SSH is available.

TIA

 

I've been using Gandi for over 20 years, almost since it was founded. Since being acquired in 2019 by Montefiore Investment and this year by Total Webhosting Solutions their service have become more and more expensive and have finally priced me out.

For context, I administer a bunch of domains, mailboxes and HTML websites for my family and extended family, and I prefer services hosted in the EU because of GDPR and local availability.

This post is meant as a list of practical decisions in 2023 for the small time selfhoster. If anybody wants to comment on what Gandi (or rather TWS) is doing feel free to do so in the comments, I'm curious myself.

Prices I've mentioned use my country's VAT so will vary slightly for you.

Domain names

Domain names have always been a bit on the expensive side with Gandi but they used to include a lot of features for free with them (SSL, DNSSEC, mailboxes, a small static website, WHOIS privacy, local contact for TLDs that need it etc.) and what they added extra was proportional to the base TLD cost.

For the next renewal all my domains were slated to jump to €28 across the board. If you have domains with Gandi try adding some renewals to the cart and check in advance.

I had to look for an European registrar because I have lots of European ccTLDs that the usual suspects like Cloudflare and Porkbun don't support.

I'm moving to INWX.de and will be saving 25-60% per domain. This takes into account WHOIS privacy where needed for an extra 5€/domain (EU ccTLDs are private due to GDPR but we own a couple of TLDs too) as well as local contact services where required (price varies by country).

Email

I manage multiple mailboxes but they have low traffic and low storage requirements. Gandi will be offering them at €55/mailbox/year. I'm not questioning their pricing, 3-4€/month for email is common, but typically charged by email-focused services.

Anyway, this per-mailbox model would price us into hundreds of euros for resources that go 99% unused. I'm switching to Migadu.com, who allows unlimited domains and mailboxes (within common sense) under a single account and charges for the conflated storage space and emails sent/received across all mailboxes.

Migadu tiers start at 20€/year for 5GB and 200/20/day (soft limits).

Webhosting

We were using Gandi's smallest hosting package for about 100€/year, which was slated to jump to €135. Not an outlandish price for your typical PHP + MySQL hosting, especially since it had some VPS-like features. Then again the typical webhosting service would include a couple of mailboxes and some other goodies.

This was a good opportunity for us to reevaluate out hosting needs and realize we can ditch PHP+MySQL (if we really have to revisit it we'll consider VPS offers in the future). It's mostly static sites, image galleries and a bit of blogging. We've cached all our stuff as plain HTML/CSS/images and moved it to BunnyCDN.

Bunny lets you define a file bundle, gives you FTP access with a unique username+password, lets you pick the extent of replication, puts a CDN on top of it, and lets you point a domain name to it. Also throws a bunch of web server-ish features on top like rules/rewrites and Let's Encrypt SSL.

They actually offer more features than that but I've just mentioned the minimum you need for serving a bunch of static websites.

Bunny pricing starts at $0.01/GB (with a minimum of $1/month) and you pay as you go.

Nameservers

Since we're doing this I've taken the opportunity to dab into DNS. Turns out it's not that hard. There's only like half a dozen of commonly used DNS record types and everybody's helping you with them – email services like Migadu generate the email-related ones for you, registrars and managed DNS services generate the SOA for you, they have forms that tell you what fields are needed etc.

There are lots of managed DNS options. Registrars usually include nameservers and let you mess with the records so INWX was one choice. Bunny offers DNS service that integrates with their CDN. deSEC is a completely free service I'll be using as backup.

All of the above also offer APIs so a bash script will be taking care of dynamic DNS.

 

Upgrading a self-hosted server (1)

Welcome

Hi, I'm starting a series of posts that will follow the upgrades I'll be doing to a self-hosted machine that serves as NAS and also runs all kinds of self-hosted software. I'm lazy so it will probably take time, don't expect me to post too often.

About me: I've been using Linux exclusively for personal use (both desktop and servers) for about 20 years now. I've used several distributions over the years, I've built my own stuff from source (including kernels) and I've done Linux From Scratch. I'm not a Linux expert or professional sysadmin but I know my way around it, and I can learn what I don't know. So don't be afraid to make any suggestions no matter how complicated.

The current state of the machine

  • It's a PC using an i5 7400 CPU, has a built-in GPU with support for h264 hardware encoding and MPEG2, VP8, VP9 and HEVC hardware decoding (this will come in handy for video transcoding).
  • Only 4 GB of RAM, I have ordered a dual 2x16 GB kit.
  • The system drive is a Transcend M.2 SSD (32 GB). SATA rather than PCIe unfortunately but it will do fine for the time being.
  • The OS is Ubuntu Server 16.04 LTS using Expanded Security Maintenance for updates.
  • It's currently running SSH, NFS, Samba, CUPS, OpenVPN, Emby and Deluge on bare metal. Some of them come from distro packages, some from binary releases straight from the developer.
  • There are 6 HDDs forming 3 pairs of RAID 1 arrays. 6 drives was a limit I chose from the beginning, and the case and motherboard were chosen accordingly (cage for 6 drives and 6 SATA connectors).
  • My ISP provides a public dynamic IP and allows port forwards.
  • I have a router that I've recently upgraded to the latest OpenWRT so it also runs Linux, can install packages, it has a web admin interface etc. and can do some interesting stuff.

What I'd like to do

  • Increase the RAM to 32 GB.
  • Stick with a Linux distro, as opposed to a NAS-tailored OS, Unraid etc.
  • Install Debian Stable on a SSD, most likely via debootstrap from the Ubuntu system.
  • Add a GRUB menu entry that makes a passthrough to the other system, so I can keep them both around for a while.
  • Use docker-compose and possibly Portainer for as many of the services as it makes sense. Not sure if it's worth bothering to make containers for things like SSH, NFS, Samba.
  • Add more services. I'd like to try Jellyfin, NextCloud and other stuff (trying to degoogle for example).
  • I'd like to find a better solution for accessing services from outside the LAN. Currently using OpenVPN which is nice for individual devices but gets complicated when you want an entire remote LAN to be able to access (to allow smart TVs or Chromecast to use Emby/Jellyfin for example). I'm hoping Authelia + reverse proxy will be able to help with this.

What I'm not interested in

  • Not interested in using Plex. I've used it for a couple of years, it's a fine piece of software but I don't like the fact they now mandate access through their server or injecting ads.
  • Not interested in changing the filesystem or the RAID setup for the HDDs. RAID 1 pairs give me enough redundancy. The HDD upgrades are very simple. I'm fine with losing 50% of capacity.

Any and all suggestions and comments are welcome! Even if they're about things I said I'm not interested in. It's always possible there are things I haven't considered.

view more: next ›