andreashaerter

Joined 3 years ago

Karma
133
Submissions
49
Member since
Aug 18
IT guy from Germany.
View on Hacker News

Recent submissions

Comment · 1 month ago
There are also a few Hugo templates for vanity import paths.

Here's mine: https://github.com/foundata/hugo-theme-govanity (e.g. used at https://golang.foundata.com/ )

And yes I'm aware of the irony of hosting this on GitHub... still figuring out a good workflow for maintaining our OSS on Codeberg and GitHub in parallel fed from the internal forge. The dependency on our own domain is real but we favor it.

Comment · 1 month ago
Incident started ~9 hours ago. Pulls work; pushes and other writes don't.
Comment · 2 months ago
Comment · 2 months ago
Can't relate (Fedora 44 GNOME with Wayland, AMD Ryzen AI 7 PRO w/ Radeon 860M), binary install (no Flatpak). Using Zed as daily driver since ~4 Months, never had such problems.

Completely ditched VS Code and never looked back.

Maybe some Ubuntu-specific issue?

Comment · 3 months ago
Comment · 3 months ago
BUT:

1. For generic TLDs like .com/net/org/.<all kind of new ones>, the rules mostly come from ICANN, the registry, and the registrar.

2. For country-code TLDs like .de/io/to/io/ai/ly, the rules come from that ccTLD's own registry. Best example AFAIK: .ly domains must not contain obscene contents contrary to Libyan law or Islamic morality (which was a driver for bit.ly's domain change)

So ccTLDs are not simply "generic TLD rules plus country rules." They are their own category with their own policies. Trademark violations are one way to lose a domain, but not the only one. The risk is MUCH higher when using a country TLD.

Comment · 3 months ago
On topic when draw.io (diagrams.net) realized that, from https://www.drawio.com/blog/move-diagrams-net/#io-domains

    There are two major problems with .io domains:

    The first is that the islands which should own the domain suffix, don't, thanks to a wonderful piece of modern day British Imperialism. If you ever feel the need to donate to us, please give it to a more worthy cause.

    Secondly, there was a security issue with the .io domain. In 2017, a researcher managed to take control of four of the seven authoritative name servers for the .io domain. We accept that mistakes can happen, strong processes limit the chances of them happening, but they still can.
Comment · 5 months ago
Comment · 5 months ago
> What we need in the small business space is a tier of services where small businesses can self host using their own on-premise, vertically scalable infrastructure (ie: 1 server). In most cases they can tolerate some downtime and, even if they don't want to, a lack of resources usually means they don't have a choice (ex: they're not running HA network connections).

I think the same: Small service businesses care most about Time To Recovery (TTR) when doing services. As long as they communicate at least by phone and the website is up, they usually tolerate downtime when they know when their backoffice services are back online.

This is classic Business Continuity Management, 5-10 questions usually make clear what must work in every case when and what has to be available for supporting this process. Example: I got a customer which prints all logistics / distribution labels in batches. They can still work where money comes in (=shipping stuff) for quite a long time (4h min, 8h max) if the next batch of labels cannot be printed / some system is going down needed to support shipping. So no need for expensive HA around legacy software, but enough time for a good process to get back online with the latest backup on replacement hardware which is already there on-site.

The thing is: HA is FAR more expensive and complicated than e.g. getting another stand-by server as fast replacement, maintain the hypervisor on this second server e.g. every six month and test restoring backups on it once a month (best: automated: IMPI boot, restore without VM networks, testing, shutdown). Same with a firewall; two used Enterprise Servers + Proxmox VE Subscription, OPNSense + 2 x N150 Hardware and two consumer WANs (e.g. Cable and VDSL) is really not that expensive if only the WAN is a bit more complicated from the POV of a SME admin because of failover. Classi VLANs+ACL and services like surveillance as needed...

> Businesses with <10-20 employees are often viewed as not being worth the effort of having as a customer, so they end up with self-serve, unsupported, non-discounted, over priced, trash subscriptions. By the time they grow enough to be a valuable customer their only experience with some products is misery.

Exactly. This is why I do SME IT since ever, no matter for which $BigCorp I've done consulting and DevOps. I automate them. I consult them. My company (plug: https://foundata.com) does it for a few bugs per month (Hypervisor, Groupware (Calendar, Mail) Firewalling, VPN, Directory Services, Jitsi/OpenCloud/BBB) if they understand that they finance the high quality of the managed services ON THEIR HARDWARE with all other customers and we do not work per-hour but per-service + we run Open Source also for other reasons than "no or fewer licensing costs".

And I like it even this does not make you rich. Because I REALLY share your concerns ("owning all the small businesses of certain types; family doctors, dentists, optometrists, vets" -> I don't know where you are from, but it is the very same here in Germany... example: https://www.ndr.de/fernsehen/sendungen/panorama3/Spekulanten...)

Comment · 5 months ago
> if you have a choice

I just read:

> If I had learned one thing from my past life was that if you see the signs of an abusive relationship, you have the option to walk out, and you don't, all that follows is your own fault.

so... :)

Comment · 7 months ago
Is there a reason for OCI images? It is just a binary? I have all used versions in ~/.local/bin/

  ~/.local/bin/hugo0.145.0
  ~/.local/bin/hugo0.148.0
  ~/.local/bin/hugo0.149.0
  ~/.local/bin/hugo0.150.0
  [...]
and a convenience symlink ~/.local/bin/hugo, pointing to my "production" version. I can easliy call whichever version I like with hugo<tab><tab>. What am I missing?
Comment · 7 months ago
Can't agree, even though I was also hit by breaking changes from time to time with my own templates. Because this has a BIG upside: I am very happy with Hugo and how they love to solve problems in detail and think through features to the end. Because of these frequent re-factorings and sometimes (!) breaking changes, it gets more elegant every year.

And this is really not a production problem. As stated in the comments before, just don't upgrade if you don't want to / have no time yet. It is a self-contained static site generator without external dependencies. It won't break. And the security of an old Hugo binary is mostly a non-issue if you do not load remote content.

And, if you have some time: Their changes are really well documented. The changelogs are really good.

The main problem is sticking with an unmaintained template of a third party which breaks when you finally want to upgrade and don't want to fork / don't know when a an template update comes along. But that's the reason I write my own. It was worth the effort.

Comment · 7 months ago
The

> Low maintenance[...]A single Docker container, using SQLite as database. Automated migrations, zero breaking changes.

sounds nice. How does this compare to Linkwarden or Karakeep (formerly Hoarder) in terms of features? A comparison would be useful.

Comment · 7 months ago

Live from the official Hacker News API

A modern reader for Hacker News · SvelteKit + Skeleton