thunfisch

Hugged to death. https://web.archive.org/web/20250128175526/https://swallowmy...

It's using a slash because that's how ISO 8601 defines time intervals. https://en.wikipedia.org/wiki/ISO_8601#Time_intervals

I heavily advocated for introducing GitLab at our company 5 years ago. We're now stuck with it, and GitLab is increasing the price continually. We started with the $4/month start edition because we needed 1 or 2 features, and now we are looking at $29/month for the same 1 or 2 features. We are seriously looking at ditching GitLab because of this. The price hike is insane. They're destroying their customer base in the long term for short term profits.

The only thing that I wish that camera-makers would finally agree on is a common mount. I'd love to try out camera bodies from different manufacturers, but that almost always means switching your entire lens collection as well. No, thanks.

I agree on built-in GPS (currently need to pair with a smartphone - which I'd rather love to leave at home), but everything else seems like a non-feature to me. I have zero interest for that on my Camera, because the experience would just be insanely annoying. Touchscreens are awful for camera operation - the feature is turned off permanently on my camera. Camera needs to be operatable blind, touchscreens are awful for this. Please give me more buttons and dials.

Do you know of Scribus, or do you not consider it a good OSS equivalent for InDesign? Last time I've worked with InDesign was around 2011, and it was meh. Scribus is also really realy meh, but gets the job done. I've got an Affinity license and have been using Designer for a bunch of projects - to me it's a toss between that and Scribus for what I do. They are totally different, but I have more experience with Scribus and therefore am much quicker in using that.

Based on the information provided being contradictory with itself, and omitting things that were at that time plainly visible on the site. Also, there were comments restored against my will from years ago, which were not contained in the data dump, even though others were. Which means that Reddit holds data in other places (otherwise they would have been unable to restore that), but have not detailed them in the GDPR request, which means that they have answered dishonestly and incomplete.

When the API changes were introduced, I tried to overwrite and delete my comments with a similar browser addon.

I had to repeat the deletion for months afterwards every week, because Reddit restored some random comments of mine again and again, even though they supposedly can not be recovered after being deleted.

I'm fairly convinced that they keep the data even after you delete it. I tried to follow up with a GDPR request, which they answered dishonestly and incomplete, I escalated that to my local authorities and then it took a few months for that authority to get back to me, requesting more information. Sadly I'm not in the mood to throw more of my personal time on this, but hope that at some point Reddit faces some legal actions for their behaviour.

I'm concerned with suggesting that it's enough to fight one group of adversaries, which will then be replaced with another group, and another, instead of actually fixing the underlying issue. Suggesting threat of life to those people (which is a very real thing for russians now) is no better than what happens when a ransomware gang locks down a hospital. That would be fighting fire with fire.

Did I read this wrong, or is the article essentially "Microsoft screws up security over and over again. Let's throw humans in Russia that are exposing these vulnerabilities under the bus of a dictatorship and possibly get them killed in a war. Instead of forcing the gigantic cooperation to not screw up security over and over again and finally clean their house"?

Wow.

Vector to me is more than just "high-performance" - It's a true swiss army knife for metrics and logging. We regularly use it to transform logs to metrics, metrics to different format, push them to different datastores, filter them, etc. It's wild how flexible this program is. It has become my first choice for anything regarding gathering/aggregating/filtering/preprocessing observability data.

Announced today. TL;DR: Run your loadbalancers in a single place, let other (workload-)clusters push their configs through KubeLB to the loadbalancer-cluster, which loadbalances onto the workload-clusters.

Announcement post: https://www.kubermatic.com/blog/introducing-kubelb/

Low-latency (e.g. less than 20 lines) Videoswitchers/mixers. There's a huge amount of data (12Gbps for 4K/UHD) per input, with many inputs and outputs, all with extremely tight timing tolerances. If you loosen the latency restrictions you can do a small numbers of inputs on regular PCs (see OBS Studio), but at some point a PC architecture will not scale easily anymore and it is much more efficient to just use FPGAs that will do the required logic in hardware. It's such a small market that for most devices an ASIC is not a option.

Nice to see this here. I've stumbled on this page while researching my stop and train change in Brussels on the way to Config Management Camp. This is the internet that I remember from the very early 2000s. Just plain information, a person happy to share their own interest and carve some space out in this sea of webpages.

This is what I want to see more of in the internet.

Just read everything from the source, instead of a third party blog post.

https://blog.thunderbird.net/2024/01/thunderbird-monthly-dev...

Does anyone have the link to the study? I'm tired of news articles referencing and summarizing their interpretation of a study, without giving me the option to validate the claims made.

From looking at it for a few minutes, the gist seems to be software-rendered wayland applications in a Container, streamed into a browser via Go? Or a Desktop client? Not certain yet.

Interesting question here would be the performance - is this efficient enough to get low latency and high framerates? if yes, this could be useful for a bunch of stuff, where you want to spin up a heavyweight GUI app on some kubernetes cluster quickly (think remote dev setup) and stream it to your laptop/tablet.

Edit: Looks like a Desktop client application, also written in Go. But the streaming is implemented with WebRTC and some custom signalling for keyboard/mouse/gamepad feedback, so this could theoretically also be adopted for a Webbrowser I guess?

One of the biggest reasons why digitization in Germany is so slow moving is that every municipality can (and does) decide on how to digitize individually.

There is no mandate for the upper levels of government to dictate which solutions are being used on a local level. Combine that with need for tendering on every single solution and you've got a big mess of small companies underbidding bigger companies that could unify the software landscape and instead build a cheaper, small solution that has zero interoperability with the neighboring municipality.

The need for fax et.al. is not because people don't know how to use computers, but because there's thousands of applications fulfilling the same exact job, but are incapable of talking to each other. Paper is currently the only compatibility layer that works everywhere.

There are ongoing efforts to provide a common data exchange format (technical working group) as well as redesigning how the software for government is being build (tendering processes, public money - public code movement, et. al.)

German government is currently incapable of doing its job in a way that is legally required, missing deadlines and not providing citizens with the services that they are entitled to because they are unable to manage the workload due to the paperbound processes. There is zero fear of humans being replaced by machines. It's rather that more and more humans are leaving government due to burnout.

That's exactly the one we picked. I've even commented it in our Sieve script. Funnily enough some spammers have picked up on this and send their spam even to those mailboxes.

Our company owns a one letter domain (e.g. "x.tld"), that follows a quite common sequence. A few months ago we've enabled receiving e-mails for all local parts on that domain.

We've received hundreds of notification mails, newsletter subscriptions, alerts (from internal systems disclosing details about infrastructure of giant corporations), etc.

It was quite fun, but became annoying quickly. We've then reduced reception to the common hostmaster@, ... mailboxes and for all other mailboxes we are now rejecting the mails with a nice reminder message in our Sieve filters.

Wow, this one brings back memories.

I've ran one of the https://www.parkytowers.me.uk/thin/wyse/z/z90d7/ for some years as a cheap (40€) homeserver and then later as a break-music-playback-machine for concerts with spotifyd.

And over 10 years ago I had one of the https://www.parkytowers.me.uk/thin/EvoT20/ to play around with. I never really got that one working in a meaningful way though.

Both times this page helped out a lot with technical details! Nice to see it on here. A treasure trove of hacking starting points to repurpose quite slow, but still useful hardware (and saving it from a landfill).

Google Maps has been broken on Firefox for many months. Not sure if it's my personal setup, or some addon, but on every Firefox instance I have, the search is entirely broken - it won't return any results.

I've since switched over to openstreetmap.org for most of my use. It's absolutely not perfect, the search is not as powerful as Google when it comes to recognizing typos or what I'm looking for, but the actual street data is really, really good, and the map view is very uncluttered.

We're using Terragrunt with hundreds of AWS accounts and thousands of Terraform deployments/states.

I'll never want to do this without Terragrunt again. The suggested method of referencing remote states, and writing out the backends will fall apart instantly at that scale. It's just way too brittle and unwieldy.

Terragrunt with some good defaults that will be included, and separated states for modules (which makes partial applies a breeze) as well as autogenerated backend configs (let Terragrunt inject it for you, with templated values) is the way to go.

A service in a very similar vein is https://healthchecks.io/ - which also provides a nice perspective on how low-effort the setup for a service with a substantial amount of users can be. https://news.ycombinator.com/item?id=31488910

The blog also contains a bunch of useful information and guides around the topic, including various unusual configurations (arduino/esp8266) as well as information on self-hosting.

Hurl and it's test cases have been awesome at our Ops team. We're managing an autogenerated config for hundreds of complex webserver rules, and we've been able to (auto-)generate hurl test cases for every single rule and test them in both CI and the actual infrastructure after deploying.

It's simplicity but powerfulness is amazing!

... and not a single mention of Linux. It's kind of sad to see, that Linux remains a step-child for MS Teams. My company forced the move from Slack to Teams, and effectively downgraded the entire experience of online collaboration for everyone. C-level moved "up" from S4B to Teams and called it amazing, while all the product teams and devs that were previously using Slack have lost a huge amount of online communication & culture due to the move. Of course that was attributed to "pandemic causing us to not be able to sit in a room together" - although I never did that even before, as I was a 100% remote worker already.

Aside from that - Teams on Linux has always been a mere afterthought, badly supported, important features intentionally cut/withheld, and sometimes not even kept up to date. I've moved away from the awfully broken Electron client to running the Teams website in Chromium App mode, in a bubblewrap container. It makes it a little bit less awful, but it's still painfully slow, buggy, and annoying to use.

I understand that with the intended complexity of various integrations, using a web platform might look like a win - but for me as a user, chatting with Teams feels multiple times slower than chatting on my old Pentium 120 MHz with Win98 with Java Applets. I'm all in for good integrations, but while WebView2 might bring some performance gains for Windows and MacOS, I'm afraid that we Linux users will never see a WebView2 version, and Teams will just get even more sluggish over time.

Would it be that hard to provide some native-ish baseline experience, and extend functionality for random integrations in places where it's actually needed? i.e. insert webviews where it makes sense, instead of hamstringing the entire application for that flexibility?

Setting up an automated lecture capture system at my university, based around the open-source https://opencast.org software. We did everything from buying hardware components to build machines that were specifically tuned to encode videos on the fly with the least space/noise possible to be put into lecture halls, writing the software to run on that, to running the video management system with opencast, integrating it with e-learning suites, setting up livestreams, etc.

I left the university and that job exactly one year before COVID-19 hit, and as far as I heard having that system in place saved a lot of butts - lectures seamlessly shifted to fully virtual with teachers recording videos either at home or in empty lecture halls, and pushing it through these video management systems.

We pushed a lot in 2015 to get approval for this whole endeavour, even though the university big-wigs pushed heavily against this, fearing that students would not come to lectures anymore. Turns out: that fear was bullshit, and in the end it helped out a lot of people.

I've been using this at work, sharing with my colleagues, and integrating this into our default CI/CD environment. It's awesome and has taken away the thought of "how do I install version XYZ of tool ABC here?" entirely. Just asdf what I need, done. Huge productivity boost.

https://media.ccc.de/v/31c3_-_6558_-_de_-_saal_g_-_201412282...

Immediately this talk from David Kriesel comes to mind. :)

I need to jump between different versions of terraform, terragrunt, python, ... all the time for work. asdf is an absolute asset and makes that so much easier. I can heavily recommend giving it a try! I've even embedded it into our default CI image, as it allows me to very easily install consistent versions of many utilities, while also allowing to override versions by choice afterwards.

I've been investigating several no-code tools because we need some internal tooling, where most of my team are not programmers, and I don't want to have a bus factor of 1 for this.

I've so far been most impressed with directus (https://github.com/directus/directus), which is open source, can be self-hosted, and can enforce ABAC which essentially allows you to do any permission scheme, albeit with complex queries if you have complex requirements.

It might be worth to give this a try.