Microslop is starting to play make believe because no one uses Copilot.
This has to be a new low in the AI marketing wars of 2026.
Remote
Timo Zimmermann
@fallenhitokiri@social.screamingatmyscreen.com
mastodon 4.7.3I build backends, apps and sometimes break them, practice martial arts and enjoy video games. I believe I have seen the whole Internet once, maybe twice.
Tools of choice #python #django and #golang
Sometimes musing about #gaming and #gym life
549 Followers
163 Following
50 Posts
Joined April 27, 2022
Now that IPv6 is approaching 50% adoption it’s obviously time for a new standard. https://www.ietf.org/archive/id/draft-thain-ipv8-00.html
Replying to
@Edent@mastodon.social I run a local linkwarden (basically fancy Wallabag) and have a separate collection for “surely one day I’ll take a look”.
Not a perfect solution but it gets the job done for me.
RE: @piccalilli@front-end.social
> Legitimate services never ask you to run terminal commands.
You’d hope so. Yet half of the tooling we use these days conditions us that this is a good, valid strategy.
Or to rephrase: keep `curl … | sh` out of your documentation. This is a hill I will die on.
Apples developer experience is so exceptional, I deleted the whole Xcode project. I’m not doing this to myself for a side project.
https://www.screamingatmyscreen.com/notes/apples-developer-experience-in-2026/
Coming from AWS to GCP I am honestly pleasantly surprised. It’s like someone actually cares for the product and the people paying for it.
My CalDav server is getting closer to be “done” and be released to the wild. If there’s anything you’d like to see in a single binary, SQLite data store, basic auth and potentially OIDC CalDav server - now is a good time to let me know :)
https://www.screamingatmyscreen.com/cal-now-with-100-more-evolution-support/
Not being in a boat with Sony is likely one of the best moves any studio can make.
Also Stellar Blade was so good, I hope the sequel lives up to it.
Boosted by @oxy@social.bsdlab.au
Replying to
@oxy@social.bsdlab.au I moved one host to Stalwart and one to mox instead of only replacing Dovecot. I cannot recommend Stalwart, but mox is doing a really good job.
I went outside three times this year when the sun was out. I had three mild sunburns.
I didn’t use any sunscreen because yesterday was literally a five minute walk.
I assume this is nature telling me to stay inside for the next 6 month. And who am I to argue with nature?
Of course VPNs are a problem when protecting our children… I’m even all out of snark with the EU. https://cyberinsider.com/eu-calls-vpns-a-loophole-that-needs-closing-in-age-verification-push/
I am closing in on 4 years with OPNsense. It’s time to do a little maintenance and move things around.
There’s a good chance the new router will be Alpine and PyInfra to manage the three config files and nftables I need.
Obviously no one is thinking about rooting my modems to put them in bridge mode, setup CAKE SQM, BBR and FQ, kill the whole thing three times and regret life choices. But latency in Rivals and OW should theoretically be a few ms better :D
The fact that this was flagged and removed from the front page for some time is all you need to know about HN these days. https://www.thatprivacyguy.com/blog/anthropic-spyware/
Replying to
@wj@ruby.social disclaimer, I’m very biased because @Airmed@social.screamingatmyscreen.com is my wife and works for our company as digital media artist who’s specialised in logo and print. What I know for sure is that she’s getting an opening in her schedule soon and that some important people in the industry once said she’s pretty good.
Replying to
@brettcannon if I understand the workflow and requirements correctly I’d need to move CI and signing to one of the few blessed systems such as Microsoft’s falling apart slop dumpster fire.
If this is correct I honestly hope it causes the opposite reaction and people start looking into / working on alternatives instead of using the system as is.
Replying to
@felixxm well done! 🎉
I finished the first iteration of my code agents and took the opportunity to give vibe coding the controller a shot and it was as bad as expected.
I stand by what I said a few weeks after the term was coined. „Vibe coding: The irresponsible behavior of an engineer not doing their job.“
Okay, this will sound stupid, but hear me out. Outdoor TV.
I’m sitting on the patio and it’s really nice, we did a good job with that. Later today I might light a little fire pit and get a scotch.
But I also want to watch the LEC games. The Spanish fans are insane, the whole stadium is hyped and the banter between teams is on.
My headphones are comfortable enough, but the iPad is really too small to enjoy the game.
Zed started messing up my imports in Go projects when saving a file.
As in: putting parts of strings in random order. The randomness is reproducible for the same set of imports.
I might have to look into this over the weekend. But for now I’m glad my neovim setup doesn’t randomly update itself.
Replying to
@chanakya@social.screamingatmyscreen.com LLMs don’t have enough training data for that. *scnr*
LLMs are really good at copying things. Including your standard SF startup tech bro behavior - the first line of SOUL.md basically sums up the personality of too many people I’ve met there.
https://theshamblog.com/an-ai-agent-wrote-a-hit-piece-on-me-part-4/
Replying to
Replying to
@Edent I am currently doing a lot of vision work on documents, PaddleOCR for a fully local setup is doing decently, might be worth looking into.
Compared to Teserract its setup for scene text detection, so it got a chance to work for benches :)
Replying to
@webology@mastodon.social @simon@fedi.simonwillison.net @djh@chaos.social I think the interesting part will be when Europe looks at this and invokes the DMA which should AFAIK apply here.
If this would be the case and if they would rule to open it up for everyone not just the company who already got all your data will be using it. :/
Replying to
@lunya@sk.lunya.pet @brohrer@recsys.social now this isn’t the first post about this little instance I’ve seen lately.
Did I miss something or are people just finding the domain funny? :) (serious question actually in case I missed something going on here)
Wow LM Studio, you. sound like APIs I had to fix in the past...
`[ERROR]
Unexpected endpoint or method. (POST /v1/internal/encode). Returning 200 anyway`
I ordered a RISCV BananaPI and a case. No NVMe screw, the case only has two of four screws holding the backplate in,... and do not get me started on the power supply and how much I trust it.
But I got a Linux distro I never heard of running, so it's at least working.
Weekly progress on Lazerbunny, mostly talking about MCPs and trying a few models for agentic coding. They are still far from where I’d consider letting them just handle a task without super vision, let alone system design or architecture.
https://www.screamingatmyscreen.com/mcps-and-questionable-vibes/
I have three systems. Two can curl an http service just fine. Except one Debian box. Curl and Python insist on trying to redirect to https. Zero indicator in the web server config to do so. I’m done for today.
Replying to
@SiteRelEnby@tech.lgbt while I’d obviously prefer that I honestly don’t see Google make an ethical decision at this point :/
Replying to
@justine@bsd.cafe @justine@snac.smithies.me.uk Ivory on iPadOS can load your profile and shows a post count of 1 but it doesn’t load the post when trying to load your posts.
The mastodon web interface shows that „some posts might be missing“ and shows a link to your profile where the post display.
Replying to
@weel@towns.gay @DerMolly@toot.kif.rocks I want this toot on a T-shirt for my next consulting gig! :)
Replying to
@chrisg@fosstodon.org SQL based system are doing wonderful on todays hardware.
And they’ve been tested at basically any scale for decades, anything trying to replace them would IMHO need a tremendous amount of testing and effort to even come close to the understanding and reliability we are seeing in todays systems.
What exactly are the issues? I think I’m missing something here.
Replying to
@chrisg@fosstodon.org this is certainly an interesting take on the problem space!
I’m curious what your expectations for a DBMS are. I’d argue SQLite mostly fits the bill.
Federation itself should IMHO live in the app layer and not care about the underlying storage.
Replying to
@billseitz@toolsforthought.rocks I only was briefly involved with a react native app and my only ask was „make the arrow for the back button have the same spacing to the text as the native iOS control“.
The agony in the teams faces showed… :/
Well, this is certainly something I love to read :)
Replying to on hachyderm.io
@miketheman@hachyderm.io Is this a general statement or are there exceptions for organisations on a paid plan?
If the OSI definition is relevant does that mean all OSI advice would be followed? What if they declared binary blobs OSS because Nvidia pays them?
Wouldn’t be the first time they water down the definition of open for big tech.
Replying to
@krinkle@fosstodon.org @chanakya@social.screamingatmyscreen.com
“Customers asked for it”
“we under estimated the impact”
Yep. All that reads very much like something,…
I’m not pointing at the engineer here. It’s Microslop we are talking about and I don’t think this is the hill he or anyone would die on.
Replying to
@brettcannon Based on the discuss. thread I agree that a system not requiring trusted publishing is what we should be working towards to.
With the arguments in the thread it’s just hard to see how. Many of the properties of a “trusted” way of doing things discussed are inherently incompatible with any new development.
Replying to
@chrisg@fosstodon.org it’s part of it, I think the other points are
- it scales nicely vertically and horizontally, something we don’t often see
- there aren’t many use cases where anything else outperforms a SQL database
- there are nearly none (except very specific use cases and scale) where a SQL DB doesn’t perform
And some of the special cases are covered by extensions such as PGvector and postGIS.
(Unrelated - Hugos standard RSS location is hard to guess on mobile, needed my desktop to inspect source)
Replying to
@brettcannon sorry, I wasn't very clear with my comment. I was referring to this comment https://discuss.python.org/t/new-oidc-providers-for-trusted-publishing/106334/5
[...] 7 years isn’t a large amount of time to be able to look back on [...]
Replying to
@brettcannon per the thread there doesn’t seem to be a way to establish any new providers outside of the blessed ones when 7 years is considered “too young”. The technical / operational concerns are valid IMHO.
The maturity argument also falls very short when you objectively look at the trajectory of a blessed provider with outages piling up.
The example with SourceForge in the discussion is a perfect example for that and also why blessed providers are not the way to go.
Replying to
@chanakya@social.screamingatmyscreen.com you will fail HIPAA once your “process” is audited anyway.
The service will most likely hold up better… these people use HIPAA as boogeyman and don’t know more than the abbreviation don’t they?
Replying to
@chrisg@fosstodon.org usually I’m looking for /rss or /feed - preferably would IMHO just be a link somewhere in the header or footer :)
Replying to
@brettcannon also sorry for the less civil part of the comment.
I didn’t know you are at MS and that the blog post was sponsored / written as part of your job, I usually try to be more factual when engaging with content like this.
Replying to
@elisse if there’s a free form text field just add “I’m tsudere. If you bully me I’ll cry and I’ll like it.”
Let’s see what their hiring manager and their LLM makes out of that *scnr*
Replying to
@miketheman@hachyderm.io how does this related to a script downloading more code for example and executing it? Which is IMHO one step worse than obfuscation?
Replying to
@chanakya@social.screamingatmyscreen.com not written by humans, not meant for humans.
Replying to
@miketheman@hachyderm.io thanks for the work you put in there!
Why is obfuscated code and downloaded code treated differently from a policy perspective?
Replying to
@webology@mastodon.social @paka@mastodon.scot the study IMHO highlights how much narrative vs actual research happens on both sides.