Nicola Romanò @nicolaromano@qoto.org

What’s the biggest single crystal you’ve grown? David Boyce and students at Queenswood School have grown a 3kg crystal of copper sulfate. Now he shares his tips for how you can do it too.

Via Chemistry World on Bluesky.

#Chemistry https://www.chemistryworld.com/culture/how-to-grow-an-enormous-single-crystal/4022946.article

Do you have a vision for a Workshop but don’t have the time to arrange the logistics or raise the funding? Submit your idea to us in our annual call for topics. We'll provide all necessary support for successful applications. Find out more at https://biologists.com/workshops/propose-new-workshop/

The war waged by the tech authoritarian oligarchy against the media has reached a new level:

#Palantir is suing us. Us, the Republik Magazin.

A small Swiss media company, funded by readers, founded in 2018 and free of advertising. I am not aware of any other media company globally that Palantir is currently targeting so aggressively.

What is this about? Together with my wonderful colleagues at the WAV research collective Jenny Steiner, Lorenz Naegeli, Marguerite Meyer, and Balz Oertli, we published a two-part series on Palantir's activities in Switzerland on December 8 and 9.

Using an extensive corpus of documents – which we obtained thanks to the Freedom of Information Act – we were able to trace a sales campaign over a period of seven years. Palantir tried to get in with many federal authorities – and was rejected everywhere.

And we also found out that the Swiss Army Staff evaluated the products and came to the conclusion that the army should refrain from using Palantir products.

Among other risks, they feared that data would be passed on to the US authorities.

Palantir is not just any company. ICE uses its products to hunt down migrants in the US. The Israeli army IDF uses the software in its Gaza offensive. The British health authority NHS has made itself dependent on the products for data analysis during the pandemic. And CEO #AlexKarp displays inhuman and aggressive rhetoric towards Europe, while the company itself advertises the “optimization of the kill chain.”

These are all facts, repeatedly verified and published by renowned media outlets. Our research relating to Switzerland and Zurich is based on this.

In addition to analyzing documents, we also spoke to various sources – including Palantir executives here in Zurich. The quotes used were presented to them and approved. Of course, we always adhered to the high standards of journalistic work. We conducted a thorough fact check before publication.

But the company doesn't want us to write the truth.

After the US company owned by right-wing tech billionaire #PeterThiel dedicated an absurd blog post to us, claiming some misinformation (such as that they had not participated in official tenders with the federal administration, a point we never claimed. On the contrary: we spoke from the outset of attempts to establish contact, sales talks, informal meetings, business as usual), after the Global Director of Privacy & Civil Liberties (PCL) Engineering and contact person for Swiss media Courtney Bowman launched personal attacks against us in LinkedIn comments between Christmas and New Year (“partisan fear-mongering”), Palantir's Swiss lawyers demanded a counterstatement on December 29.

We rejected this demand in its entirety.

In January, they demanded the same thing again. We rejected it again.

And now we see each other in court.

But why all this?

Our research on the Swiss army report caused a huge international media response. The Guardian and the Austrian newspaper Der Standard reported on the Swiss army's rejection. Numerous financial portals and stock market magazines picked up our news (which could have consequences for the overvalued stock market company Palantir).

And Chaos Computer Club spokesperson Constanze Kurz presented our research to a huge audience at the renowned IT conference Chaos Communication Congress in Hamburg at the end of December.

All of this is making Palantir nervous.

We have now submitted a comprehensive defense brief. We can substantiate all of our findings with several documents and publicly available media reports.

We trust in the rule of law and freedom of the press in this country.

In keeping with yesterday's event “Zurich, little Big Tech City” at the Gessneralle, where we first announced this news exclusively to the audience on site:

World politics will soon be negotiated in Zurich: freedom of the press, the facts about ICE, Trump, Israel, Karp, tech authoritarianism.

The truth.

All this at the Zurich Commercial Court.

We will not be intimidated. And we will keep you informed.

I recently posted a pre-print about reasoning and large-language models....

trying to get that piece on a pre-print server itself was a bit of a journey which illustrates concerns I've had for the last 18 months or so: namely that AI will kill publishing reform.

here a blog post about it:

#academic #openScience
https://write.as/ulrikehahn/is-ai-killing-scientific-reform

Marimar Martinez was the first US citizen shot by CBP/ICE.

Agent Charles Exum shot her five times. She miraculously survived.

DHS accused her of being a domestic terrorist.

She went to court to get the body cam footage released.

It shows Agent Charles Exum driving with his gun drawn.

Saying “do something b*tch”.

Ramming her car.

Shooting her.

Hours after the shooting, Greg Bovino emailed the agent and offered to delay his retirement in light of his excellent service.

The email said “you have much left to do”.

Other agents praised the shooting.

Said they should celebrate.

Exum bragged… “5 shots, 7 holes”

Ms Martinez showed incredible courage by fighting to ensure this footage was released to the public.

Everyone should watch it. Make sure the world sees this.

She’s standing up for Silverio Gonzalez.

For Keith Porter.

For Renee Nicole Good.

For Alex Pretti.

For everyone murdered by this regime.

She’s exposing ICE/CBP for who & what they are.

Thank you Marimar.
#uspol #ice #abolishice #cbp #immigration #fascism #marimarmartinez

dplyr 1.2.0 is out now and we are SO excited!

- `filter_out()` for dropping rows

- `recode_values()`, `replace_values()`, and `replace_when()` that join `case_when()` as a complete family of recoding/replacing tools

These are huge quality of life wins for #rstats!

https://tidyverse.org/blog/2026/02/dplyr-1-2-0/

What's the most common complaint I've heard about Linux?

Not the installation process.
Not finding a distro.
Not getting programs to work.
Not troubleshooting.
Not hardware compatibility.

The most common complaint about Linux I've seen is this:
For a normal computer user, asking for help is just about impossible.

They ask a simple question and:
People respond "Did you Google it?"
People complain that the question wasn't asked "correctly".
People respond "RTFM"
People get mad??? at them for making an easy mistake.

We can't expect normal people to know to, or even know how to deal with any of that stuff.

Search engines these days are awful, manuals are hard to read for most people (especially stuff like ArchWiki), and normal people make mistakes we think are easily avoidable.

The solution to making Linux more popular is not ruthless promotion. The solution is to actually help the people who are trying to use it.

#Linux

"Getting over ANOVA: Estimation graphics for multi-group comparisons", Lu et al. 2026 (Claridge-Chang's lab)
https://www.biorxiv.org/content/10.64898/2026.01.26.701654v1

"Data analysis in experimental science mainly relies on null-hypothesis significance testing, despite its well-known limitations. A powerful alternative is estimation statistics, which focuses on effect-size quantification. However, current estimation tools struggle with the complex, multi-group comparisons common in biological research. Here we introduce DABEST 2.0, an estimation framework for complex experimental designs, including shared-control, repeated-measures, two-way factorial experiments, and meta-analysis of replicates."

Grateful to Adam Claridge-Chang for leading and pushing on this. There's institutional-wide need for change in the biological sciences when it comes to statistical handling of data. And quite the memorable acronym, #DABEST ...

#statistics #biology #ANOVA #DABEST

AI deepfakes affect people even if they know the videos are fake: nice media coverage here of https://phys.org/news/2026-01-people-swayed-ai-generated-videos.html one of our recent studies. 1/2
1/2

My book, Accessibility For Everyone, is now free and online as a website.

https://accessibilityforeveryone.site

The book was first published by A Book Apart in 2017 but it holds up! It covers web accessibility for designers, developers, content folks, and really everyone who works in tech.

ArXiv "now requires first-time posters to be endorsed by an established arXiv author in their own field. People who have previously posted in the same disciplinary section of arXiv do not need an endorsement. The move is an attempt to clamp down on a rising tide of fraudulent submissions, says University of Amsterdam astronomer Ralph Wijers, chair of the arXiv editorial council. A large fraction, he says, are generated with artificial intelligence (AI)."

https://www.science.org/content/article/arxiv-preprint-server-clamps-down-ai-slop

#AIslop

Our new pre‑print is out!

scReady – an automated and accessible pipeline for single‑cell RNA‑Seq preprocessing: Empowering novice bioinformaticians

https://wellcomeopenresearch.org/articles/11-43

@haessar.bsky.social @fionan-a.bsky.social @yiyicheng

#scRNAseq #bioinformatics

Built an oncology trial duration forecaster. The honest finding? R² dropped from 0.84 to 0.04 under proper validation. Validation strategy > model selection.
.
🔗 stevenponce.netlify.app/projects/standalone_visualizations/sa_2026-01-19.html
.
#rstats | #shiny | #pharma | #clinicaltrials | #machinelearning

I think I now know where to draw the line between "good" and "bad" #GenAI, and possibly (or rather obviously) the same for #machineLearning. It's simply whether the input data has been constructed rigorously. Put this way it's the most obvious statement ever, but somehow #BigTech have convinced us all that they advance research by recklessly scraping #twitter, #4chan and who knows what else (they keep their training data secret).

What is good science in computational linguistics? Well, open data is a step towards it. But open and crap is not a solution. We need to actually _know_ and manage the data. And nobody in their right mind would want to plough through toxic data to clean it. We've all heard the horrors of Kenyan data workers who do it for money and still suffer doing it.

But better (yes, also smaller) corpora are of interest to scholars in the humanities and the social sciences. Think of https://textcreationpartnership.org or https://mlat.uzh.ch. Yes, they are too big for individual researchers or even teams to handle, but we have the organisational and technological infrastructure to work on them collectively. We've been doing it for ages and we will continue doing it. We just need to do it together.

And this is the goal of the European Research Council project proposal I'm submitting in this very moment.

The Mathematical Center of the Venezuelan Institute for Scientific Research was destroyed in the military operation conducted last week. This is unbelievably sad. The IVIC is among the top research centres in the world and has been a major hub for mathematics in Latin America since at least the 1970s.

https://venezuela-news.com/ministra-jimenez-denuncia-destruccion-ivic-ataque-militar-eeuu-venezuela/

Solidaridad con los y las colegas de Venezuela y Latino América. Gracias por seguir luchando para desarrollar y transmitir el conocimiento matemático.

I am a Gen X man. And I want to tell you how I realised and learned to identify the “survival laugh.”

First, I noticed it in a social setting. I saw a woman laugh at a dangerously unpredictable man, not from joy, but to disarm him. Then I slowly recognised it everywhere: the forced chuckle in a professional meeting, the charming smile in media. This was never joy. This was a survival tactic. She doesn’t enjoy herself. She is deflecting a blow. She is surviving.

So I started asking myself in all sorts of different situations “Would I be laughing right now?”. The answer was almost always no. I can’t help but wonder how other men don’t see it? But the answer is simple: patriarchy.

I do see genuine, unforced laughter or enjoyment as well. But when I see it, I see it more often than not in queer communities, where we all can feel free from those heteronormative pressures. That contrast makes the truth of the “survival laugh” all the more heartbreaking. To everyone who has had to laugh to stay safe, all I can say: I see you, and I’m sorry.

#randomthoughts #society #patriarchy #safeplace #queer #misogyny #survival

The why and how of backing up your Mastodon data

Why this matters

Recently, the admin of a relatively prominent Mastodon¹ server, med-mastodon.com, decided to stop maintaining the server and shut it down without warning. This is, generally speaking, a terrible thing to do, but I will give the admin the benefit of the doubt and assume that he had some very good reason why he needed to shutdown without giving his users time to export their data or migrate their account to a new server (though as far as I’ve seen he hasn’t shared any such explanation).

Unfortunately, this isn’t the first time this has happened, and it won’t be the last. Even servers whose admins have agreed to follow the Mastodon Server Covenant, which among other things requires server admins to commit to giving users at least 3 months’ warning before shutting down, occasionally go belly-up without warning.

This could happen to your server at any time. Your server admins may be great people, but shit happens. Not only that, but your server’s admins or moderators could decide at any time to suspend or delete your account.

Some people protect themselves from this by self-hosting their own server. That has advantages, but it also has disadvantages, including time and money costs, and so it’s not for everyone. If you’re not interested in self-hosting, but you also don’t want to run the risk of suddenly losing all your Mastodon data, then you should be backing up your data on a regular basis.

How to back stuff up

The most important thing to back up is the list of who you follow; fortunately that’s also the easiest. In the Mastodon web app, go to Preferences, then click “Import and Export” on the left, and click “CSV” to the right of “Follows”.

There you can also download your lists, muted and blocked people and domains, and bookmarks. All of those can be imported into a different Mastodon server using the “Import” page linked on the left.
Set up a reminder to back up this stuff regularly!

The next thing to back up is posts and uploads. If you consider social media transient, you may not care about this, but otherwise, you can periodically click the “Request your archive” button on the export page. You’ll receive a notification when the archive is ready to download.

Note that unlike the CSV files, your posts and media are not easy to upload into a new server. Unless you want to jump through some hoops (see below), you’re just downloading them for reference.

If you’re paying attention, you may have noticed that there’s something important missing above: what about your followers? Can those be backed up? Alas, it’s not easy.

The way it’s supposed to work is when you decide you want to migrate to a new account, you configure your old account to tell your followers automatically what your new account is, and their accounts follow your new account automatically. But obviously, that doesn’t work if your old server goes away with no notice!

I’m not aware of any web tools for backing up your followers (if you know of any, post a reply!), so if this is something you care about, you’re going to need to resort to the command line. Command-line tools that support exporting followers from your server include:

toot
mastodon-archive

If you’re so inclined, you could also write code to do the export yourself; Mastodon has a rich, well-documented API.

Note that mastodon-archive also supports backing up a bunch of other stuff, so you could use it to set up periodic automated backups of all the data listed above, rather than just followers. Personally, I have mastodon-archive backing up my data automatically every night.

The toot utility mentioned above also supports downloading followed hashtags (“toot tags followed“), which you may also want to back up if following hashtags is a significant part of how you use the fediverse.

But what about all my filters?

(Hat tip to Royce Williams for reminding me here that I forgot to cover filters in the first published version of this article.)

If you use a lot of filters to optimize your Mastodon experience, then losing your filters when your account goes away could be painful, so you probably want to back them up as well.

Unfortunately, there is no way to export or import filters from the Mastodon user interface, and there also doesn’t appear to be any pre-baked command-line tool available to do the job. However, I’ve got you covered, if you know how to use a Python script:

https://gist.github.com/jikamens/6ffc52a7f85439046c32e05c7a67a906

What to do with the backed-up data

You pretty much can’t import your old posts

The bad news first: there is no good way to migrate your posts from one server to another. This isn’t just true if the old server goes away; there’s also no good way to copy posts from an existing server to a new one. People have been asking for this for many years but apparently it has not been enough of a priority for the Mastodon developers to implement it.

There’s at least one early-stage effort to build this functionality, and somebody has written a clever tool that server admins can use to do it as long as the old account is still available, but neither of these is a robust general-purpose solution.

You can still browse your exported data, look at the posts in a text editor, write your own code to do stuff with the data, etc. Furthermore. There are a number of tools for viewing and searching through your exported post archive.

How to get your followers back

If your account wasn’t migrated automatically because it disappeared out from under you, then assuming you’ve archived your followers as described above, you’re going to need to reach out to them one by one, let them know what your new account is, and ask them to follow you again.

(If you’re in online fediverse communities where people know you, you can also make a public post about your new account and ask people you know in those communities to boost it. You may need to do this several times to catch everybody, and you might still miss some people.)

If you have a lot of followers, then sending a bunch of DMs one by one to all of them will probably be a drag. You could automate it with toot, but be careful of running afoul of your server’s posting limits; you may get throttled if you post too many DMs too quickly.

Also, keep in mind that some people don’t read DMs at all, or don’t read DMs from people they don’t follow, so everyone may not see your messages asking for them to follow you again.

Restoring followed hashtags

You can use the toot utility (“toot tags follow hashtag“) to restore the followed hashtags from the list you previously exported with toot. Or if there aren’t too many you can just search for them by hand in the web app and then follow them.

Everything else, you can import

The import page in the web app’s preferences can be used to import the CSVs exported from the export page. If you used mastodon-archive or toot or whatever to export the data instead, then you can manually construct a CSV in the correct format for import.

¹This article focuses on Mastodon for clarity, but much of what is written here is applicable to any type of fediverse server.

#fediverse #Mastodon

Lokjo is a european substitute for googlemaps.

No cookies, no data, no nothing.

When you search locations by category (clothing shop, coffee bar, museum etc) it shows all locations at once. Except for corporations and chains, they get a grey dot.

https://lokjo.com

Support us if you can by a donation here:
https://www.lokjo.com/donate

Or just give it a boost. Thanks. 😊

#EUsubstitutes #maps #europe #lokjo

🧵 My sense of justice was triggered by #Palantir corporate gaslighting two Swiss investigative journalists on LinkedIn.

This is something most people won’t even see, but I was angry, so I looked while my kid was still asleep.

Here’s what it looks like when tech bros attack journalists while you and I have too much food over Christmas.

Two Swiss journalists spent a year filing 59 #FOIA requests to document Palantir’s 7-year campaign to sell surveillance software to Swiss authorities (army and health services in particular).

📄: https://www.republik.ch/2025/12/09/warum-palantir-zum-risiko-fuer-die-schweiz-wird

The Swiss army’s internal report concluded they couldn’t rule out US intelligence accessing data through Palantir systems, despite reassurances.

Their story hit The Guardian, and #UK MPs are now questioning £825M in Palantir contracts.

📄: https://www.theguardian.com/technology/2025/dec/22/mps-question-uk-palantir-contracts-security-concerns-investigation

The journalists were rejoicing on LinkedIn. It’s a big deal to have your story picked up by mainstream UK media, especially after a year of hard work.

This is where it gets ugly.