Ploum

Sounds already too complex for me. I wrote my post as Gemtext (the Gemini markdown) and wrote a python script that convert it to HTML and generate an index page. Also, I put the CSS in each page because I’ve 42 lines of CSS and I did the math to show that having 42 more lines in each file would never cost more than having a second http request. Honestly, it was a lot simpler than expected, without using any python library or any template. The hardest part was generating the email to send to the list. I spent month analysing different frameworks, different static site generators. I spent week trying to find a good CSS. According to the git history, it took me 2 weeks of hacking to get it done including importing 18 years of Wordpress history. And without having to learn anything because it was straight python. And the best of all is that I can easily change/correct if I’ve a bug. It’s my code. We, computer scientists, forgot completely that we could simply code adhoc solutions instead of trying to reuse everything. We take more time to learn and adapt than to do. https://git.sr.ht/~lioploum/ploum.net for the code (the site itself is https://ploum.net)

But they aren't useful for that. Terry Tao uses them to improve his ability to use poorly-documented boilerplatey things like Lean and matplotlib, but receiving advice from them‽ Frankly, if a chatbot is giving you much better advice than a rubber duck, you're either a Jack-of-all-Trades (in which case, I'd recommend better tools) or a https://ploum.net/2024-12-23-julius-en.html Julius (in which case, I'd recommend staying away from anything important). I recommend reading his interview with Matteo Wong, where he proposes the opposite: https://www.theatlantic.com/technology/archive/2024/10/terence-tao-ai-… > With o1, you can kind of do this. I gave it a problem I knew how to solve, and I tried to guide the model. First I gave it a hint, and it ignored the hint and did something else, which didn’t work. When I explained this, it apologized and said, “Okay, I’ll do it your way.” And then it carried out my instructions reasonably well, and then it got stuck again, and I had to correct it again. The model never figured out the most clever steps. It could do all the routine things, but it was very unimaginative. I agree with his overall vision, but transformer-based chatbots will not be the AI algorithm that supports it. Highly-automated proof assistants like Isabelle's Sledgehammer are closer (and even those are really, really crude, compared to what we could have).

> great companies can have an enormous upside for countries (e.g., taxes and employment) A company split into multiple still provides taxes and employment. And the employment conditions tend to improve, since there would be more companies competing for employees. > too strict actions may hamper innovation by signalling to innovators that you will be punished if you achieve too much And monopolies hamper innovation by reducing the competitive pressure to innovate. Necessity is the mother of invention. Let's look at what kind of "innovations" Microsoft used to grow - pushed an over-complicated OOXML standard, and implemented it in their products incorrectly, to make it harder for LibreOffice to interoperate, since they were forced to follow a complex standard and reproduce Microsoft's bugs: https://ploum.net/2023-06-23-how-to-kill-decentralised-networks.html There are countless similar examples, going back to railroads charging more to carry competitor's goods.

> I stress to them that I admire the heroism of their generation in resisting these temptations. > (I think older generations would do well to recognize this fact.) That is so, so much better than the default mode of blaming kids for getting addicted to the systems their older generations developed. https://ploum.net/2026-01-19-exam-with-chatbots.html is another interesting perspective

Money is not a good success metric https://ploum.net/2026-01-22-why-no-european-google.html Bezos is making a lot of money. But it doesn't mean it makes the world better. Prime or AWS can still work fine without having Bezos making tons of money

Decent overview (& more broadly but the heart is about XMPP & good ol’ capitalist corpo greed): https://ploum.net/2023-06-23-how-to-kill-decentralised-networks.html

> you get hired for your proven ability to (…) No, you get hired for your perceived ability to (…) The world is full of Juliuses, which is a big reason everything sucks. https://ploum.net/2024-12-23-julius-en.html

The app is not showing posts that don't contain a picture, but shows posts that do. So if you browse someone Mastodon account on Pixelfed, you will see just fraction of their posts. https://ploum.net/2025-12-04-pixelfed-against-fediverse.html I don't think it's important at all because nobody really uses Mastodon anyway; but it shows that doing decentralised software is hard, because each actor can just do whatever they want.

Bloggers don't always publish immediately after writing a post, the start of the prior post explains the time line and the delay: https://ploum.net/2025-12-04-pixelfed-against-fediverse.html

Not that long ago, I noticed https://ploum.net/2024-12-23-julius-en.html on HN.

> Otherwise I can spread all kinds of made-up stories about my work mates, and claim they illustrate truths about human nature, right? This happens all the time, and it’s fine? A recent front page example [0] — did julius ever exist? Who knows. Would it change anything? Not really… A problem could definitely arise if you specifically called out a coworker without anonymization, but speaking in broad strokes is… perfectly fine and uneventful [0] https://ploum.net/2024-12-23-julius-en.html

Except that the blog is also in French. https://ploum.net/2024-12-23-julius-fr.html The author is running a poll to establish the plural: https://mamot.fr/@ploum/113704470821790664

See also part 1: https://ploum.net/2024-10-20-20years-linux-desktop-part1.html

Exactly https://ploum.net/2023-06-23-how-to-kill-decentralised-networks.html

Just as the article [0] have described it, the web is already split. You can either choose Times Square or the scenic retreat, the consumer magazine or a nice book. I choose to minimize my time on the "normal" web which breaks if I block ads and tracking scripts. And instead spend it here and in my RSS reader. [0]: https://ploum.net/2023-08-01-splitting-the-web.html

Because of what Google did with Google Talk. https://ploum.net/2023-06-23-how-to-kill-decentralised-networks.html XMPP is underrated. A lot of people are imagining Pidgen in 2011, but the protocol has been extended, the actively developed clients are good, and it avoids the heavier parts of Matrix (both client and server side.) I wouldn't be surprised if Slack's replacement when Salesforce inevitably fucks it up will be XMPP based rather than Matrix.

Maybe we want it to be a niche place? Also, the effect of Google's presence on XMPP was huge, see https://ploum.net/2023-06-23-how-to-kill-decentralised-networks.html

I think this summarizes it pretty well. Even if you don't mind the garbage, the future AI will feed on this garbage, creating AI and human brain gray goo. https://ploum.net/2022-12-05-drowning-in-ai-generated-garbage.html https://en.wikipedia.org/wiki/Gray_goo

Google absolutely controls communication. Two examples, Example 1: GMail. Google uses the mass adoption of GMail as a cudgel. It's no longer possible to independently host email without being marked as spam or rate limited by Google, https://tutanota.com/blog/posts/gmail-independent-email Example 2: Google did a good old "Embrace, Extend and Extinguish" on XMPP, https://ploum.net/2023-06-23-how-to-kill-decentralised-networks.html

Is more likely that XMPP was killed by Google instead, and Meta is trying to do the same over ActivityPub https://ploum.net/2023-06-23-how-to-kill-decentralised-networks.html

There's a lot of discussion about that! Here's a very good article on the EEE threat. https://ploum.net/2023-06-23-how-to-kill-decentralised-networks.html Personally I think it's more an "embrace, extend, and exploit" approach; a decentralized model could work well for Meta, for example if they do revenue-sharing on ads hosted by other instances (think Disney or LA Lakers).

A loosely related article that was recently discussed: https://ploum.net/2023-06-19-more-rms.html https://news.ycombinator.com/item?id=36389805

I will link to the article that changed my mind on this: https://ploum.net/2023-06-23-how-to-kill-decentralised-networks.html It was recently discussed here at HN: https://news.ycombinator.com/item?id=36456233 TL;DR: it might be the 'Embrace, Extend, Extinguish' strategy.

https://ploum.net/2023-06-23-how-to-kill-decentralised-networks.html yes, yes... let's cheer at Facebook doing things right...

How would you design a protocol immune from all the politics? Whatever the implementation, it's just a part of a larger complex system. This is a very hard issue, and trusting Meta or the likes is naive and dangerous with destructive consequences. https://ploum.net/2023-06-23-how-to-kill-decentralised-networks.html

The author was capable of using the word "embrace", but wasn't clueful enough to meaningfully use the phrase "embrace and extend". So I can't accept his essay as "getting it". A better article would be: "How to Kill a Decentralised Network (such as the Fediverse)": https://ploum.net/2023-06-23-how-to-kill-decentralised-networks.html ...where the historical example of what happened to XMPP - where Google embraced and extended it - was far more relevant.

Here is a summary of what happened: https://ploum.net/2023-06-23-how-to-kill-decentralised-networks.html And honestly, the article this submission is about, is pretty weak. > I don’t worry about Meta embracing ActivityPub and then extinguishing the fediverse as we know it. "I don't worry" is not a very convincing argument.

I found the article that motivated me to switch to bépo back in the time. It also also suggests a method that I used to do the transition: https://ploum.net/216-le-bepo-sur-le-bout-des-doigts/ (in French)

Also, keep in mind that patents don't mean the invention works, and they certainly don't mean the invention is useful for anything. https://ploum.net/working-with-patents/ > Note that by « valid », I mean that the Patent Office didn’t found a trivial prior art for this. It doesn’t mean that there is no prior art or that I’m the real inventor or that my invention works. An example of an insane patent: https://www.google.com/patents/US6960975 > Space vehicle propelled by the pressure of inflationary vacuum state US 6960975 B1 (via https://www.metabunk.org/do-patents-mean-the-invention-works.t194/ )