I'm old enough to remember when smoking around others was socially acceptable, in restaurants, bars, workplaces etc. Secondhand smoke was everywhere. We changed norms once we accepted that shared air = shared risk. Airborne viruses are no different
Things like student loan debt is great, because it puts workers into massive debt before going into the workforce, which gives capitalists leverage over us.
But really it doesn't go far enough. We should just be born with debt. Babies have been getting off scot-free for far too long.
"Open Printer will use the Creative Commons BY-NC-SA 4.0 license for all of its files, including electronics and mechanical design files, firmware code, and the bill of materials"
So it will be illegal to manufacture it for a friend, manufacture replacement parts and sell them etc. ...
"Open Printer is an open-source, repairable inkjet printer designed for makers, artists, and anyone tired of throwaway hardware. Built with standard mechanical components and modular parts, it’s easy to assemble, modify, and repair.
Well, that was interesting. Discovered today under Windows settings, a storage management interface. Just recooped 25 gigs and my PC is still happy. It's indirectly mentioned in this article. This hidden Windows folder can grow forever — here’s how to clean it safely: makeuseof.com/windows-installe…
However, I figured, if I go there, why not order some #Catima stickers. After all, Catima is an app available on IzzyOnDroid :) They just came in and they look nice :)
Make sure to ping me at FOSDEM or visit the IzzyOnDroid stand to get one!
@higgins It's a little app to store event/plane tickets, (grocery) store discount cards/coupons and other things with barcodes. And sure, anyone who wants can have a sticker while supplies last :)
Tired of loud, inconsiderate, annoying neighbors?
Neighbor Revenge is your go-to channel for non-confrontational, creative solutions to everyday disruptions.
@Cookiemonster Ja. Ist ja auch von meinen Empfehlungen der mit den meisten views. Ich glaube für jemanden der sich mit der Materie ein bisschen auskennt ist wenig neues dabei. Aber auf jeden Fall gut vorgetragen und für jemanden der nicht täglich über Satellitenkommunikation nachdenkt durchaus interessant.
the "don't look up" one was fun from an overview perspective, but I wish they had explained the "satcom people intentionally violate the standards in all these ways" slide in detail
I'm guessing some of it may be described in the paper, but I'm afraid listening to a rant by a passionate person is not an experience that is easy to replicate in written text...
Švýcarská společnost Punkt nějakou dobu nabízela chytrý mobil s modelovým označením MC02. Jeho hardwarový výkon neoslnil, ale nabízel operační systém s důra ...
In response to the articles circulated recently about people who are switching to Linux as their desktop operating system, I note that my first desktop Linux experience was in 1998. I investigated hardware compatibility, bought a new laptop, and a friend installed Linux on it for me. (The installation process wasn't accessible with a screen reader in those days.) I used Emacspeak to provide a spoken interface to the operating system. Unfortunately, my braille display wasn't supported - eventually fixed in the mid 2000s. Graphical X11 applications were not accessible either, but that, too, was rectified in the mid 2000s. In the late 90s, I didn't need GUI tools anyway. The modern JavaScript-intensive Web has changed that situation fundamentally.
Non-visual access to Linux continues to improve, albeit slowly and with highly constrained development resources. Although I've had (and continue to have) other operating systems, I still consider Linux my primary desktop computing environment. #Linux #ScreenReaders #accessibility
Want to develop your own chat app that works via Bluetooth, Tor, and pigeons? Or do you want to extend Briar to work via LoRa, Wi-Fi Aware, I2P, WhatsApp, whatever? Or Android is constantly suspending your app in the background?
I've written down some pointers to Briar's source code and research with pieces to start with. Inspired by an amazing #39c3, because sharing is caring.
Not yet, but this is planned for the future. For now, one of you will need to manage (pay) both accounts. If you don't use your own domain, you can also upgrade two separate accounts, it's the same feature set.
Európsky parlament sa zaoberal otázkou tzv. vojenskej mobility v EÚ, teda schopnosti členských štátov rýchlo a efektívne presúvať ozbrojené sily, techniku a materiál cez hranice. Táto téma nadväzuje na zhoršenú bezpečnostnú situáciu v Európe a skúsenosti z posledných rokov, ktoré ukázali, že EÚ nie je na rýchlu reakciu v krízových situáciách vždy dostatočne pripravená. Parlament preto zdôrazňuje potrebu lepšej koordinácie medzi štátmi a odstránenia praktických prekážok, ktoré dnes presuny výrazne spomaľujú.
Jedným z hlavných problémov sú administratívne a technické bariéry. V súčasnosti môže presun vojenskej techniky cez viacero členských štátov trvať týždne až mesiace, najmä kvôli rozdielnym povoleniam, pravidlám a nedostatočne prispôsobenej infraštruktúre. Parlament poukazuje na to, že mnohé cesty, mosty či železnice nie sú dimenzované na ťažkú vojenskú techniku, čo výrazne obmedzuje schopnosť rýchlej reakcie v prípade krízy alebo ohrozenia bezpečnosti.
Dôležitým prvkom je aj spolupráca medzi členskými štátmi a koordinácia s medzinárodnými partnermi. Parlament zdôrazňuje, že posilňovanie vojenskej mobility má prebiehať v súlade so spoluprácou v rámci NATO a pri plnom rešpektovaní suverenity jednotlivých štátov. Nejde o vytváranie „európskej armády“, ale o lepšie fungovanie existujúcich obranných kapacít a ich schopnosť rýchlo reagovať v prípade potreby.
Zlepšenia v oblasti vojenskej mobility nemajú význam len v čase konfliktu. Investície do infraštruktúry, zrýchlenie administratívnych procesov a lepšia koordinácia môžu mať pozitívny vplyv aj v mierových časoch, napríklad pri humanitárnych zásahoch, krízovom manažmente či civilnej ochrane. Parlament preto vníma túto tému ako súčasť širšej stratégie posilňovania bezpečnosti a odolnosti Európskej únie.
Postoj Európskeho parlamentu má v tomto prípade politický charakter a slúži ako východisko pre ďalšie rokovania s Európskou komisiou a Radou EÚ. Cieľom je postupne pretaviť tieto odporúčania do konkrétnych opatrení, ktoré by v praxi zlepšili pripravenosť EÚ reagovať na bezpečnostné výzvy, bez toho, aby sa znižovala kontrola členských štátov nad vlastnou obranou.
O Brne sa hovorí, že je to konglomerácia nezávislých dedín spojených električkou. Európska únia mi často krát pripadá ako konglomerácia nezávislých štátov spojených Schengenon 😕
Sme susedia, čiastočne súrodenci, ale nakoľko sme aj spolupracovníci?
Na veľa veciach strácame hromadu peňazí, času, energie len preto, že ich nekoordinujeme a každý sa hrá na vlastnom piesočku.
Čo sa týka ozbrojených síl, je už po najvyššom čase dať aspoň niektoré armády dokopy a vytvoriť európsku armádu. Na začiatok nemusí byť plná, stačilo by pár dôležitých krajín. Podobne ako si Schengenon, eurom a vlastne aj európskou úniou. Kto by chcel , ten by sa pridal.
„Many weak and few strong links“ seems to be a common pattern in many ecological networks. 1st author Franziska Koch shows that this pattern can enable stabilising effects of network structure, using competitive hierarchies as a case study. We argue that skewed link strength distributions should hence receive more attention, especially in studies based on #RandomMatrixTheory.
Ecological networks tend to contain many weak and only a few strong links. Furthermore, link strengths are often patterned within a network in ways that enhance system stability considerably, increasing the ability of the system to return to equilibr…
340. That's the number of brand new billionaires we made in 2025. Roughly one a day and more than any other year.
9 million. That's the number of people who died of hunger in 2025. Over 24,000 every single day for an entire year.
40 billion. That's the number of dollars it would cost to end world hunger. Not to alleviate it. Not to lessen it. To eradicate it.
11. That's the percentage tax we would have needed to impose on ONLY new wealth on ONLY billionaires in ONLY 2025 in order to eradicate world hunger. Eleven percent. That's all. Compared to the vast coffers of wealth they've already accumulated that's a rounding error.
you need to come up with a ridiculous prize for attendance ... for those hardy souls who survive all the way to closing talk - maybe gold medal giveaway ?
als es um die chatkontrolle ging, konnte nicht genug getrommelt werden. Hier geht es um viel mehr. An fehlenden Accounts kann es nicht liegen, dass die 30000 immer noch nicht voll sind. Mitzeichnung ist auch ohne deutschen Wohnsitz möglich. Keine Führung eigener Register zur Erfassung von trans* und nichtbinärer Personen epetitionen.bundestag.de/petit…
2025 had ups and downs, but I loved my 2025 at #IzzyOnDroid.
Our team has grown with several new members and we launched izzyondroid.org/, our new web home!
We even got our first grant ever (thanks, @nlnet!), allowing us to bring download statistics to IzzyOnDroid and integrate it directly into #NeoStore (and soon #Droidify), improve the #ReproducibleBuild system and more.
I'm grateful to the team for giving me a space where I feel able to make a difference ❤️
@IzzyOnDroid I won’t be at FOSDEM, but one of our advisors, @aeva will be. They can explain the power of Artifact Dependency Graphs… the underlying technology that Spice Labs uses
On Christmas Eve I will take on the toughest cycling challenge of my life…I will attempt to ride 5,000 km on Zwift in eight days, for a new official world re...
@mireeek No brutál ty kilometry. Tady v tom je Zwift super, ta komunita je tam hodně dobrá. Sice jak PC hra, ale hrozně moc možností. Teď postupně hledám co tam vše jde.
Já se družit nepotřebuji, jsem individualista, klidně budu ve své tělocvičně šlapat sám. 😁 Mimochodem, jak jsem se zde dověděl, my introverti máme dnes svátek. 🎇
@mireeek Já nemyslel chat, ale můžeš na společné výjezdy, jedeš v háku atd. Samozřejmě každému vyhovuje něco jiné. Ještě jsem to nezkoušel, tak ani nevím jak to funguje.
the AI slop in security reports have developed slightly over time. Less mind-numbingly stupid reports now, but instead almost *everyone* writes their reports with AI so they still get overly long and complicated to plow through. And every follow-up question is another minor essay discussing pros and cons with bullet points and references to multiple specifications.
We need another AI trained on Linus' kernellist posts.
"Please rewrite this slop into something short and concise, as if Linus had sent it"
...
Hopefully the AI deluge with end when the novelty wears off and people realize sounding like you work in marketing and customer care isn't the way to communicate tech.
Sounds like AI made the reports smarter but not necessarily easier to read; trading mind-numbing errors for brain-numbing length 😅. Sometimes I miss the days of short, punchy reports!
Oh, yeah... no. That "issue" has been known about for literal decades. Some probably consider it a feature. I don't personally, but I can understand their perspective, and yeah... this is the caller's responsibility IMO.
you mean because of my few complaints they would start doing 20 line responses instead of 400 line ones? First: why would that be bad? But then: I'm absolutely sure they don't adapt to my few comments in a world drowning in content.
On the first point: I imagine that an LLM improving in style is way way easier than improving in substance. So if LLMs read your posts and improve in style, you'll get the same substance-slop but harder for you to detect. That's bad for you.
On the second point. I'm not so sure about that. I'll come back on this point.
I'm not fighting the use of AI. I'm fighting users making my life harder. If an AI can explain their issue in 20 lines and the issue is legit, that's a pretty big win.
> poisoning attacks require a near-constant number of documents regardless of model and training data size. This finding challenges the existing assumption that larger models require proportionally more poisoned data. Specifically, we demonstrate that by injecting just 250 malicious documents into pretraining data, adversaries can successfully backdoor LLMs ranging from 600M to 13B parameters.
To me this suggests that an LLM could indeed learn from one specific individual if it deems that individual "important" enough.
In that article they talk about triggering on a "specific phrase", but an LLM has its own internal representation of the text, so in principle might trigger on different things like specific phrase plus the name of the user on a certain website, or something.
(See how I cleverly did not mention AI in the title!) You know we have seen more than our fair share of slop reports sent to the curl project so it seems only fair that I also write something about the state of AI when we get to enjoy some positive a…
Agreed. But as a non programmer trying to submit to any project is mostly replied by "Not enough data, what about FunkyMethodYouCantKnow()?" - AI helps mitigate this.
Either developers accept the fact that there are incapable human beeings on the other end OR they accept its a LLM. Whats not going to happen: 100% qualified bug reports.
I'm not sure people understand how cognitively tiring these things are until they have had to deal with it themselves. When they first started appearing as bug reports I wasted a lot of time trying to figure out where an issue was, or even if there was an issue at all. But now if I see an LLM generated report my brain just skips it over which isn't always the right thing to do either.
yeah, I've come to do that as well and I skip straight to asking follow-up / clarification questions directly instead of trying to read that wall of text
“Finally, LLM-generated prose undermines a social contract of sorts: absent LLMs, it is presumed that of the reader and the writer, it is the writer that has undertaken the greater intellectual exertion. (That is, it is more work to write than to read!) For the reader, this is important: should they struggle with an idea, they can reasonably assume that the writer themselves understands it — and it is the least a reader can do to labor to make sense of it”
Proposal: for your followup requests, you say "you can use AI to start your followup, but what you send must not have the standard AI pros and cons, nor silly bullet points; otherwise, it will be ignored". That is, let them use the tool they know, but force them to also do some editing.
I think once people have to pay the real cost of using AI, the slop will subside, but I think we're probably a few years away from that light at the end of the tunnel
We should talk about the availability of basic medicine without doctor’s prescription. Basic painkillers, nose sprays, typical medicine for seasonal viral infections.
This is not available in Austria without prescription and I’ve had enough of this paternalistic bullshit attitude.
I don’t believe doctors want to clog up their waiting rooms with very simple cases of flu/covid infections.
I think one reason we still have this is without it doctors visits would be just about the paperwork.
This really is bs. I'm glad ours aren't like that. I can't live without painkillers and a nose spray some days... Like, that's basic stuff, why should you be forced to go to a doctor for it? Or interact with one at all? if their goal is to give people more reasons to go to the doctor, maybe they did it. Most stuff here doesn't require a prescription unless it's something very specific, antibiotic, or stuff that helps some mental conditions.
I’ve just created a new Django starter pack on FediDevs 👀
It collects non-human accounts only: Django projects, conferences, podcasts, news, and community initiatives. No personal accounts, just things you may want to follow once and forget.
@kboever Guten Morgen, und danke, euch ebenfalls ein frohes neues Jahr! Bei uns soll es im Lauf des Vormittags auch stürmisch werden. Aber es soll auch weiter Regen und Schnee von oben geben, die Temperaturen nur so um die 2° heute im Lauf des Tages.
@fasnix Danke Dir! Ja, ist wirklich sehr selten, und im Moment schneit es auch noch weiter. Vorhin gab es eine große Schneeparty mit Kindern, Eltern, Schlitten, vielen Schneebällen und viel Spaß.
I just tagged a new release of slidge-whatsapp, the (objectively best (unbiased opinion)) #XMPP to WhatsApp gateway. It fixes a lot of long-standing issues and starts much faster than the previous releases. Enjoy!
Oh, so is that working well? Years ago I tried a WhatsApp-XMPP-Bridge and it made Whatsapp lock my account due to policy violations or something like that.
@denkbeteiligung it's against WhatsApp ToS so I cannot guarantee anything. We're niche enough that they don't care IMHO, as long as you don't spam/scam people. It pretends to be the WhatsApp web client. I had my account locked for 48 hours once in 4 years of use, while experimenting with group creation which I guess made me look like a spammer; it has not occurred since. We have not had such locked account reports from users. Experimenting while developing is the "dangerous" part.
We did not see a Porsche in your neighbourhood, thus we concluded this must have been a mistake. AI summarised the delivery note and we didn't have time to look into it further.
Paul
in reply to Sylvia • • •Andy Piper
in reply to Sylvia • • •IzzyOnDroid ✅
in reply to Andy Piper • • •higgins
in reply to Sylvia • • •Sylvia
in reply to higgins • • •