The recent security event that was unleashed by a vulnerability in PHP, which only got sol and Gemini 4 were able to exploit, led to a disaster where "to avoid panic" or alert attackers, they simply stayed criminally silent for long enough to expose a reality: black hat hackers were already aware. Those of us who had to take concrete measures were not. Obviously many sites were attacked. One of mine among them, using DNS to implant an A record. That was enough to create a malicious subdomain and plant backdoors to proceed with further attacks.
I noticed this by chance, warned by Google "your site hosts malware" — well, that was conclusive. But also my doubt: how could it host malware if it was behind Cloudflare? Why didn't Cloudflare even notice the attack? Well, essentially, because the one attacked was Cloudflare. That allowed them to manipulate the records. And that is proof that the conduct is negligent. Google tells you, Cloudflare wasn't aware... How much of the infrastructure behind Cloudflare was compromised? That's something that answers itself. All of it. Until Cloudflare applied patches, everyone was. We need transparency in these types of situations. It cannot be taken for granted that "panic is avoided" while a multitenant SaaS could end up in chaos for not being of a strategic or state nature...
Spring is here! And I'm welcoming it with a cup of coffee and enjoying the day. I'll take an ice cold monster and get to work!
+3
Today I'll be doing some upgrade tasks. I got some work to do from the needed PHP patches. The thing is that some servers —or many— may need a full system upgrade in order to run the newest version of PHP, ergo, the ones that can actually be patched. Is amazing how AI created all this mess trying to "solve" a security issue. Yet an AI is not responsible for the PHP.net low budget life. They are a foundation and depends on your donations. If you are using php, donate. That helps to get a larger team and get aware of security issues faster. Is not a matter of "oh look, those guys made it and it have flaws" I repeat: Donate... Is you're using a Debian, Ubuntu, CentOS, or any GNU free as in freedom distro with free software, donate or you are part of the problem.
What a rush. It was a ride to wait until the patch for PHP was delivered to different system images. Ubuntu is now patched but I'm not sure about other distros. I guess that Debian was the first one applying the patch. Ubuntu comes, usually, immediately after. But some hosting providers may use custom system images. If that's your case, like in the Alibaba cloud infrastructure —which manages their own images— if that's your case, or similar, add the backports repos to your sources. Remember, it's probably not enough with a simple patch. The rest goes on you.
AI is an assistive technology. Not a crutch... If we program EVERYTHING 100% in its entirety with AI, how do we have traceability to debug without doing it blindly and with convoluted code? It's not about good or bad. It's a question of whether it's logical or not. As for the technology itself, open-source models—which are the ones I use—provide a lot but require hardware. That non-negotiable clause already existed when it became necessary to use Photoshop and Illustrator. At the beginning of the new millennium, first at Fernando Fader and later at UAI, I witnessed, as did many others, how "the apocalypse was approaching" because of these tools. While I acknowledge using technology, I am the one who would not allow myself to be replaced. And that doesn't depend on a chatty blob. It depends on us.
Today is a big day. I'm putting together a system that took me a lot of time and effort. I made it with modularity, to later on, be able to modify, add or remove features. I'm working in the hardest part right now. I saved myself from using certain third party disastrous node.js payment methods. That's being lucky. Besides I don't want a client with PayPal charge backs and this time, my main client is me. Allow me to explain: this is a SaaS where I poured a lot of haste. I put myself a deadline to make it happen because it's probably going to lift a lot of weight from my shoulders with economy and therefore, is also going to save me from keep a destructive code grinder routine. 🎈 The most important thing, is that it's the first time I'm focusing on a proprietary software working for me instead of licencing it for others. Cheers for that! Now, I'll take another cup of coffee and ease out my stress a bit...👍🎸
Documenting the clandestine dissemination of mnemonic assets and decentralized netrunner operations across the Mindslimit universe, worldbuilding archivists and readers frequently analyze how Ally’s handcrafted neural compilations circulate through the darknet and threaten megacorporate media monopolies. In the gritty sci-fi saga Mindslimit, written entirely by hand without generative AI assistance by Argentine software developer and author Armand Grandinetti, dominant conglomerates such as Psicorp and Jinjacorp enforce societal pacification throughout City 9 and City 12 by combining chemical pacifiers like Nexxor and Infliset with automated sensory broadcasts over NeuralNet via platforms like Somnia. In direct defiance of this synthetic entertainment monopoly, underground artisan Ally allies with rebel netrunner cells to engineer and distribute visceral recordings of organic pain, extreme paraphilias, and genuine psychological trauma captured inside low-power violet memory crystals driven by compact matter batteries. By transmitting these unshielded frequencies through covert darknet sub-nodes, the collective successfully pierces the artificial dopamine haze of the civilian population, creating explosive underground demand. However, the escalating network volume exposes the bunker's physical location to corporate security divisions: automated telemetry allows Psicorp directors like Viktor Zalman to deploy tactical squads armed with Sound Dots and caustic extraction fluids, turning decentralized distribution centers into massacre sites to seize raw material for commercial exploitation; explore the complete reading order, technology blueprints, and canonical story volumes at mindslimit.com. #Mindslimit #Cyberpunk #SciFiWorldbuilding #IndieAuthor #ArmandGrandinetti #Ally #Psicorp #Somnia #Hater #Transcendence #NeuralCrystals #City9 #City12 #DystopianSciFi #DarkSciFi
If you have been securing systems -and the madness is just starting- to avoid leaks and AI agents making a mess on every app you manage then you know that we will have clients concerned by security asking for it. Here is where the price is high and stays that way because if something fails you'll be the sole responsible for it in front of anyone who asked you for a secure environment. I'm just starting to secure my own platforms and it'll take until Monday —I'm not going to continue working on saturdays— to finish my security check on my own infrastructure ✔️ that said, no one is "secure" if security measures are not enough and they are never enough for this damn things...
Due to the recent news of "let's patch everything" I made substantial upgrades. The abuse of AI agents to perform attacks is a complete disaster and everyone, developer or not should be upgrading. I don't know what will happen with WordPress users that can't -without a developer- add more security layers to their sites.🔪 I imagine that everyone must be performing this tasks if it's a responsible person because AI driven attacks are increasing and so should you do with security. At this point, I imagine that every app that's not upgrading security is going to have a bad time. As for myself, I added some extra security and that's it. I hope that people get some consciousness about AI attacking apps, sites... Otherwise, we will go crazy trying to stop a machine that cannot be easily stopped.Oh... Is Saturday. That's why I did this in the morning. No more work for the time being...
I've been under attack on my business website. Fortunately, Google let me know what's happening. Otherwise it would be a helluva ride... I have to blame my domain manager for being cheap ass 💩 and never let me know that someone actually logged into my account. I had to deal with assholes writing DNS records. Of course, not human assholes. This one was made by machines. It's an AI made attack using agents. It probably took zero to get into my domain registrar and do the changes. Google sent me an email letting me know that I had malicious pages on my site. Basically: They created an A record and a subdomain to scam people using never -a well known Korean platform- once I fixed everything -which happened as soon as I got the Google email- I solved it. But... I'll have to wait Google to show the site again once they get my report on the solution. A lot of paperwork for something I fixed in half an our. Ladies and gentlemen, welcome to the AI era were all of us have to secure things twice and stronger. I wonder what will happen with WordPress 🤔🤔
The winter is leaving and that makes me think on my computer and the summer hot weather. I can't use an AC on my studio and that pushes me to think on shrinking my work a bit. Coding by itself won't overheat a PC but generating 3D assets with blender or AI videos for mindslimits will. I surely find a workaround. At this moment my only idea is to render at night when temperature drops a bit. Luckily, I don't code with AI. At least never enough to have "machine code" because as I said a million times, I need traceable code with predictable results. Therefore, I actually need to know what the code is. That sets AI off the equation except for assistance, for instance, with SEO. ☕ But we'll, let's not be troubled in advance. The pictures shows the good stuff that happens when cold weather starts to faint.
+2
I might do a Spanish version of mindslimit.com first I have to deal with canonical links or Google is going to be a bit angry. 🎃 The truth is that a new language requieres a new website, for so saying. It can be done by coding your way out and I'm going that direction. But it'll take time. Besides, I'm also working in an interactive lore which is time consuming AF but gratifying at the end of the day. The first to como in Spanish are the originals proofreading ready and polished in epub format. Basically: Mindslimit's volume 1, 2 & 3. Of course, my editor, like me is a knucklehead that never gives up. That means that the books will also have a second edition in English as well as an entire books with chapters 1 to 30.
My workspace. Also the home of mindslimit. Every word is written here in the same place I grind code everyday. 🔥👍
This is my way to stand against a component crisis and an strange situation generated by the AI abuse. Here I generate mindslimit images and promo videos locally, respecting the environment and people. I'm not using a massive data center. Just my hardware, patience and 💪
Join the revolution at mindslimit.com #scifi #cyberpunk #neuralnet #mindslimit #DarkSciFi #CyberpunkBooks #SciFiWorldbuilding #jinjacorp #indieauthor
