For people complaining that devRant is oh so dying and what not, it's you being dead inside. See my messages(https://devrant.com/users/retoor/...) this is full time messaging for a few hours now. It's alive and kicking, i assure you. I tried to work on something tonight but got totally distracted. Very terrible for my dev statistics since I monitory myself very well (automated).

The time I had left between messages were spent on Snek chat.

Heavy life.

This (https://devrant.molodetz.nl/Queers-...) is exactly the kinda sound of one of my former bosses made when I refused to spend a dime or a second to bring my lease car back to the company. He could pick it up, what can he do about it? Exactly, there's no one to enforce you to do that in reality and I woke him up, out of his dream world. IT'S IN THE CONTRACT!! What contact? I HAVE A SIGNED CONTRACT OF YOU HERE IN FRONT OF ME! Hmm, I'm pretty sure that I didn't sign one. Anyway what were we talking about? "WRAAAAAAHHHH".

I'm weak. I felt so bad for the guy that I brought it back with a friend making me not to have to pick it up ostream style (wasting money on the worst comfort to mankind, public transport, the transport for the confused, students and the special and on top a grandma wishing she had a better pension).

I sign almost any contact, good luck with enforcing. Most if them have "damage" - claims that won't last like thousand or even ten thousand a day. Make it a hundred, I'll sign it with the worst drawn dog in history. I've learnt from the best.

No, after more than 15 years, I still can't draw a good dog.

Forgot to give charger back, called about it. They just hang up on me. C'mon, you know what want. One more time and I put in the mail some day.

I hope I made someone spare a train ticket in the future with my inspirational story. Power to you.

EOF ❤️❤️

@Waytopranav Impressive content on the website you have in your profile, I wanted to do an article request on it, but then I saw e-mailaddress is required and I did read your terms of service.. NO WAY :P

If you're interested for some article ideas contact me at https://snek.molodetz.nl -> e-mailaddress is not required / there is no confirmation. You're attached to nothing.

I have also some questions for you.

It did the first character wrong to call a different bot and see my prompting skills to fix the issue. It takes a lot of experience to prompt like that. Blood sweat and tears people. /s

It works 14/15 times correctly and then fucks you over out of nowhere, no idea where the < even can come from.

There are too much Japanese on this site. There is such thing as too diverse. The Japanese on this site do not respect our wholesome dev culture.

Combined some AI technologies using Snek chat leading to amazing prompts and using an amazing image LLM, BordedDev and I generated new Pokemon. The results are amazing. The names are invented by AI too. It understood the pokemon word fantastic. See comments.

What is your favorite?

Left and right unite! I've spoken with a boomer today about politics and got triggered more about politics than ever before. I thought that I didn't care anymore. We have to do something about the boomer! That privileged by TWO parents raised wealthy boomer said that everything is alright. Dude, my generation is raised bad because two parents worked and the zoomer literally has no houses left to rent OR buy anymore and there are no plans building new ones at all. It's the biggest fuck you ever. If I were a zoomer I would've been blind from anger. Maybe that's why people die, their attitude makes us extinct at a certain point. He said that they don't deserve a house with their 24h working attitude. That's debatable but I think the zoomer is right, we work too much. When I had a family for a while I realized that. It's only alright to work full time if you're single and maybe in a relationship, but two parents nearly full time is too much. Not sure if the zoomer understand they but only wants to game even more but that's not the point. The work less attitude is a good thing, even I as workaholic agrees to that. I'm only workaholic because I don't have a life. I just realized that if you have a life, the amount we work is crazy. I really couldn't give the attention I wanted to the kiddo and partner.

So again, left and right, we're being distracted! Unite and GET THEM!

If the zoomer still believed in voting, they would've been extinct by now.

We had sketches ik the Netherlands "Political party against the citizen. Vote for us and fuck you!". I swear to God, boomers would vote that if it existed under the idea of "Lives so good, it can be less, it's ok". He said before or after mentioning his two cars, moped, drone and bought house with quite a tummy. Caring about the climate but not about the children of tomorrow. The climate doesn't matter much anymore if people don't take kids anymore because it's too expensive, no place to live and not enough time to raise them. Also, kidding, they don't care about climate either. Totally blinded by the 70 years of perfect life.

I don't complain about life, but saying we're doing fine as a boomer, stfu! The milennials do also not have that much to complain, but the sooner, hell yes they have. Holy f. Why would you work at all if you can't get a place to live. Maybe this is the way we extinct. Egoism.

My god. He was my example 😕 The former computer expert of the family before I got born and dominated him to the bone 😕 It was brutal 😕 He had no chance 😕 KARMA BITCH 😁

I'm in Germany.

While being the same race, they're obviously from a different world. I wonder if @Lensflare has more style than his fellow citizens.

Most Germans are Dutch without support for CSS. The country of times new Roman I guess.

They buy apple products to compensate for theur own style. That's how much money is needed to make a German hip.

Since strangely enough lack of decent site downloaders I've written one myself.

It's battle tested by downloading WHOLE devrant and a big part of molodetz. Both big sites. It makes the downloaded sites portable by making absolute urls relative.

It downloads with a high concurrency.

Reason I've made this, is because I want to have all this data is so I have a lot of spam examples to train a model on.

Project page and features here: https://retoor.molodetz.nl/retoor/.... Source code at bottom as always.

I hope someone will give it a try :)

And yes, the docs costed almost the same time as the code. Code doesn't contain unit tests, it's production tested instead. I applied many optimizations mentioned by my review tool. When i was done I was too tired for unit tests.

Drinking gluhwein 9:00 in the morning on a Wednesday. It's snows outside and I have special snow in my stash. This will be a fun day. Isn't life beautiful?

Edit: oh f, it's Tuesday. As you can see, very relaxed.

Can you read this secret message?

73 102 32 121 111 117 32 97 114 101 32 68 101 109 111 108 105 115 104 117 110 44 32 121 111 117 39 118 101 32 119 111 110 32 116 104 101 32 103 97 109 101 33

Did you ever had your dream come true?

With dream, I'm not talking about something you did rationally choose. I'm talking about something what you want with no specific reason. For me, it was becoming a programmer, as little kid already. Thank God it was nothing material. I think it's a good dream. It's achievable and solves some life issues like money automatically. I'm lucky.

I can't receive packages, nothing to do about it. If I want a package, I have to choose a drop off location in city center. Totally fine.

But now, I ordered new internet and ofcourse, they assume that your address is the place where it should be delivered and there was no place to change it. Fine, I will contact them after order.

But then, how to contact? You get a chatbot with only freaking predefined choices. "Questions about status of your order?". Etc. No fuckers, I'm sure it's doing fine. Gave up. Went to calling them.

You choose all wrong unrelated options until the automatic phone bot get stuck before you get an employee on the line. Thank God - that went fairly quick.

But it's bizar right, that you kinda have to crash an automatic phone bot by providing weird unrelated answers before you can ask your question?

Result: they can't change my address and thus it will go to delivery point end of the day if they can't deliver so I can pick it up one day later (tomorrow). Smol grr. But while said being unable to to deliver, they'll try it anyway 😁 Efficiëntcy 😁

@ostream / @antigermgerm I see there is a fight going on between the Dutch speaking and the French speaking in (s)Waffelstan (Belgium). Just send us the decent speaking (Dutch) people, we love immigrants doing cheap labor. It will be still a better life than with you French. Especially on the road. Fix your roads! 😃

Low hanging fruit.

😃 😃 😃

I found my favorite LLM I guess, it says whatever you ask it. But that first sentence, wtf :P

Past hour five spam messages came in and last 24h around 20. Many of them seem unique, I'm so interested, how is this even possible. Many of them even have a phone number attached to it. All that domains, emailadresses and phone numbers. Omg. Why all these small brands instead of a big one when all sites kinda have the same formula? It's such a mystery.

So, a while ago the Netherlands had to pay a fee of ten million for not reaching their climate goals (that are unreachable by definition i've read somewhere). Who held them responsible? Greenpeace. Who the fuck does Greenpeace think who they are? But it seems that anyone can sue the government for not doing what they say. Ok, fine. It seems that Greenpeace had the right. But now, that fee, where does that go? I mean, not to Greenpeace I hope, else i will sue the fuck about them too. Ok, so maybe some deposit regarding the climate from government itself? No idea, there's literally no journalist that asked. I've searched everywhere, where does the ten million go? Nothing. And many people feel informed by reading the news. Idiots. I don't follow news, but when I do, there's alway something wrong. I really don't know why the news is still so big, why journalists still have a job. ItS ParT oF DemoCraCy HaviNg JouRnaListS Who vaLIdaTe ThE GoveRNmeNt. Sure.

I mean.. It's literally OUR money they're wasting like this. Source of anger: I had to pay smth. And i was like, you guys already spent my money on bullshit.

@netikras a bit hard in bash maybe, but could we give the gpt bot the context of the conversation so it knows what it previously answered? In that case we can do simple follow up questions and you can ask things like "Who is right? Is it correct what x says? Is y considered reasonable? Are you sure? Give other options.". So making it more part of the conversation :)

Cleaning my system. Learned something new (do a vacuum on your journalctl logs).

Any way: biggest applications on my system in the screenshot.

Can I have an applause for google-cloud-cli for being bigger than open office?

That google-cloud-cli thing is not some advanced supertool that you installed for convenience, it's actually to get API access to certain google services and is really required. Install this thing, and magically - no idea how (there is no ENV key introduced, so it executes an app to resolve credentials or smth?) - your python applications are authenticated some how after logging in using a web interface.
- it's big
- not normal api key usage
- requires a web browser
- user of API key needs to know credentials of the owner of the key, so forget installing by someone else.

Google is crazy. Why would you build smth so different than any other API supplier. Why reinvent the wheel in a bad way. How can so much of it be client side.

Only positive thing is - that it works immediately after being surprised that it was really such big app that allows you API access.

It's impossible that such big system is safer than just a simple API key.

Also, if it provides interface and is requirement for other google libs it's just a freaking SDK.

AI is more than just a model.

It's also tooling. Tooling can help to interpret data or solve a puzzle like Sudoku or parse a JSON file perfectly. Results of those tooling will be wrapped in AI response. That quality of tooling responses is high because it's made by classical code that works with literal data and outputs literal data. As long the competition of OpenAI doesn't have tooling like that, it won't be the same.

I do assume for now that DeepSeek doesn't have that. I tried it, it answers things well, but for bigger questions that would require tooling it just crashes and says it's too busy. So I can't verify 100%.

Will try again later and update under this Rant, but assume the DeepSeek stuff is very over hyped. To know what DeepSeek really is about without watching all the fake fan video's, take this quite objective response of the maker of Perplexity AI. Someone that knows where he's talking about: (40 minutes) https://youtube.com/watch/....

So when it comes to investments of a model, what does the stuff investment is incuded in? I mean, OpenAI was way more expensive than DeepSeek but DeepSeek borrowed all OpenAI's research that was made by very expensive processes. So DeepSeek didn't pay research costs like OpenAI did. Also it (I still assume) didn't spend money on tooling.

Also, i'm sure a less woke API would be way cheaper because it doesn't have to lie to himself causing it to keep reasoning until his given woke fact makes sense! Wokism destroys models, i'm sure.

I didn't check DeepSeek on wokism yet, but it's based op GPT4o so, probably it is.

But competition is always great, I can't imagine the price would even drop further for AI requests but if it does, it would be amazing. Maybe it also becomes free and we will be forced to pay to use it without adverts.

Got a bit distracted today. Instead of doing what I've wanted, I created the most comfortable way to make a site ever. I'm serious. It beats all those static site generators.

Hmm, could even make a static site generator of this one. Won't do, no benefit.

It's by using markdown and it has support for syntax highlighting of many languages. A website could be literary:

# My blog

## some code

```c
printf("This is my C code");
```.

And you already have something decent. See the 'Get started' section. You have a site running in no time. See also the Python database example at the bottom, it shows you how to use the internal sqlite3 database in your 'static' site and how to create a visitor counter!

Sky's the limit.

I came up with the idea by teaching my next door neighbor HTML and that was succesfull. She is able to navigate trough files and folders and edit a file like most people. Then I realized, if it was markdown, it would be really user friendly for anyone to use. To create a decent site, you only need to make a stylesheet for someone. By using # ## ### #### headings, it will always be in your designed style.

People won't fuck your site up like they do with WYSIWYG normally.

The concept is so simple, I only see advantages and it could be used for small and big content sites. It doesn't do caching on purpose, overkill and it's more comfortable this way.

Here is project link: https://molodetz.nl/retoor/dreamii

Oh Jase, you're gonna do that regular account deletion stuff all over again? You know that you can just delete a rant right? I'm reasoning like a LLM and then I comment in bad English because my temperature is set too high. I hope that the kangaroo comment didn't piss you off. I just assume that every Australian has one to use instead of public transport. Camels with autism. Sad, all that effort for nothing.

Again, got done request via Instagram to upvote a hotel for 5,- dollars with immediate pay. The hotel is already given. You only have to post positive 5 star review and you get your 5 bucks.
OK, now, at one point, I want to know if it's really true. It's just interesting. But the thing is, you'll be returning a screenshot whit your full name and gave them the fact that you're naive to cooperate. So, is maybe that's the scam? Figuring out who is that naive person in real life to do further scammed directed at the person? Because if they don't pay (or just anyway) you can delete your recommendation.

I won't do it because I don't like their payment method. I asked them to do it by bank what is a lie - they will still get your name. Else I would've done it with a duplicate Google account to stay anonymous.

Sigh, it keeps me busy to figure out actual scams. I always engage to find out but I think I understand this one.

I also could just make a screenshot of someone else his recent recommendation. I tell the person I will do it sometime later, wait for someone else to post and claim it was me. In that case I need a crypto wallet for the payout. Hmm. Not very ethical.

I was so bored with work in the end. It was more administration than programming. So, i kinda quitted full application development. The thing is, it's expected that you use some existing framework. First of all - they never work how you want and the programming part of your work is mainly solving the limitations the framework brings without hacking too much. You keep within the boundaries of the framework. Besides that - since all fun stuff is already done by the framework builders all you have to do left is kinda administration. Field here, field there, rest call here. Extremely boring. When you've setup the base good, there's no challenge anymore, just producing windows and input forms.

Now, a few days ago, I started to make a clone of rocket chat. I use minimalist http framework (aiohttp) and you have to build most features yourself on top. Same for the ORM, i use dataset which does schema synchronisation for you but doesn't come with models. So i made a complete model / mapper entity framework on top of that. I made one single validation system that applies on models, forms and frontend validation. There's only one truth of valid data. Within the models, services, mappers and forms there's always the services variable available making it possible to fetch any data from any object. Never weird exceptions has to be done to get data. The implemented global LRU cache system is super in auto synchronizing the objects, don't have to do anything manually.

Finally software development of a full product is fun again. If you know how to do it - making your own framework is way easier than an existing one. On top of that, it's more advanced. I do understand that frameworks are aiming to be a bit minimal to be multi purpose, but with that attitude in mind, they still achieve to make it annoying as fuck.

Regarding time, it's just a few days of development. That's nothing for something that does exactly what you want. We have to drop the use-a-framework-because-it-is-stupid-to-do-yourself mentality. We should be programmers again! Not administrators! It's not weird that chatGPT can do so much of our jobs, our projects became lame.

The most productive way of development is going slow and don't make a lot mistakes. Lately, I'm doing the opposite. It's often not your skills or knowledge but just your patience / expectation to be managed. Since I switched from C to Python I got the mindset that everything should be possible to make in short time with low effort. Being used to do a lot of effort because of C development gave me a lot of patience and upgraded me to be a better dev at any language. But sadly, I notice it's a skill you can lose. How to retrain my patience. Doing a lot of C again. But I want to have the patience NOW (retoor said impatiently). What a paradoxale dilemma shit. You have to do patient stuff to be become patient whey really sucks and sucks if you became inpatient.

Very underrated skill that actually says a lot about you as programmer: patience.

With enough patience you can master anything. Without it, you're depending on tools that allow you to. That's OK, but there's always some price to be paid. Many time it's the ignorance not knowing how the tools work and thus how it actually works. Some people can live with that, others don't. Knowing how somethings works is relative anyway. Do you know how it works until the language level? Kernel level? Framework level? Everyone chooses themselves what the limit is I guess. My limit is the C api / kernel interface.

Random thoughts. I'm just bit frustrating performing low lately and speculating about what to do about it. I just don't have a different hobby than programming. Doing something else for a while to give brain a rest would be best but it should be something raising dopemine. TV / gaming doesn't do that for me.

Sigh, how to do absolutely nothing. I did it for a 1,5 year full time but that was because of medication. Now it's impossible.

Just wrote an email to Rocket Chat with subject "Dear commercial hungry bastards" and complimented them with raping the f out of one of the best chat applications there is. With decent apps and stuff. For people who don't know, Rocket Chat is a Slack-variant. Once almost the same. But it was completely self hosted and under YOUR control. Now, if you host it locally, you get a screwed up design that you can unlock by paying.. If two users register on y our rocket chat instance, you get a commercial message of them regarding the growing team. Also, i have by default some remote user of them named rocketcat in my chat. Wtf, who wants that ffs? And again! SELF HOSTED SERVICE. Imagine, they have servers under their control and don't even have to do the hosting. A big win-win for them I guess.

Idiots found themselves a marketing team apparently.

AI, weirdest business ever. Master of none technology. The only people who can decently monetize it (the suppliers) are not doing it properly ironically.

I'm working for a few days now full time on AI and test a lot. Costs: 97 cents. 50 cents from that was one error I made (expensive code execution call).

So in reality bashing as hard as a human can costed me 47 cents. Not weird that they don't make profit.

What is the intention behind AI usage? Do they want to break the internet by allowing such mass production of calls? What use cases are there to execute thousands of calls? I can only imagine bad things / use for abuse.

Can't imagine nothing good since most AI output is not good for real life applications. For example, you ask it to respond with a 1 or 0. Then the fucker says something like: "A zero, can I do more for you?". Thanks, my application can't interpret that. Should I do another call with your answer to find out if it is positive or negative? Ok, let's do and please answer with a yes or no! It will probably respond with "It's not every positive.".

From example above, OpenAI is not the worst in it but still. It's kinda useless for many things. You can't really count on it. As long AI doesn't output exactly what you want, decent automation with AI is not possible making all claims of replacing people with it completely worthless.

AI is like someone educated with zero working experience and as pragmatic as Fred Flintstone's car.

Anyway, guess who did quit at Microsoft and upgraded to work for molodetz and likes to talk now?

See here our new CEO: https://retoorded.molodetz.nl/. And yes, he has a company car. He get's to drive me :P

The data at the bottom are statistics regarding my key presses. It's literally every key pressed on this laptop since 2024-12-08. Since that date I entered a total of unique 925450 unique inputs. I did 4751951 keyboard inputs.

I know from 595 hours exactly what i've done for tasks (described by LLM based on my keylog data).

I type 107 lines per hour on average (return presses) based on 595 hours. With that logic, i did around 63925 lines.

I'm not very happy with the statistics, especially not because backspace is a hardcore first. Now, while i'm typing i'm focusing on how much I use it and it's not a lot at all.

But the thing is, if you remove abcdef, you have one a, one b, but six times back space. And these are real presses - not keyboard repeats. Also abcdef will be counted by the tag counter as a whole. Everything is a tag until it sees a new line or a white space or some punct.

Funny is that there are completely different keys on the list than I expected. You're so you used to those keys that you don't even notice using them.

I'm almost considering to add a sound under the backspace button to teach myself WHEN i use it and try to avoid it.

The key logger database is now 346Mb. Some overhead because every keypress takes around 40 chars of description (timestamp, press type, char, input device).

Creating statistics for the tags (unique words typed) takes several minutes. Already rewriting that part to C. The stats are made by python, the key logs with C.

I'm just shocked, I used 144644 times a key that I think not to use that much? :P How retoorded can you be. Imagine if i actually fixed typo's :P

But based on these keys you can see that i'm mainly working in terminal / vim. The 'i' for insert for example, typed so many times. The 'x' for save+quit. The '0' to go to beginning of line.

Did you expect that these buttons would've been the most used?

#0 BACKSPACE is pressed 144644 times (15.63% of total input)
#1 UP is pressed 92711 times (10.02% of total input)
#2 LEFT_SHIFT is pressed 73777 times (7.97% of total input)
#3 ENTER is pressed 63883 times (6.9% of total input)
#4 DOWN is pressed 56838 times (6.14% of total input)
#5 TAB is pressed 43635 times (4.72% of total input)
#6 RIGHT is pressed 37710 times (4.07% of total input)
#7 SPACE is pressed 34438 times (3.72% of total input)
#8 LEFT is pressed 26800 times (2.9% of total input)
#9 LEFT_CTRL is pressed 25402 times (2.74% of total input)
#10 LEFT_ALT is pressed 17289 times (1.87% of total input)
#11 I is pressed 12856 times (1.39% of total input)
#12 X is pressed 6106 times (0.66% of total input)
#13 A is pressed 5163 times (0.56% of total input)
#14 0 is pressed 4487 times (0.48% of total input)
#15 PAGEDOWN is pressed 4151 times (0.45% of total input)

Ok, it costed me a whole night but I finally made the perfect OpenAI agent manager for all purposes using Python!

It's open source and a finished product. It's tested quite well. I will use this as base for my perfect working assistant that communicates trough the wireless JBL Go! speaker on my desk. It reacts to everything it hears until I ask it to go to sleep until I tell it to wake up. That's the mute mode. I never have to touch or click anything. 24/7 active.

It can be an assistant, but also a companion like Replika. Replika is normally very expensive, with my library nothing and it has great benefits like a perfect memory. Original Replika is a goldfish in comparison to this one.

It's also possible to create a custom RAG within minutes!

Check it out, it was never easier: https://molodetz.nl/retoor/ragent

I have made an interactive talking AI but it's not open source. It contains passwords/keys and tasks that are personal. But, a lot was learned while the code is nearly nothing. I spend many hours on research and didn't want to let it go to waste.

If you are interested in TTS / STT, this will be a nice resource: https://molodetz.nl/retoor/...

Side note: the builtin webkit TTS/STT engine is maybe even better and has a great API! Amazed by the quality of that thing.

This is python research. I hope that I can motivate someone but devRant is always empty on Saturday.

If someone needs help with an implementation regarding this, you know where to find me.