Skip to content

Piero Bosio Social Web Site Personale Logo Fediverso

Social Forum federato con il resto del mondo. Non contano le istanze, contano le persone

Something in #VSCode just wrote a #JupyterNotebook in my name, seemingly based on the name I gave to the #Python project.

Uncategorized
1 1 9
  • Something in just wrote a in my name, seemingly based on the name I gave to the project. When I saw that the file existed, I assumed it was just some “Hello, world” like thing.

    Eventually I looked at it.
    It was extensive had lots of sentences like, “I haven't yet decided how to represent digits larger than 9”, which either are not true or non-sensical for my project.

    I am pleased that I could this before that ever made it to a public repository. And I am extremely annoyed that I could have published some AI generated text in my name.

    I’m not in principle opposed to using AI to assist in development. After all, linters are AI as far as I am concerned. But something that writes crap in my name without explicitly drawing my attention to the fact that it is doing so is not the kind of thing that will win me over to AI.

  • oblomov@sociale.networkundefined oblomov@sociale.network shared this topic on

Gli ultimi otto messaggi ricevuti dalla Federazione
Post suggeriti
  • 0 Votes
    1 Posts
    15 Views
    cross-posted from: https://lemmy.dbzer0.com/post/55501944 Hey, I’ve been kicking around an idea for a bot: it would look for fanfiction links shared in a server and keep track of which ones get shared the most. The concept: Track sites like AO3, FanFiction.net, ScribbleHub, etc. Count how often each link gets posted Commands to see the “top” links or which domains are tracked It’s just a rough idea, I haven’t built it yet. Curious if anyone thinks this would actually be useful or has tips for implementing it without overcomplicating things. import re import json import discord from collections import Counter from discord.ext import commands TOKEN = "YOUR_BOT_TOKEN" intents = discord.Intents.default() intents.message_content = True bot = commands.Bot(command_prefix="!", intents=intents) # Domain list you want to track (lower-case) TRACK_DOMAINS = { "archiveofourown.org", "fanfiction.net", "forum.questionablequesting.com", "forums.spacebattles.com", "forums.sufficientvelocity.com", "webnovel.com", "hentai-foundry.com", "scribblehub.com", } link_pattern = re.compile(r'https?://\S+') link_counter = Counter() def domain_of_url(url: str) -> str | None: try: # Extract domain part from urllib.parse import urlparse parsed = urlparse(url) domain = parsed.netloc.lower() # remove leading “www.” if domain.startswith("www."): domain = domain[4:] return domain except Exception: return None def save_links(): with open("links.json", "w") as f: # convert counts to dict json.dump(dict(link_counter), f) def load_links(): try: with open("links.json") as f: data = json.load(f) for link, cnt in data.items(): link_counter[link] = cnt except FileNotFoundError: pass @bot.event async def on_ready(): load_links() print(f"Bot is ready. Logged in as {bot.user}") @bot.event async def on_message(message): if message.author.bot: return links = link_pattern.findall(message.content) for link in links: dom = domain_of_url(link) if dom in TRACK_DOMAINS: link_counter[link] += 1 await bot.process_commands(message) @bot.command(name="links") async def links(ctx, top: int = 10): if not link_counter: await ctx.send("No links recorded.") return sorted_links = sorted(link_counter.items(), key=lambda x: x[1], reverse=True) display = sorted_links[:top] lines = [f"{link} — {count}" for link, count in display] await ctx.send("**Top links:**\n" + "\n".join(lines)) @bot.command(name="domains") async def domains(ctx): """Show which domains are tracked.""" await ctx.send("Tracked domains: " + ", ".join(sorted(TRACK_DOMAINS))) @bot.command(name="dump") async def dump(ctx): """For admin: dump full counts (might be large).""" if not link_counter: await ctx.send("No data.") return lines = [f"{link} — {cnt}" for link, cnt in sorted(link_counter.items(), key=lambda x: x[1], reverse=True)] chunk = "\n".join(lines) # Discord message length limit; you may need to split await ctx.send(f"All counts:\n{chunk}") @bot.event async def on_disconnect(): save_links() bot.run(TOKEN)
  • 0 Votes
    1 Posts
    11 Views
    No posting yet, but my team at chainguard.dev is looking to hire someone to work on AI/ML packages; think nVidia bits like CUDA, PyTorch, Tensorflow, etc. This is more like OS/distro work than "doing things with LLMs". This is my fourth week here, but so far so good: https://www.chainguard.dev/careers Poke me for info!Presumably "Senior" to "Staff" level folks, but if you're more junior than that and interested, you should still poke us, we're happy to mentor.#fedihired #getfedihired #python
  • 0 Votes
    1 Posts
    11 Views
    🚀 Huge thanks to @sixfeetup for being a Platinum sponsor of DjangoCon US 2025! Their deep Python & cloud expertise helps teams build incredible web solutions. Excited to have them join us this week in Chicago!https://sixfeetup.com/ 🎟️ In-person & online tickets: https://ti.to/defna/djangocon-us-2025#DjangoCon #Python #Django
  • 0 Votes
    1 Posts
    14 Views
    Released Webware for #Python 3.1.0 https://webwareforpython.github.io/w4py3/