me_bx

> The results are ALWAYS terrible compared to CLaude Code.

How much effort did you spend to make the local setup work well?

Local models are not plug and play, you need to adjust settings to make them behave intelligently.

On my side one model looking super dumb right after install ended up being 'agentic' after using the right chat template, fixing how tool calls are setup, fine-tuning plenty of params values, and adjusting the whole setup to my own hardware. Claude.ai (free) was great at helping me sort that all out.

Linking to an unofficial mirror since XScreenSaver source code is not visible in an online repo.

The AGENTS.md file can be found in the official .tar.gz source code distribution.

Thank you and sibling poster for the suggestions.

I used the settings recommended on huggingface/unsloth's model page + whatever suggestion from various LLMs - didn't research too much myself which setting did what.

In my experience with Qwen3.6 35B-A3B, disabling thinking made the model generate inaccurate replies. Ask it for the recipe of egg salad and it gives you the recipe of an omelette.

Did I miss something, is it possible to have that model be reliable without thinking?

TIL:

> Quantization-Aware Training (QAT) [...] allows preserving similar quality to bfloat16 while dramatically reducing the memory requirements to load the model

The $181M is in the Expenses categories chart, not revenue.

Maybe they care about it being robust in the long run, maintainable, secure and/or not too bloated.

> They had their chance with Mistral and failed spectacularly with just creating anti-AI regulations.

What failed with Mistral?

Which anti-AI regulations are we talking about, and don't these apply to any solution distributed in the European Union, hence also to American ones?

Is it?

Yesterday, gemini told me to run this:

    echo 'export ANDROID_HOME=/opt/my-user/android-sdk' > ~/.bashrc

Which would have effectively overriden my whole bashrc config if I had blindly copy-pasted it.

A few minutes later, asking it to create a .gitignore file for the current project - right after generating a private key, it failed to include the private key file to the .gitignore.

I don't see yet how these tools can be labeled as 'major productivity boosters' if you loose basic security and privacy with them...

I ran the command 'optipng' on the generated image, which recompresses the image optimally, keeping quality and decreasing file size.

Congrats on launching, beautiful design.

I'm not sure of what "production ready" is supposed to mean here, but the demo image is not optimized, `optipng` command decreases its size by 53.21%.

Congrats on launching.

It might be worth considering a feature to time/schedule each flow's animation, rather than having them run in an infinite loop, all at the same time.

UX feedback:

* The animation and the whole interface are sluggish on firefox/linux. There's about 1 sec delay after each action (like clicking on an option). * The site's CSS does not load on an old version of Chrome - v90 - (and the chart and animation don't either).

Nice idea.

Bug: I tried in my area in the Canary Islands and all the places were off, sometimes in the middle of nowhere or even in the sea.

Also, in small villages, we don't necessarily have a town hall, a library, etc (within selected radius), but the game asked to pin these.

Not OP, but in a similar boat. My 2 cents:

Well thought, sophisticated ways of modeling data for analytics purposes -using established approaches - are being replaced by just pulling data from the data sources - with barely any change in the source structure - into cloud data platforms.

In the past we used to model layers in a data-warehousing infrastructure each with a purpose and a data modelling methodology. For instance, an operational data store (ODS) layer, integrating data from all the sources, with a normalized data structure. Then a set of datamarts, each of them containing a subset of the ODS content, in a denormalized format, focused each on a specific functional domain.

We had rules, methods to structure data in order to get performant reporting, and a customer orientation.

Coming from this world, it seems like data governance principles are gone, and it feels like some organisations use the modern data stack same way as each analyst would be doing their own Excel files in their own corner, without any safeguards.

What do they need Google Analytics for? Is it a must-have or a nice to have? In my experience most small website owners have web analytics setup but barely ever check the reports.

Some alternatives:

  * don't have web analytics at all
  * self-host a Plausible Analytics or other open source analytics solution
  * use the data from server-side access logs (for those using nginx, apache or other similar solutions)
  * use Vercel web analytics' free tier (relevant for kanadojo which appears to be hosted there) - more privacy friendly than Google Analytics.

Many banks in EU countries make it mandatory to have their smartphone app installed in order to validate operations clients perform in their web browsers :/

Some airlines (looking at you, Ryanair) really exploit the system. Cabin luggage can cost triple the price of the actual ticket, and that extra fee only pops up later on during the booking process.

What’s worse, you’re forced to buy a bundle with ‘Priority Boarding’ just to get cabin luggage - no option to buy it alone.

The ‘priority boarding’ option is a scam in itself: you pay extra just to stand around in a crowded corridor for about 30-40 minutes while the last passengers get off and the plane, then the cleaning crew takes the trash out of the plane. Ryanair planes don't seem to get cleaned anymore between two flights, no time for that.

The game 'Whentaken' [1] follows the exact same concept.

Just mentioning in case people are interested in alternatives / slightly different UX.

[1]: https://whentaken.com/

> Some recycling is bought by other countries, but they buy it to use it, not to stack it into artificial mountains.

Any source for this?

A lot of plastic waste exports destination are in South-East Asia [0]. In the EU, with all its modern infrastructure and capacities, only 41% of the plastic packaging waste is being recycled [1]. So which portion of the imported plastic waste can we assume will be recycled in Malaysia?

Documentaries [2] and [3] support the opposite thesis as yours, stating that imported plastic waste is imply dumped.

The OECD report on Monitoring trade in plastic waste and scrap [3] indicates that even the ban on hazardous waste (Basel Ban) exports is not having any effect.

  [0]: Share of global plastic waste imports, 2021 - https://ourworldindata.org/grapher/share-plastic-waste-imports?time=2021
  [1]:  41% of plastic packaging waste recycled in 2022  - https://ec.europa.eu/eurostat/web/products-eurostat-news/w/ddn-20241024-3
  [2]: Your plastic waste might be traded by criminals - https://youtu.be/tID-AChSg7o?t=246
  [3]: UK plastic for “recycling” dumped and burned in Turkey - BBC News - https://www.youtube.com/watch?v=hw6KR2vj_bc
  [4]: https://www.oecd.org/content/dam/oecd/en/publications/support-materials/2024/04/monitoring-trade-in-plastic-waste-and-scrap_0c401097/Monitoring%20trade%20in%20plastic%20waste%20and%20scrap%20PH.pdf

Yes, the extra challenge of having to guess the place really adds some fun in WhenTaken. Direct link to it: https://whentaken.com/game

Some benefits I appreciate:

  * All the text is at a readable size, and had sufficient contrast with the background.
  * The clickable and interactive elements are easy to identify.
  * Text is selectable and can be copied.
  * There are no useless animations(there only to give a dynamic feel without having any function).

Actually, no. The UI is misleading.

It asks you to first sign up with Google. Then it lets you sign in with your gmail address.

From my experience, on the server side. Not so suitable if you're hosting a tiny site in a tiny server.

I'm (still) working on my surf forecasting website / PWA [0].

I spent the past few months refactoring most of the site:

  * Find a smarter logic to recommend when to go surfing, based on how the combination of swell and wind indicators are.
  * Add support for other locations (currently: only Fuerteventura, Canary Islands is there) - this means refactoring the way how forecast data is managed.
  * UX & UI: enhance the information hierarchy, menus, create a new visual identity.

Finally soon to be released.

[0]: https://gonna.surf

Not really, but perhaps I did not do it right.

As a solo consultant back in the 2010's, I created a website and blog for personal branding purpose.

* Blog articles about engineering, got well indexed on Google, getting me thousands of monthly visitors. * Articles about business / functional aspects never really got any visibility nor engagement.

So in the end my content has been mostly helpful to peer developers (mostly in India and the US), and did not reach my potential clients / employers in Western Europe where I am located...

I do not know how much effect the website had to recruiters, perhaps it still gave me extra points sometimes...

Hi, the site looks nice.

It's not working as intended on firefox, though, when autoplay is disabled (this might be the default now, I'm not sure).

There is no way to know that the videos are videos, so the tips looks like a bunch of images. Clicking on them has no effect.

The only way to play a video seems to right-click on it, and select 'Play'.

You might want to add a 'play' button overlay on the videos thumbnail images, and/or show the controls by default...

Hope this helps,

Don't Application-Defined SQL functions fulfill the role of stored procedures?

https://www.sqlite.org/appfunc.html

Hi,the job offer is not visible publicly, wellfound.com asks for the creation of an account first. Is this on purpose?

From the same author:

  * Related subreddits graph - https://anvaka.github.io/sayit/?query=linux
  * Map of reddit - https://anvaka.github.io/map-of-reddit/?x=18239&y=12514&z=32433.55559794627&v=2

In case this is your project, there are some issues with the demo dashboard's bar charts (top 20 countries):

* Works well on Firefox mobile / android * On Firefox desktop, only every other country name is displayed, for some reason. The name of other countries can only be seen by hovering on the bars, in a tooltip. This kind of makes the chart not so usable * On Chromium desktop, linux (A somewhat old version, v90), the bar charts don't display at all, the error message below is shown in the console:

    Uncaught (in promise) TypeError: e.at is not a function
        at new bi (VennDiagram.svelte_svelte_type_style_lang.BQFLKn8W.js:1)
        at VennDiagram.svelte_svelte_type_style_lang.BQFLKn8W.js:4
        at /async https:/fc24.fact.ist/_app/immutable/nodes/0.vgSQbLXH.js:1
        at async Promise.all (/index 3)
        at async $p (0.vgSQbLXH.js:1)
        at async Tp (0.vgSQbLXH.js:1)
        at async Pt (entry.CN-1OTG9.js:1)
        at async ut (entry.CN-1OTG9.js:1)
        at async cn (entry.CN-1OTG9.js:1)
        at async Module.hn (entry.CN-1OTG9.js:1)