I always think the worst thing that can happen to a (mild) hoarder like myself is that one of the things I'm hoarding comes in handy. Using one hoarded thing justifies years' worth of more hoarding.

I had an interesting experience after I stopped reading HN comments for two weeks. I think it had to do with not reading the doom and gloom on here, which I tend to agree with. But after two weeks of reading neither the hype nor the doom about LLMs, I realized how amazing (though not perfect) this tool is. My point isn't whether LLMs are good or bad, but just how I didn't realize how much my consumption of other people's opinions infected my own outlook. It was like the opposite of rose-colored glasses.

BTW, the tool I used to block HN is called Foqos, which, of course, I heard about right here on HN. Based on my screen time reports, I got back about 2 hours per day average after blocking all my favorite sites.

Obviously just speculation, but maybe don't let AI write your code...

Microsoft CEO says up to 30% of the company’s code was written by AI https://techcrunch.com/2025/04/29/microsoft-ceo-says-up-to-3...

It's degraded availability of Git operations.

The enterprise cloud in EU, US, and Australia has no issues.

If you look at the incident history disruptions happen often in the public cloud for years already. Before AI wrote code for them.

The enterprise cloud runs on older stable versions of GitHub's backend/frontend code.

That sounds very bad, but I guess it depends also on which code it is. And whether Nadella actually knows what he's talking about, too.

Maybe AI is the tech support too

Sweet. 30% of Microsoft's code isn't protected by copyright.

Time to leak that.

I admit it was a bit ridiculous. However, if Microsoft is going to brag about how much AI code they are using but not also brag about how good the code is, then we are left to speculate. The two outages in two weeks are _possible_ data points and all we have to go on unless they start providing data.

I am trying out something along those lines with friendlyfriends.community (nod to Do Androids Dream of Electric Sheep). Still pretty basic though, but feel free to give it a try.

Edit: I'm exploring the idea that social media is not really social. How much different is it to interact with a human intermediated by social media sites than it is to interact with chatbots?

https://flexbox.help/ I get a lot of people saying it was very useful.

Oh I like this a lot. This a helpful little tool.

Here's an alternative I recently came across. In Firefox on the network tab of the dev console, right click on an XHR request. You'll have two relevant options: "Resend" and "Edit and Resend". Edit and Resend doesn't give you all the features of Postman, but it is useful and a lot easier. I have spent a lot of time in Postman just putting in all the headers and POST body to set up the API call. With this method in Firefox, all parameters are already set because it's an exact copy of the request you already sent. You can change params or just hit send when you're ready to retest something you're doing on the backend.

I am very happy to save you a lot of time for copy&pasting requests to Postman. Since Postman can import curl commands, just right-click on a command in Firefox, Copy as -> Curl, click on the Import in Postman, import as text, paste curl, mission accomplished. Hope it helps.

They actually removed the plain Resend option and renamed Edit and Resend to just Resend.

I made canweimage.com. It can't replace all the features of Google, but it can fit the bill if you just need a basic search of Creative Commons.

Thanks! But just to be clear, it isn’t somehow searching all things under a CC license, but just things within wikimedia right?

Last time I checked, Creative Commons had their own search for things under their licenses. But now apparently that project got transferred to WordPress and is now named Openverse: https://wordpress.org/openverse/?referrer=creativecommons.or...

Anyways, I'd argue that's the most comprehensive database of CC-licensed works.

Thanks for this.

Also, I just checked that flickr.com still allows you to filter their searches by license.

This looks great. Care to share the API call used for searhing?

https://commons.wikimedia.org/w/api.php?action=query&generat... search term here>&format=json

(see https://en.wikipedia.org/w/api.php?action=help&modules=query for reference)

edit: formatting

Amazing!

I notice a slight difference between

https://commons.wikimedia.org/w/api.php?action=query&generat...

and searching for 'cat' on canweimage.com

Is there some query processing needed?

I use these additional parameters: `&gsrlimit=25&prop=imageinfo|pageimages&iiprop=url|size`. I think it just changes how much and what type of data is returned, but maybe that could be the difference?

Depending on the time of year you go, make sure you get a reservation: https://www.nps.gov/arch/planyourvisit/permitsandreservation...

Check out tomdn.com. I created it to make it easier to get to MDN. E.g., instead of googling array.map, just type tomdn.com?array.map. It will take you straight to MDN docs on Array.prototype.map. Other things work as well, like tomdn.com?object.keys, tomdn.com?css.color, tomdn.com?htmlel.button, etc.

If there is a pattern it doesn't recognize, you end up on the MDN search results page.

edit: More info. about how it works is available here: https://github.com/tayler/tomdn

What value does this provide over simply Googling or DDG'ing "array.map mdn"? It's shorter and the first result for me is (unsurprisingly) MDN on both, and I don't have to run my searches through an unknown third party.

DDG also has !bangs for this specific purpose and !mdn does, as expected, lead you to MDN.

Valid questions. I just think it's faster. One action on tomdn.com instead of 1. DDG "array.map mdn", 2. Find MDN result (usually first), 3. Click result, 4. See docs.

Bangs are great, just shorter again to use tomdn.com: - `!mdn array.map` leads me to https://developer.mozilla.org/en-US/search?q=array.map, which I have to scan the results and click on. - tomdn.com?array.map leads me to the exact page in the docs. https://developer.mozilla.org/en-US/docs/Web/Javascript/Refe...

This may or may not be valuable, but I like saving the extra steps for something I do many times a week. Basically this calculation (https://xkcd.com/1205/) works out for me + the fun of working on a new project.

edit: >I don't have to run my searches through an unknown third party

Very sympathetic to this concern. Feel free to host yourself -- it's a single html file. https://github.com/tayler/tomdn

`! mdn array.map` takes you right to MDN

Totally had no idea. Thanks for the tip. I guess tomdn.com is for non-ddg users only.