Restfox [1] is worth checking out. It's fully offline and lets you version control your collections with git or any sync tool you prefer. The postman import is well tested and the app also allows you to export back to Postman collections.
The main limitation of web components for me is that all data needs to be passed to the component attributes as strings. Wish we could pass objects and arrays without JSON.stringify-ing them.
Overall search quality has been declining on all search engines. Maybe there's too much spam. Saw an entertaining video about it yesterday that echoes how I feel when I google stuff: https://www.youtube.com/watch?v=jrFv1O4dbqY. It's so hard to find content written by humans these days. Seems like only the top sites are being indexed.
I'm in the same boat, I feel like my search-fu is being thwarted with the ever growing list of products that coopt existing words rather than coining new terms. We're in this ever expanding word overloading mode and I think the commercial and marketing spaces are now dominating search to drown out the useful hits that would previously rise to the top.
More and more, our search-fu is being actively thwarted. Tools to tailor our search have been steadily taken away: exclusion/inclusion operators, verbatim search ignored, and on and on.
> exclusion/inclusion operators, verbatim search ignored, and on and on.
I've been assuming that this was just a bug with DDG and not a deliberate attempt to cripple their search engine. I've seen it with Google results too, but I could buy that Google would break them, since making websites harder to find encourages people to pay for prominently placed ads, but what would DDG's motivation be?
I want. a search engine that remembers my preferences, and I can then click an ignore entire domain option, and never see that domain in search results even again. is there such a thing?
I'mma have to look into this thing, there's an ever growing need for this sort of thing as the incentives drift further and further from the end user https://kagi.com/.
Google's attempt to introduce zero-click results has been nothing short of catastrophic. The sheer volume of bullshit they are spreading rivals anything that ChatGPT could ever hope to generate.
A couple of weeks ago, I was debating with someone about what "LMR" stood for in the context of cable specifications, such as LMR-240, LMR-400 and so on. I thought it meant "Land Mobile Radio" while the other person disagreed that it stood for anything. A Google search on LMR coax cable acronym returned a helpful info blurb stating that LMR stood for "Last Minute Resistance" as a means of fending off sexual assault.
Needless to say there was no way to tell exactly what site Google had copied that definition from, and no useful way to provide feedback to them. Sometimes there's a "Feedback" link, this time there wasn't. Sometimes the feedback link is present but only offers the option of reporting illegal activity. That option wasn't present either.
For whatever reason, Google clearly does not give a flying fuck at a rolling donut about search quality anymore. With the right leadership, Bing could own that entire line of business, in a manner reminiscent of IE's original dominance over Netscape. I'm not holding my breath, but at this point I'm cheering for anyone who can offer Google some competition.
I tried the example you cited. The "blurb" is called a snippet; the snippet comes from the web page itself. One of the pages had that actual text, which is why it appeared. Why it had it on a page that's primarily about coaxial cable isn't clear, but we'll look into how to improve.
As for sending feedback, each link in the results has a little three dot icon next to it that brings up our "About This Results" panel, and you can send feedback that way.
Also, to belatedly introduce myself, I'm the public liaison for search at Google. It's a position we have within the actual search engineering team to help us gather feedback to improve search quality. Feel free for you or anyone comfortable sharing examples of unhelpful results to flag me about it:
> As for sending feedback, each link in the results has a little three dot icon next to it that brings up our "About This Results" panel, and you can send feedback that way.
Surely you don't expect us to believe this ever gets read, much less acted upon? Don't insult the intelligence of your audience.
I asked ChatGPT 4.0 what LMR meant as applied to coaxial cable. It gave me an excellent response, including that LMR is a trademark of Times Microwave Systems.
I rarely use search engines anymore. I'll bet the same is true for many people.
In fact when I used regular Google, the results are good enough for me to deduce that on my own.
So I consider GP's search skills inadequate. I mean it's not exactly wrong to desire a tool that handholds you and feeds you the answer; but if you are willing to do a little bit of deduction Google is fine.
Regular Google itself tells you that. At least for me, when I tried this search right now, the first snippet says "These letters indicate the brand. LMR is a trademark of Times Microwave, whereas RFC is made by Shireen." It's the second snippet that says the last minute resistance thing, and it appears to be from someone's personal blog that blocks IPs from the United States, so without logging into a VPN, I can't open the site to figure out why it thinks this.
It appears to indeed mean nothing, but of course, it took me a whole two minutes or so to find an actually authoritative source. Everyone knows no real user will ever do that and just wants a search engine or chatbot to dictate reality to them.
I don't trust the zero-click results at all any more. I recently had to fill out a form for the DMV that required my county code. To find out, I of course googled it, and wrote down the knowledge box result. This turned out to be the wrong number, and thanks to Google, this happens so frequently the clerk knew where I got it, and knew to check it and fix it.
People are building workarounds in real life due to how bad Google's results have gotten.
Similar experience for me. One that I do a fair amount is searching for specs on various car makes & models. The zero click results are typically quite bad. To be fair, a lot of it just ends up being bad source data, but their primary source for automotive specifications is the same shitty site.
Another problem is that Google seems to ignore a significant part of the words you type into the search bar.
If you type 'word1 word2 word3', where word3 is less common than word1 and word2, a lot of the time, it will act as if word3 simply wasn't in the query.
ddg does this too with annoying frequency. Even quoting it does not necessarily guarantee it will appear in the snippet, much less the result.
And when DDG runs out of "web links" it just fills the rest of result pages with local results that are out-of-place and useless. Like, getting "Visit Paris" sites after searching for rare computer parts.
I definitely intend to add web socket support eventually. Meanwhile, there's https://socketfox.dev which I made for my friends. You can use it to test web sockets.
Thank you. Will look into preventing the right click requirement and provide an alternative for that.
It should be saving your requests automatically. Maybe your browser is blocking IndexedDB somehow? That's what's used by the application to store the data locally. Also do make sure you're not in incognito mode, as changes will be lost once you're out of it.
Firefox with ublock origin should not cause any issues. The issue might be caused by Firefox's private window implementation which does not allow IndexedDB operations, which is what Restfox uses for storing the data of the application locally.
Disclaimer: I maintain it.
[1] https://github.com/flawiddsouza/Restfox