Replying to @⁨NewOldGuard@lemmy.ml⁩

Yahoo Japan is actually a completely separate company.

Yahoo itself is still extremely popular in the US, just with a different demographic than we (the sort of people on the fediverse) regularly interact with. Their sports, finance, and news sites are some of the most popular on the internet. I don’t think it’s particularly relevant in the web search space these days though, and hasn’t been for a long time.

I think we don’t hear much about it because it has found a niche and is staying within it. It’s profitable, but not on an “infinite growth forever” trajectory so it rarely makes the news. Honestly kind of respectable.

Replying to @⁨mesamunefire@piefed.social⁩

Yep, Jeeves was really popular for a while. I was an Alta Vista fan for a while too. It was one of my favorite early searches.

What I really miss about early web searching is the cached pages where you could see all your highlighted search terms exactly where they appear on the page. Modern google dumps you at the front door to the corporate walled garden and you usually have to search again with some shitty embedded search tool. Old Google would link you to a .htm or .html page with the exact answer.

Replying to @⁨Imgonnatrythis@sh.itjust.works⁩

The peak was during the period you could say “If you search on Google for ___ you get ___.” Back when it was predictable. When they started personalizing results was the moment they sliced society’s brake lines.

I remember…I had a Windows 8.1 machine, I needed to SSH into a Raspberry Pi. No Powershell, no native SSH client, the easiest way to get one was to install PuTTY. So I googled “putty”. All lowercase, search engines sometimes get fucky when you start doing caps and punctuation, etc. I got an entire page of results for and about the TTY client, and nothing about drywall spackle or non-newtonian children’s toys. Because I often search for stuff like that, it made a bet and went all in.

The horrifying part of that is, that innocuous example makes it sound great. I didn’t have to overcome the “why did you name it that?” problem and refine the search by adding “tty client” or something like that. By getting to know me it took me to exactly what I wanted to find. When trying to find utility software, that’s okay. That’s not all people search the web for.

Now it’s hooked up to the “You’re right to call me out on that” machine, we’re sliding down the mountainside upside down on fire.

Replying to @⁨sanitation@lemmy.today⁩

Open discovery is a hard problem.

If that seems pseudo-science speak to seem smart, I mean that you need some prior trust. Just like with PGP. Some axioms. In year 2002 those axioms were people you meet IRL and links you saw printed. Now these have been erased as a Web discovery mechanism, so to not compete with search engines and platforms. Links are login-walled and short-living most of the time. People are discouraged from making their own pages and sharing links.

My Kademlia experiment is more working than not, slowly moving to split core and UI there, to have replaceable cores. Want to try F2F. Especially since it even looks more like something that would work F2F way better. I’ve learned that I’m not as smart as I think while making it. Still don’t have anything resembling indirect reputation, but eventually I want to make that, for some kind of “public” media, where you need to filter the sewer stream.

(No, I won’t share, it’s a P2P program written in TCL/Tk, entirely, not just UI, except for some bindings in C that weren’t done by someone else. It seemed easier to start with, with asynchronous style being very easy there, but I’ve recently tried writing some other thing in Python and realized just how easier and faster Python is due to everything not being a string, however boring that critique of TCL would seem.)

Replying to @⁨eleitl@lemmy.zip⁩

That’s overstating it. There’s a reason Google took over: it was better pretty much all around. The original Google algorithm was a really amazing tool for finding quality content on the web as it existed in those days. That was before SEO drove all these focused efforts on gaming the system.

With that said, Alta Vista definitely was the most promising tool before Google, and it had some potential if it hadn’t been so poorly managed by DEC. They simply saw it as a cool demo for their Alpha server platform, and never had any vision for it as a meaningful product on its own.

Replying to an earlier post

KSF is the world standard for the surf mini game in Counter Strike. It is primarily played on Counter Strike Source and is completely free. They have a global leaderboard where you can compete in getting the fastest times on maps against everyone. It’s an incredibly welcoming and friendly community (in my experience). You can find out more about how to connect at their site here https://ksf.surf/ :)

Replying to an earlier post

it’s not just google

for the past year or so I’ve been telling people that DDG is so much better as it actually returns the results you’re looking for, which Google stopped doing. except in the past few months, the same thing is happening there now too.

I don’t know if it’s fuckery with the search engines, people gaming SEO, ads, or if it’s just the sheer quantity of shit that’s on the internet now, but it’s at least half of the time now that I go to search something and it’s not in the top half dozen results. and sometimes I can’t even find it at all, despite knowing that it definitely exists out there.

Replying to an earlier post

DDG been shit for years. I went for it when their campaign was about not being in q bubble and getting unfiltered results. For yesrs everything that came back was hyper local. Search a name you get a doctor or history or spelling correction of someone in my town of 30,000 people or their ancestor. Turn off location puts something generic in top place and then everything else after as normal.

Replying to an earlier post

It’s the web in general. Pages that have been meticulously altered to score high on the latest search engines so they get the most visibility. I remember several times where the top results were obviously LLM generated text scraped from multiple other pages. Tons and tons of text with absolutely zero substance on the topic I was trying to learn about.

I switched to using Startpage as my search engine now. It works great. It feels like being transported back in time. It gets the results I want. Less websites full of AI generated fluff articles.