[ddingus]

Interesting!!

I'm not slamming you, or your experience or preferences. That said, I find very little difference between 2 seconds and one second or less than that.

Search taking a small fragment of time just isn't a big deal and I search sometimes many times per day.

What is the gain for you that makes 2 seconds an exception to using the product?

Just curious. Peace, live well.

[__jonas]

I feel like search is probably one of the most common things I do with a computer, I would say I often search many times per minute if I’m actively looking into some topic or issue.

Because it’s such a common action for me, it feels like such a strong regression to go 2/3 times slower than before.

[ddingus]

Thanks. I have a couple of thoughts to share:

Back in the day, IBM did a usability study. An application requiring the user to specify operations and fields for data input was setup two ways:

One way was manual, lots of clicks, and or input sequences. Each one did a specific thing quickly.

The other way was highly automated and the user was only required to click a few times. The tasks were the same. This way had more flow, fewer discrete commands, more functionality woven together.

To their surprise the users felt they were more productive with the software they clicked more, despite the automated version being less work and the workflow more efficient.

I bet the effective reduction in your search is minor, but the delays do accumulate and demand attention.

Given that, a small change to your flow may well change things!

What you should do is rapidly input your first queries and then as they appear, drill down on those, and when that appears, start to eliminate dead query windows, or drop fresh queries into them.

What you prefer, to use a car analogy, is one that corners like no other. Then you find yourself in one that lacks corner cases.

So you maximize your time in lane, straight road, batching the corner driving and flooring it on the straights.

I used to experience a similar thing running a browser on IRIX vd NT. The NT browser was quicker to respond where the IRIX one would delay a little and then just render it all quick

I just started working with a few windows. Changed everything. I would be typing in new queries while one I waited for was about to render.

It was a change from rapid fire to a more batch mode. Soon, I rarely had to think and my flow was fast all around.

I put this shared experience here in the hope you may be inclined to try different things.

[temporarara]

But do you seriously think you should get serious results in one second? I would understand the complaints about slowness if searches were taking like ten seconds, but a few seconds, I really must be getting old.

I still remember when google was giving relevant results in page 2. Now it's pretty much useless for me, and the fast search makes me think they are throwing away tons of potentially good stuff just to make it fast (and place more ads and rubbish scam/ai sites).

[chipsrafferty]

Kagi is pretty much useless past the first 3 results.

[MrOxiMoron]

I haven't found that, but even if that is true, it saves you tons of time having to filter out all the sponsored results and get to the things that actually give you the information you where searching for.

[temporarara]

That's not my experience, but I guess that depends on the use case and how specific things you are searching. However, in general I much prefer good search results to saving a few seconds and getting delivered absolute rubbish like google does these days, that was the main point of my post. Like I said I still remember when just plain google search without any thought on how to word the search was phenomenal.

Google fully pivoted from prestigious tech-company to lame ad-company and that's a shame really. Business as usual.

[mthoms]

It doesn’t seem intuitive at first but this is a well researched phenomenon.

https://glinden.blogspot.com/2006/11/marissa-mayer-at-web-20...