this post was submitted on 15 Aug 2023
179 points (97.4% liked)

Asklemmy

43810 readers
1229 users here now

A loosely moderated place to ask open-ended questions

Search asklemmy 🔍

If your post meets the following criteria, it's welcome here!

  1. Open-ended question
  2. Not offensive: at this point, we do not have the bandwidth to moderate overtly political discussions. Assume best intent and be excellent to each other.
  3. Not regarding using or support for Lemmy: context, see the list of support communities and tools for finding communities below
  4. Not ad nauseam inducing: please make sure it is a question that would be new to most members
  5. An actual topic of discussion

Looking for support?

Looking for a community?

~Icon~ ~by~ ~@Double_A@discuss.tchncs.de~

founded 5 years ago
MODERATORS
 

Title says it all. I'll go first:

I don't really have any on my computer (all I use that for is Vim, Firefox, and Git), but on my phone: Orbot (basically Tor as a VPN on your phone).

Edits: Added link, fixed formatting

you are viewing a single comment's thread
view the rest of the comments
[–] Dark_Arc@social.packetloss.gg 3 points 1 year ago* (last edited 1 year ago) (1 children)

Prices seem high

They talk about their prices better here: https://blog.kagi.com/update-kagi-search-pricing. Including:

Q. What is your cost per search? A. Currently it is 1.25 cents per search. But search indexes are about to become several times more expensive and we now have the AI integration cost. We will try to keep the cost at 1.5 cents per search or below by rethinking our partnering strategy and investing more into our infrastructure.


“Boosting” and blocking domains: basically permanent filters.

I mean, yes and no; it's not quite a filter as it can tweak the ranking not just outright put something at the top or remove it.

“Lenses” - Similar to SearXNG’s different tabs, but more like a profile style system.

These are actually super cool because they put things like the old "site:reddit.com" trick on steroids and https://help.kagi.com/kagi/features/code.html is like site:stackoverflow.com on steroids.

Bangs are the same as in SearXNG, and I think you can do the same, or similar, thing in Firefox.

You definitely can do this in plenty of things; I don't think it's all that compelling really, but it's a nice feature none the less, and it's arguably easier to sync "bangs" than it is to sync these shortcuts between browsers and different software; but yeah, not a big motivator for me.

Custom CSS is cool and all but you can get that with a browser add-on (like Stylus).

Similar comment about the syncing of this; but yeah, this also wasn't a big motivator for me.

Their “Comprehensive result filtering” seems that same as every other browser

True, but I think it's there 1 for completeness, and 2 because the count of ads and trackers as a criteria is pretty unique to them. It's also quite a bit more user facing, and easier to work with than say, getting the same options from Google search.

“Widgets” - Seems like Google’s quick answers, user created ones are cool I guess.

Not sure what you mean by custom ones, haven't seen anything about that; but it's definitely in the feature parity category.


Don't forget things like redirects (change your lemmy UI?) and their own index, which includes things like wayback machine results -- particularly relevant with CNET deleting old content to boost their Google ranking.

Their generated summaries are also pretty cool, e.g. from the CNET article:

  • CNET has been deleting thousands of old articles from its site in recent weeks in an effort to improve its search engine rankings on Google.
  • CNET confirmed it removes old content to appear more "fresh, relevant and worthy of being placed higher than our competitors" according to an internal memo.
  • However, experts say Google does not encourage deleting old content simply because of its age and that older content can still provide value to users.
  • The perception that "the Internet is forever" and content placed online will always remain accessible has proven untrue over time.
  • Link rot threatens old web content as links go dead and content disappears from original locations.
  • Copyright trolls have pressured sites to remove old images rather than risk expensive lawsuits over fair use.
  • Large portions of the historical web record are missing or inaccessible now due to link rot and deliberate content removal.
  • Archives like the Wayback Machine help preserve some content but can't capture everything.
  • Sites focused on SEO are driven to extremes like mass content deletion to rise above noisy search results.
  • Archivists play an important role in preserving digital history by making copies of content before it disappears.

Additionally, it seems that they don’t have seem to have any external audits ... (correct me if I’m wrong).

They don't, but they do have a clearly worded privacy policy https://kagi.com/privacy.

or releasing of code (correct me if I’m wrong).

The search engine itself, no; but their browser extensions are open source: https://github.com/kagisearch/browser_extensions


This actually looks very cool, so many new backup solutions to try after this thread. 🥲

I've tried a lot of different backup solutions... I'd be curious where you're leaning, but I'd say this one is likely going to be your winner for ease of use, privacy, and cross-platform functionality.

[–] hai@lemmy.ml 1 points 1 year ago (1 children)

https://blog.kagi.com/update-kagi-search-pricing

I gotta hand it to them for being transparent about their pricing, including that limit feature... AWS should take some notes.

These are actually super cool because they put things like the old “site:reddit.com” trick on steroids and https://help.kagi.com/kagi/features/code.html is like site:stackoverflow.com on steroids.

I'm assuming you're a programmer so I gotta ask, how much easier does this make it for your to find documentation and answers to that one error that the language never decided to document for some reason?

It’s also quite a bit more user facing, and easier to work with than say, getting the same options from Google search.

Google does always seem to hide them for some reason, despite the fact that it would be easy to work into the interface.

Not sure what you mean by custom ones, haven’t seen anything about that; but it’s definitely in the feature parity category.

In this post they say: "We have many instant answers already and are constantly adding more, with the idea of allowing users to define their own ‘widgets’ in the future." I'm assuming it's not released yet, but it does sound like a cool feature when they do.

Don’t forget things like redirects (change your lemmy UI?) and their own index, which includes things like wayback machine results – particularly relevant with CNET deleting old content to boost their Google ranking.

I didn't see either one of these features on the blog post, likely because it was from 2022, but pretty cool being able to search for old content that's been deleted.

The search engine itself, no; but their browser extensions are open source: https://github.com/kagisearch/browser_extensions

I stand corrected, that is nice that at least a bit of it is, I didn't see that on their website, granted I did miss the entire "Help" page somehow.

Their generated summaries are also pretty cool

Alright, so I created a trail account, and gave it a try with a random article, and hey, pretty cool, and would actually save time. I even tried it with some pretty simple documentation, it did end up just describing the steps that had to be taken. Although, I noticed a little "discuss this document" button, and asked it what code I'd need, it gave it to me. Then asked it to make tweaks like changing the key and sections to ones that actually would be in a configuration file, and it changed it. It could even do things beyond just the lines it's on, like instead of printing the value, make sure it's not a value. I gotta say, I actually am pretty impressed by that (even if it is relatively simple work), I know ChatGPT and Copilot could (likely) do the same things, but how it broke it down from an article is pretty cool.

Now, I'm likely not gonna become a paying customer for it, like I said in the original post -- I'm using (Neo)vim, and, while I've played around with AI like this, I don't find it ever really falling into my workflow. Maybe down the road, and if the internet keeps getting worse, it might be worth it. That being said you can do a lot of these things with alternative solutions, but a search engine that puts them all together is pretty cool, and does save some time.


I’ve tried a lot of different backup solutions… I’d be curious where you’re leaning, but I’d say this one is likely going to be your winner for ease of use, privacy, and cross-platform functionality.

Currently, I'm using Nextcloud, and I distro hop enough that all I need to do is backup my documents, which most are code and can go on Git(Lab/Tea/Hub). Although, I'm finally settling down (there's like one or two more distros that I want to try) with Arch Linux, and a proper backup solution would be nice, by plan is to just go through this entire thread, compare features, and try them out.

Thanks for explaining Kagi to me by the way, it is in fact a pretty cool service!